{
  "best_global_step": 107424,
  "best_metric": 0.7989339828491211,
  "best_model_checkpoint": "saves_multiple/prompt-tuning/llama-3-8b-instruct/train_math_qa_123_1760637721/checkpoint-107424",
  "epoch": 20.0,
  "eval_steps": 6714,
  "global_step": 134280,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0007447125409591898,
      "grad_norm": 16.875,
      "learning_rate": 8.936550491510277e-06,
      "loss": 2.0541,
      "num_input_tokens_seen": 2976,
      "step": 5
    },
    {
      "epoch": 0.0014894250819183796,
      "grad_norm": 11.4375,
      "learning_rate": 2.0107238605898124e-05,
      "loss": 2.1465,
      "num_input_tokens_seen": 5920,
      "step": 10
    },
    {
      "epoch": 0.002234137622877569,
      "grad_norm": 7.3125,
      "learning_rate": 3.1277926720285965e-05,
      "loss": 1.6496,
      "num_input_tokens_seen": 8896,
      "step": 15
    },
    {
      "epoch": 0.002978850163836759,
      "grad_norm": 7.28125,
      "learning_rate": 4.244861483467382e-05,
      "loss": 1.4091,
      "num_input_tokens_seen": 11744,
      "step": 20
    },
    {
      "epoch": 0.0037235627047959487,
      "grad_norm": 1.6484375,
      "learning_rate": 5.361930294906166e-05,
      "loss": 0.6603,
      "num_input_tokens_seen": 14496,
      "step": 25
    },
    {
      "epoch": 0.004468275245755138,
      "grad_norm": 6.59375,
      "learning_rate": 6.478999106344951e-05,
      "loss": 0.8271,
      "num_input_tokens_seen": 17440,
      "step": 30
    },
    {
      "epoch": 0.005212987786714328,
      "grad_norm": 13.4375,
      "learning_rate": 7.596067917783735e-05,
      "loss": 0.8999,
      "num_input_tokens_seen": 20352,
      "step": 35
    },
    {
      "epoch": 0.005957700327673518,
      "grad_norm": 6.1875,
      "learning_rate": 8.71313672922252e-05,
      "loss": 0.9275,
      "num_input_tokens_seen": 23328,
      "step": 40
    },
    {
      "epoch": 0.006702412868632708,
      "grad_norm": 7.125,
      "learning_rate": 9.830205540661304e-05,
      "loss": 0.7633,
      "num_input_tokens_seen": 26304,
      "step": 45
    },
    {
      "epoch": 0.0074471254095918975,
      "grad_norm": 17.375,
      "learning_rate": 0.0001094727435210009,
      "loss": 0.9321,
      "num_input_tokens_seen": 29184,
      "step": 50
    },
    {
      "epoch": 0.008191837950551088,
      "grad_norm": 9.8125,
      "learning_rate": 0.00012064343163538874,
      "loss": 0.8288,
      "num_input_tokens_seen": 32544,
      "step": 55
    },
    {
      "epoch": 0.008936550491510277,
      "grad_norm": 11.875,
      "learning_rate": 0.0001318141197497766,
      "loss": 0.7911,
      "num_input_tokens_seen": 35520,
      "step": 60
    },
    {
      "epoch": 0.009681263032469467,
      "grad_norm": 31.125,
      "learning_rate": 0.00014298480786416443,
      "loss": 0.9028,
      "num_input_tokens_seen": 38400,
      "step": 65
    },
    {
      "epoch": 0.010425975573428656,
      "grad_norm": 7.65625,
      "learning_rate": 0.00015415549597855229,
      "loss": 0.7724,
      "num_input_tokens_seen": 41408,
      "step": 70
    },
    {
      "epoch": 0.011170688114387846,
      "grad_norm": 5.875,
      "learning_rate": 0.00016532618409294012,
      "loss": 0.7992,
      "num_input_tokens_seen": 44352,
      "step": 75
    },
    {
      "epoch": 0.011915400655347037,
      "grad_norm": 4.78125,
      "learning_rate": 0.00017649687220732795,
      "loss": 0.7498,
      "num_input_tokens_seen": 46912,
      "step": 80
    },
    {
      "epoch": 0.012660113196306225,
      "grad_norm": 13.625,
      "learning_rate": 0.00018766756032171581,
      "loss": 0.8477,
      "num_input_tokens_seen": 49472,
      "step": 85
    },
    {
      "epoch": 0.013404825737265416,
      "grad_norm": 5.78125,
      "learning_rate": 0.00019883824843610365,
      "loss": 0.8396,
      "num_input_tokens_seen": 52096,
      "step": 90
    },
    {
      "epoch": 0.014149538278224605,
      "grad_norm": 5.8125,
      "learning_rate": 0.0002100089365504915,
      "loss": 0.82,
      "num_input_tokens_seen": 54944,
      "step": 95
    },
    {
      "epoch": 0.014894250819183795,
      "grad_norm": 6.875,
      "learning_rate": 0.00022117962466487934,
      "loss": 0.7666,
      "num_input_tokens_seen": 57664,
      "step": 100
    },
    {
      "epoch": 0.015638963360142984,
      "grad_norm": 7.9375,
      "learning_rate": 0.0002323503127792672,
      "loss": 0.8512,
      "num_input_tokens_seen": 60448,
      "step": 105
    },
    {
      "epoch": 0.016383675901102176,
      "grad_norm": 5.5,
      "learning_rate": 0.00024352100089365507,
      "loss": 0.8179,
      "num_input_tokens_seen": 63808,
      "step": 110
    },
    {
      "epoch": 0.017128388442061365,
      "grad_norm": 7.125,
      "learning_rate": 0.00025469168900804287,
      "loss": 0.7675,
      "num_input_tokens_seen": 66464,
      "step": 115
    },
    {
      "epoch": 0.017873100983020553,
      "grad_norm": 7.09375,
      "learning_rate": 0.00026586237712243073,
      "loss": 0.7479,
      "num_input_tokens_seen": 69856,
      "step": 120
    },
    {
      "epoch": 0.018617813523979745,
      "grad_norm": 9.375,
      "learning_rate": 0.0002770330652368186,
      "loss": 0.8035,
      "num_input_tokens_seen": 72768,
      "step": 125
    },
    {
      "epoch": 0.019362526064938934,
      "grad_norm": 9.375,
      "learning_rate": 0.0002882037533512064,
      "loss": 0.7734,
      "num_input_tokens_seen": 75616,
      "step": 130
    },
    {
      "epoch": 0.020107238605898123,
      "grad_norm": 10.5625,
      "learning_rate": 0.00029937444146559426,
      "loss": 0.7944,
      "num_input_tokens_seen": 78656,
      "step": 135
    },
    {
      "epoch": 0.02085195114685731,
      "grad_norm": 4.75,
      "learning_rate": 0.0003105451295799821,
      "loss": 0.8096,
      "num_input_tokens_seen": 81760,
      "step": 140
    },
    {
      "epoch": 0.021596663687816504,
      "grad_norm": 5.03125,
      "learning_rate": 0.00032171581769437,
      "loss": 0.7637,
      "num_input_tokens_seen": 84928,
      "step": 145
    },
    {
      "epoch": 0.022341376228775692,
      "grad_norm": 2.421875,
      "learning_rate": 0.0003328865058087578,
      "loss": 0.845,
      "num_input_tokens_seen": 88096,
      "step": 150
    },
    {
      "epoch": 0.02308608876973488,
      "grad_norm": 5.96875,
      "learning_rate": 0.00034405719392314565,
      "loss": 0.7076,
      "num_input_tokens_seen": 90912,
      "step": 155
    },
    {
      "epoch": 0.023830801310694073,
      "grad_norm": 20.0,
      "learning_rate": 0.0003552278820375335,
      "loss": 0.8118,
      "num_input_tokens_seen": 93696,
      "step": 160
    },
    {
      "epoch": 0.024575513851653262,
      "grad_norm": 44.5,
      "learning_rate": 0.0003663985701519213,
      "loss": 0.9098,
      "num_input_tokens_seen": 96992,
      "step": 165
    },
    {
      "epoch": 0.02532022639261245,
      "grad_norm": 6.65625,
      "learning_rate": 0.0003775692582663092,
      "loss": 0.7708,
      "num_input_tokens_seen": 99744,
      "step": 170
    },
    {
      "epoch": 0.026064938933571643,
      "grad_norm": 5.9375,
      "learning_rate": 0.00038873994638069704,
      "loss": 0.767,
      "num_input_tokens_seen": 102432,
      "step": 175
    },
    {
      "epoch": 0.02680965147453083,
      "grad_norm": 27.125,
      "learning_rate": 0.0003999106344950849,
      "loss": 0.7739,
      "num_input_tokens_seen": 105440,
      "step": 180
    },
    {
      "epoch": 0.02755436401549002,
      "grad_norm": 9.3125,
      "learning_rate": 0.0004110813226094727,
      "loss": 0.8483,
      "num_input_tokens_seen": 108320,
      "step": 185
    },
    {
      "epoch": 0.02829907655644921,
      "grad_norm": 7.71875,
      "learning_rate": 0.00042225201072386057,
      "loss": 0.7432,
      "num_input_tokens_seen": 111264,
      "step": 190
    },
    {
      "epoch": 0.0290437890974084,
      "grad_norm": 6.5,
      "learning_rate": 0.00043342269883824843,
      "loss": 0.6927,
      "num_input_tokens_seen": 114432,
      "step": 195
    },
    {
      "epoch": 0.02978850163836759,
      "grad_norm": 5.71875,
      "learning_rate": 0.00044459338695263624,
      "loss": 1.0435,
      "num_input_tokens_seen": 117248,
      "step": 200
    },
    {
      "epoch": 0.03053321417932678,
      "grad_norm": 14.125,
      "learning_rate": 0.0004557640750670241,
      "loss": 0.8675,
      "num_input_tokens_seen": 120512,
      "step": 205
    },
    {
      "epoch": 0.03127792672028597,
      "grad_norm": 12.125,
      "learning_rate": 0.00046693476318141196,
      "loss": 0.7546,
      "num_input_tokens_seen": 123776,
      "step": 210
    },
    {
      "epoch": 0.032022639261245156,
      "grad_norm": 5.09375,
      "learning_rate": 0.0004781054512957998,
      "loss": 0.7107,
      "num_input_tokens_seen": 126880,
      "step": 215
    },
    {
      "epoch": 0.03276735180220435,
      "grad_norm": 9.5625,
      "learning_rate": 0.0004892761394101877,
      "loss": 0.9462,
      "num_input_tokens_seen": 130112,
      "step": 220
    },
    {
      "epoch": 0.03351206434316354,
      "grad_norm": 15.375,
      "learning_rate": 0.0005004468275245754,
      "loss": 0.8655,
      "num_input_tokens_seen": 132704,
      "step": 225
    },
    {
      "epoch": 0.03425677688412273,
      "grad_norm": 12.6875,
      "learning_rate": 0.0005116175156389634,
      "loss": 0.8188,
      "num_input_tokens_seen": 135968,
      "step": 230
    },
    {
      "epoch": 0.03500148942508192,
      "grad_norm": 4.59375,
      "learning_rate": 0.0005227882037533513,
      "loss": 0.7882,
      "num_input_tokens_seen": 138752,
      "step": 235
    },
    {
      "epoch": 0.035746201966041107,
      "grad_norm": 9.6875,
      "learning_rate": 0.000533958891867739,
      "loss": 0.8318,
      "num_input_tokens_seen": 141792,
      "step": 240
    },
    {
      "epoch": 0.036490914507000295,
      "grad_norm": 5.4375,
      "learning_rate": 0.0005451295799821269,
      "loss": 0.7506,
      "num_input_tokens_seen": 144640,
      "step": 245
    },
    {
      "epoch": 0.03723562704795949,
      "grad_norm": 7.96875,
      "learning_rate": 0.0005563002680965147,
      "loss": 0.7676,
      "num_input_tokens_seen": 147648,
      "step": 250
    },
    {
      "epoch": 0.03798033958891868,
      "grad_norm": 188.0,
      "learning_rate": 0.0005674709562109025,
      "loss": 0.8602,
      "num_input_tokens_seen": 150432,
      "step": 255
    },
    {
      "epoch": 0.03872505212987787,
      "grad_norm": 7.3125,
      "learning_rate": 0.0005786416443252904,
      "loss": 0.7636,
      "num_input_tokens_seen": 153312,
      "step": 260
    },
    {
      "epoch": 0.03946976467083706,
      "grad_norm": 7.6875,
      "learning_rate": 0.0005898123324396783,
      "loss": 0.8312,
      "num_input_tokens_seen": 156288,
      "step": 265
    },
    {
      "epoch": 0.040214477211796246,
      "grad_norm": 106.0,
      "learning_rate": 0.0006009830205540662,
      "loss": 1.1462,
      "num_input_tokens_seen": 159136,
      "step": 270
    },
    {
      "epoch": 0.040959189752755434,
      "grad_norm": 15.5,
      "learning_rate": 0.0006121537086684539,
      "loss": 1.0654,
      "num_input_tokens_seen": 162144,
      "step": 275
    },
    {
      "epoch": 0.04170390229371462,
      "grad_norm": 15.375,
      "learning_rate": 0.0006233243967828418,
      "loss": 0.8154,
      "num_input_tokens_seen": 164832,
      "step": 280
    },
    {
      "epoch": 0.04244861483467382,
      "grad_norm": 13.125,
      "learning_rate": 0.0006344950848972297,
      "loss": 1.1192,
      "num_input_tokens_seen": 167744,
      "step": 285
    },
    {
      "epoch": 0.04319332737563301,
      "grad_norm": 12.5625,
      "learning_rate": 0.0006456657730116174,
      "loss": 0.8284,
      "num_input_tokens_seen": 170528,
      "step": 290
    },
    {
      "epoch": 0.043938039916592196,
      "grad_norm": 93.0,
      "learning_rate": 0.0006568364611260053,
      "loss": 1.1555,
      "num_input_tokens_seen": 173440,
      "step": 295
    },
    {
      "epoch": 0.044682752457551385,
      "grad_norm": 16.0,
      "learning_rate": 0.0006680071492403932,
      "loss": 0.7943,
      "num_input_tokens_seen": 176160,
      "step": 300
    },
    {
      "epoch": 0.045427464998510574,
      "grad_norm": 217.0,
      "learning_rate": 0.0006791778373547811,
      "loss": 2.7377,
      "num_input_tokens_seen": 178816,
      "step": 305
    },
    {
      "epoch": 0.04617217753946976,
      "grad_norm": 187.0,
      "learning_rate": 0.0006903485254691689,
      "loss": 2.2127,
      "num_input_tokens_seen": 181664,
      "step": 310
    },
    {
      "epoch": 0.04691689008042895,
      "grad_norm": 72.0,
      "learning_rate": 0.0007015192135835567,
      "loss": 1.075,
      "num_input_tokens_seen": 184384,
      "step": 315
    },
    {
      "epoch": 0.04766160262138815,
      "grad_norm": 18.125,
      "learning_rate": 0.0007126899016979446,
      "loss": 0.8721,
      "num_input_tokens_seen": 187072,
      "step": 320
    },
    {
      "epoch": 0.048406315162347335,
      "grad_norm": 17.25,
      "learning_rate": 0.0007238605898123323,
      "loss": 0.8984,
      "num_input_tokens_seen": 189920,
      "step": 325
    },
    {
      "epoch": 0.049151027703306524,
      "grad_norm": 27.375,
      "learning_rate": 0.0007350312779267203,
      "loss": 0.9418,
      "num_input_tokens_seen": 192608,
      "step": 330
    },
    {
      "epoch": 0.04989574024426571,
      "grad_norm": 15.25,
      "learning_rate": 0.0007462019660411082,
      "loss": 0.8673,
      "num_input_tokens_seen": 195744,
      "step": 335
    },
    {
      "epoch": 0.0506404527852249,
      "grad_norm": 37.25,
      "learning_rate": 0.000757372654155496,
      "loss": 0.9487,
      "num_input_tokens_seen": 198272,
      "step": 340
    },
    {
      "epoch": 0.05138516532618409,
      "grad_norm": 6.75,
      "learning_rate": 0.0007685433422698838,
      "loss": 0.9024,
      "num_input_tokens_seen": 201152,
      "step": 345
    },
    {
      "epoch": 0.052129877867143286,
      "grad_norm": 14.125,
      "learning_rate": 0.0007797140303842716,
      "loss": 0.8935,
      "num_input_tokens_seen": 203936,
      "step": 350
    },
    {
      "epoch": 0.052874590408102475,
      "grad_norm": 4.1875,
      "learning_rate": 0.0007908847184986595,
      "loss": 0.8749,
      "num_input_tokens_seen": 206976,
      "step": 355
    },
    {
      "epoch": 0.05361930294906166,
      "grad_norm": 17.875,
      "learning_rate": 0.0008020554066130472,
      "loss": 0.9481,
      "num_input_tokens_seen": 210048,
      "step": 360
    },
    {
      "epoch": 0.05436401549002085,
      "grad_norm": 11.25,
      "learning_rate": 0.0008132260947274352,
      "loss": 0.9795,
      "num_input_tokens_seen": 212896,
      "step": 365
    },
    {
      "epoch": 0.05510872803098004,
      "grad_norm": 2.609375,
      "learning_rate": 0.0008243967828418231,
      "loss": 0.9713,
      "num_input_tokens_seen": 215872,
      "step": 370
    },
    {
      "epoch": 0.05585344057193923,
      "grad_norm": 10.625,
      "learning_rate": 0.0008355674709562109,
      "loss": 0.7804,
      "num_input_tokens_seen": 218624,
      "step": 375
    },
    {
      "epoch": 0.05659815311289842,
      "grad_norm": 23.25,
      "learning_rate": 0.0008467381590705987,
      "loss": 0.9118,
      "num_input_tokens_seen": 221408,
      "step": 380
    },
    {
      "epoch": 0.057342865653857614,
      "grad_norm": 6.625,
      "learning_rate": 0.0008579088471849866,
      "loss": 0.8948,
      "num_input_tokens_seen": 224224,
      "step": 385
    },
    {
      "epoch": 0.0580875781948168,
      "grad_norm": 2.375,
      "learning_rate": 0.0008690795352993744,
      "loss": 0.8262,
      "num_input_tokens_seen": 227136,
      "step": 390
    },
    {
      "epoch": 0.05883229073577599,
      "grad_norm": 5.1875,
      "learning_rate": 0.0008802502234137622,
      "loss": 0.7863,
      "num_input_tokens_seen": 230272,
      "step": 395
    },
    {
      "epoch": 0.05957700327673518,
      "grad_norm": 2.515625,
      "learning_rate": 0.0008914209115281501,
      "loss": 0.9703,
      "num_input_tokens_seen": 233568,
      "step": 400
    },
    {
      "epoch": 0.06032171581769437,
      "grad_norm": 14.375,
      "learning_rate": 0.000902591599642538,
      "loss": 0.7932,
      "num_input_tokens_seen": 236384,
      "step": 405
    },
    {
      "epoch": 0.06106642835865356,
      "grad_norm": 5.90625,
      "learning_rate": 0.0009137622877569257,
      "loss": 0.9386,
      "num_input_tokens_seen": 239424,
      "step": 410
    },
    {
      "epoch": 0.06181114089961275,
      "grad_norm": 3.28125,
      "learning_rate": 0.0009249329758713136,
      "loss": 0.8366,
      "num_input_tokens_seen": 242240,
      "step": 415
    },
    {
      "epoch": 0.06255585344057193,
      "grad_norm": 356.0,
      "learning_rate": 0.0009361036639857015,
      "loss": 3.4608,
      "num_input_tokens_seen": 244800,
      "step": 420
    },
    {
      "epoch": 0.06330056598153112,
      "grad_norm": 181.0,
      "learning_rate": 0.0009472743521000893,
      "loss": 7.805,
      "num_input_tokens_seen": 247968,
      "step": 425
    },
    {
      "epoch": 0.06404527852249031,
      "grad_norm": 4.03125,
      "learning_rate": 0.0009584450402144771,
      "loss": 5.2863,
      "num_input_tokens_seen": 250720,
      "step": 430
    },
    {
      "epoch": 0.06478999106344951,
      "grad_norm": 3.96875,
      "learning_rate": 0.0009696157283288649,
      "loss": 0.8318,
      "num_input_tokens_seen": 253952,
      "step": 435
    },
    {
      "epoch": 0.0655347036044087,
      "grad_norm": 63.0,
      "learning_rate": 0.0009807864164432527,
      "loss": 1.1154,
      "num_input_tokens_seen": 256576,
      "step": 440
    },
    {
      "epoch": 0.06627941614536789,
      "grad_norm": 4.15625,
      "learning_rate": 0.0009919571045576407,
      "loss": 1.9513,
      "num_input_tokens_seen": 259488,
      "step": 445
    },
    {
      "epoch": 0.06702412868632708,
      "grad_norm": 20.375,
      "learning_rate": 0.0010031277926720286,
      "loss": 3.3339,
      "num_input_tokens_seen": 262112,
      "step": 450
    },
    {
      "epoch": 0.06776884122728627,
      "grad_norm": 14.625,
      "learning_rate": 0.0010142984807864164,
      "loss": 3.7901,
      "num_input_tokens_seen": 265120,
      "step": 455
    },
    {
      "epoch": 0.06851355376824546,
      "grad_norm": 5.5625,
      "learning_rate": 0.0010254691689008044,
      "loss": 2.2158,
      "num_input_tokens_seen": 267872,
      "step": 460
    },
    {
      "epoch": 0.06925826630920465,
      "grad_norm": 13.0625,
      "learning_rate": 0.0010366398570151921,
      "loss": 1.3114,
      "num_input_tokens_seen": 270464,
      "step": 465
    },
    {
      "epoch": 0.07000297885016384,
      "grad_norm": 25.625,
      "learning_rate": 0.00104781054512958,
      "loss": 1.6267,
      "num_input_tokens_seen": 273152,
      "step": 470
    },
    {
      "epoch": 0.07074769139112302,
      "grad_norm": 8.0625,
      "learning_rate": 0.0010589812332439678,
      "loss": 0.9056,
      "num_input_tokens_seen": 275968,
      "step": 475
    },
    {
      "epoch": 0.07149240393208221,
      "grad_norm": 10.3125,
      "learning_rate": 0.0010701519213583556,
      "loss": 1.0221,
      "num_input_tokens_seen": 278848,
      "step": 480
    },
    {
      "epoch": 0.0722371164730414,
      "grad_norm": 3.921875,
      "learning_rate": 0.0010813226094727436,
      "loss": 1.0297,
      "num_input_tokens_seen": 281664,
      "step": 485
    },
    {
      "epoch": 0.07298182901400059,
      "grad_norm": 3.90625,
      "learning_rate": 0.0010924932975871313,
      "loss": 1.0363,
      "num_input_tokens_seen": 285056,
      "step": 490
    },
    {
      "epoch": 0.07372654155495978,
      "grad_norm": 5.375,
      "learning_rate": 0.001103663985701519,
      "loss": 0.9797,
      "num_input_tokens_seen": 287872,
      "step": 495
    },
    {
      "epoch": 0.07447125409591898,
      "grad_norm": 5.5,
      "learning_rate": 0.001114834673815907,
      "loss": 0.9748,
      "num_input_tokens_seen": 290752,
      "step": 500
    },
    {
      "epoch": 0.07521596663687817,
      "grad_norm": 3.359375,
      "learning_rate": 0.0011260053619302948,
      "loss": 0.8764,
      "num_input_tokens_seen": 293824,
      "step": 505
    },
    {
      "epoch": 0.07596067917783736,
      "grad_norm": 4.6875,
      "learning_rate": 0.0011371760500446825,
      "loss": 0.9853,
      "num_input_tokens_seen": 296672,
      "step": 510
    },
    {
      "epoch": 0.07670539171879655,
      "grad_norm": 8.375,
      "learning_rate": 0.0011483467381590707,
      "loss": 0.9858,
      "num_input_tokens_seen": 299776,
      "step": 515
    },
    {
      "epoch": 0.07745010425975574,
      "grad_norm": 4.15625,
      "learning_rate": 0.0011595174262734585,
      "loss": 1.0254,
      "num_input_tokens_seen": 302720,
      "step": 520
    },
    {
      "epoch": 0.07819481680071493,
      "grad_norm": 13.875,
      "learning_rate": 0.0011706881143878462,
      "loss": 0.9879,
      "num_input_tokens_seen": 305472,
      "step": 525
    },
    {
      "epoch": 0.07893952934167411,
      "grad_norm": 2.453125,
      "learning_rate": 0.0011818588025022342,
      "loss": 0.7589,
      "num_input_tokens_seen": 308672,
      "step": 530
    },
    {
      "epoch": 0.0796842418826333,
      "grad_norm": 3.109375,
      "learning_rate": 0.001193029490616622,
      "loss": 1.1814,
      "num_input_tokens_seen": 311904,
      "step": 535
    },
    {
      "epoch": 0.08042895442359249,
      "grad_norm": 50.75,
      "learning_rate": 0.0012042001787310097,
      "loss": 1.0363,
      "num_input_tokens_seen": 314688,
      "step": 540
    },
    {
      "epoch": 0.08117366696455168,
      "grad_norm": 25.25,
      "learning_rate": 0.0012153708668453977,
      "loss": 2.9199,
      "num_input_tokens_seen": 317472,
      "step": 545
    },
    {
      "epoch": 0.08191837950551087,
      "grad_norm": 1128.0,
      "learning_rate": 0.0012265415549597854,
      "loss": 1.3287,
      "num_input_tokens_seen": 320128,
      "step": 550
    },
    {
      "epoch": 0.08266309204647006,
      "grad_norm": 16.5,
      "learning_rate": 0.0012377122430741734,
      "loss": 1.1768,
      "num_input_tokens_seen": 322944,
      "step": 555
    },
    {
      "epoch": 0.08340780458742925,
      "grad_norm": 5.59375,
      "learning_rate": 0.0012488829311885611,
      "loss": 0.9197,
      "num_input_tokens_seen": 325856,
      "step": 560
    },
    {
      "epoch": 0.08415251712838845,
      "grad_norm": 3.296875,
      "learning_rate": 0.001260053619302949,
      "loss": 1.0059,
      "num_input_tokens_seen": 328800,
      "step": 565
    },
    {
      "epoch": 0.08489722966934764,
      "grad_norm": 1.546875,
      "learning_rate": 0.0012712243074173369,
      "loss": 0.8212,
      "num_input_tokens_seen": 331424,
      "step": 570
    },
    {
      "epoch": 0.08564194221030683,
      "grad_norm": 3.046875,
      "learning_rate": 0.0012823949955317246,
      "loss": 0.8438,
      "num_input_tokens_seen": 334112,
      "step": 575
    },
    {
      "epoch": 0.08638665475126601,
      "grad_norm": 2.6875,
      "learning_rate": 0.0012935656836461126,
      "loss": 0.9073,
      "num_input_tokens_seen": 336992,
      "step": 580
    },
    {
      "epoch": 0.0871313672922252,
      "grad_norm": 3.03125,
      "learning_rate": 0.0013047363717605006,
      "loss": 0.9097,
      "num_input_tokens_seen": 340224,
      "step": 585
    },
    {
      "epoch": 0.08787607983318439,
      "grad_norm": 1.421875,
      "learning_rate": 0.0013159070598748883,
      "loss": 0.9657,
      "num_input_tokens_seen": 343040,
      "step": 590
    },
    {
      "epoch": 0.08862079237414358,
      "grad_norm": 13.125,
      "learning_rate": 0.001327077747989276,
      "loss": 0.7222,
      "num_input_tokens_seen": 345696,
      "step": 595
    },
    {
      "epoch": 0.08936550491510277,
      "grad_norm": 5.1875,
      "learning_rate": 0.001338248436103664,
      "loss": 0.9915,
      "num_input_tokens_seen": 348992,
      "step": 600
    },
    {
      "epoch": 0.09011021745606196,
      "grad_norm": 2.796875,
      "learning_rate": 0.0013494191242180518,
      "loss": 0.9974,
      "num_input_tokens_seen": 352032,
      "step": 605
    },
    {
      "epoch": 0.09085492999702115,
      "grad_norm": 1.4375,
      "learning_rate": 0.0013605898123324395,
      "loss": 0.7878,
      "num_input_tokens_seen": 354656,
      "step": 610
    },
    {
      "epoch": 0.09159964253798034,
      "grad_norm": 2.359375,
      "learning_rate": 0.0013717605004468275,
      "loss": 0.8325,
      "num_input_tokens_seen": 357504,
      "step": 615
    },
    {
      "epoch": 0.09234435507893952,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0013829311885612153,
      "loss": 0.849,
      "num_input_tokens_seen": 360256,
      "step": 620
    },
    {
      "epoch": 0.09308906761989871,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0013941018766756032,
      "loss": 0.9104,
      "num_input_tokens_seen": 363072,
      "step": 625
    },
    {
      "epoch": 0.0938337801608579,
      "grad_norm": 2.1875,
      "learning_rate": 0.001405272564789991,
      "loss": 0.8422,
      "num_input_tokens_seen": 366272,
      "step": 630
    },
    {
      "epoch": 0.0945784927018171,
      "grad_norm": 1.0625,
      "learning_rate": 0.0014164432529043787,
      "loss": 0.8472,
      "num_input_tokens_seen": 369376,
      "step": 635
    },
    {
      "epoch": 0.0953232052427763,
      "grad_norm": 2.796875,
      "learning_rate": 0.0014276139410187667,
      "loss": 1.0073,
      "num_input_tokens_seen": 372224,
      "step": 640
    },
    {
      "epoch": 0.09606791778373548,
      "grad_norm": 1.765625,
      "learning_rate": 0.0014387846291331547,
      "loss": 0.8092,
      "num_input_tokens_seen": 375232,
      "step": 645
    },
    {
      "epoch": 0.09681263032469467,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0014499553172475424,
      "loss": 0.9883,
      "num_input_tokens_seen": 378272,
      "step": 650
    },
    {
      "epoch": 0.09755734286565386,
      "grad_norm": 1.53125,
      "learning_rate": 0.0014611260053619304,
      "loss": 0.8369,
      "num_input_tokens_seen": 381152,
      "step": 655
    },
    {
      "epoch": 0.09830205540661305,
      "grad_norm": 1.078125,
      "learning_rate": 0.0014722966934763181,
      "loss": 0.8597,
      "num_input_tokens_seen": 384096,
      "step": 660
    },
    {
      "epoch": 0.09904676794757224,
      "grad_norm": 0.6484375,
      "learning_rate": 0.001483467381590706,
      "loss": 0.8207,
      "num_input_tokens_seen": 387296,
      "step": 665
    },
    {
      "epoch": 0.09979148048853143,
      "grad_norm": 0.474609375,
      "learning_rate": 0.0014946380697050939,
      "loss": 0.7681,
      "num_input_tokens_seen": 390656,
      "step": 670
    },
    {
      "epoch": 0.10053619302949061,
      "grad_norm": 3.890625,
      "learning_rate": 0.0015058087578194816,
      "loss": 1.0177,
      "num_input_tokens_seen": 393568,
      "step": 675
    },
    {
      "epoch": 0.1012809055704498,
      "grad_norm": 0.609375,
      "learning_rate": 0.0015169794459338694,
      "loss": 0.9052,
      "num_input_tokens_seen": 396320,
      "step": 680
    },
    {
      "epoch": 0.10202561811140899,
      "grad_norm": 0.78125,
      "learning_rate": 0.0015281501340482573,
      "loss": 0.8006,
      "num_input_tokens_seen": 399328,
      "step": 685
    },
    {
      "epoch": 0.10277033065236818,
      "grad_norm": 3.046875,
      "learning_rate": 0.001539320822162645,
      "loss": 0.8001,
      "num_input_tokens_seen": 402176,
      "step": 690
    },
    {
      "epoch": 0.10351504319332737,
      "grad_norm": 2.125,
      "learning_rate": 0.0015504915102770328,
      "loss": 0.8743,
      "num_input_tokens_seen": 405088,
      "step": 695
    },
    {
      "epoch": 0.10425975573428657,
      "grad_norm": 0.462890625,
      "learning_rate": 0.0015616621983914208,
      "loss": 0.8391,
      "num_input_tokens_seen": 408288,
      "step": 700
    },
    {
      "epoch": 0.10500446827524576,
      "grad_norm": 0.39453125,
      "learning_rate": 0.0015728328865058086,
      "loss": 0.8356,
      "num_input_tokens_seen": 411488,
      "step": 705
    },
    {
      "epoch": 0.10574918081620495,
      "grad_norm": 0.59375,
      "learning_rate": 0.0015840035746201965,
      "loss": 0.801,
      "num_input_tokens_seen": 414272,
      "step": 710
    },
    {
      "epoch": 0.10649389335716414,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0015951742627345845,
      "loss": 2.2374,
      "num_input_tokens_seen": 417184,
      "step": 715
    },
    {
      "epoch": 0.10723860589812333,
      "grad_norm": 0.380859375,
      "learning_rate": 0.0016063449508489723,
      "loss": 0.8344,
      "num_input_tokens_seen": 420096,
      "step": 720
    },
    {
      "epoch": 0.10798331843908252,
      "grad_norm": 0.50390625,
      "learning_rate": 0.0016175156389633602,
      "loss": 0.7791,
      "num_input_tokens_seen": 422944,
      "step": 725
    },
    {
      "epoch": 0.1087280309800417,
      "grad_norm": 0.62890625,
      "learning_rate": 0.001628686327077748,
      "loss": 0.9282,
      "num_input_tokens_seen": 425824,
      "step": 730
    },
    {
      "epoch": 0.10947274352100089,
      "grad_norm": 0.52734375,
      "learning_rate": 0.0016398570151921357,
      "loss": 0.7881,
      "num_input_tokens_seen": 428832,
      "step": 735
    },
    {
      "epoch": 0.11021745606196008,
      "grad_norm": 0.3046875,
      "learning_rate": 0.0016510277033065237,
      "loss": 0.8267,
      "num_input_tokens_seen": 432032,
      "step": 740
    },
    {
      "epoch": 0.11096216860291927,
      "grad_norm": 1.109375,
      "learning_rate": 0.0016621983914209115,
      "loss": 0.8291,
      "num_input_tokens_seen": 434752,
      "step": 745
    },
    {
      "epoch": 0.11170688114387846,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0016733690795352992,
      "loss": 0.7834,
      "num_input_tokens_seen": 437568,
      "step": 750
    },
    {
      "epoch": 0.11245159368483765,
      "grad_norm": 0.26171875,
      "learning_rate": 0.0016845397676496872,
      "loss": 0.8327,
      "num_input_tokens_seen": 440736,
      "step": 755
    },
    {
      "epoch": 0.11319630622579684,
      "grad_norm": 0.306640625,
      "learning_rate": 0.001695710455764075,
      "loss": 0.7988,
      "num_input_tokens_seen": 443456,
      "step": 760
    },
    {
      "epoch": 0.11394101876675604,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.0017068811438784627,
      "loss": 0.7976,
      "num_input_tokens_seen": 446464,
      "step": 765
    },
    {
      "epoch": 0.11468573130771523,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0017180518319928507,
      "loss": 0.8453,
      "num_input_tokens_seen": 449216,
      "step": 770
    },
    {
      "epoch": 0.11543044384867442,
      "grad_norm": 0.361328125,
      "learning_rate": 0.0017292225201072384,
      "loss": 0.7951,
      "num_input_tokens_seen": 452064,
      "step": 775
    },
    {
      "epoch": 0.1161751563896336,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0017403932082216266,
      "loss": 0.8149,
      "num_input_tokens_seen": 454912,
      "step": 780
    },
    {
      "epoch": 0.1169198689305928,
      "grad_norm": 0.6171875,
      "learning_rate": 0.0017515638963360143,
      "loss": 0.8142,
      "num_input_tokens_seen": 457984,
      "step": 785
    },
    {
      "epoch": 0.11766458147155198,
      "grad_norm": 0.33203125,
      "learning_rate": 0.001762734584450402,
      "loss": 0.8235,
      "num_input_tokens_seen": 460704,
      "step": 790
    },
    {
      "epoch": 0.11840929401251117,
      "grad_norm": 0.408203125,
      "learning_rate": 0.00177390527256479,
      "loss": 0.8177,
      "num_input_tokens_seen": 463552,
      "step": 795
    },
    {
      "epoch": 0.11915400655347036,
      "grad_norm": 0.314453125,
      "learning_rate": 0.0017850759606791778,
      "loss": 0.8046,
      "num_input_tokens_seen": 466400,
      "step": 800
    },
    {
      "epoch": 0.11989871909442955,
      "grad_norm": 0.2099609375,
      "learning_rate": 0.0017962466487935656,
      "loss": 0.806,
      "num_input_tokens_seen": 469312,
      "step": 805
    },
    {
      "epoch": 0.12064343163538874,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0018074173369079535,
      "loss": 0.8303,
      "num_input_tokens_seen": 472256,
      "step": 810
    },
    {
      "epoch": 0.12138814417634793,
      "grad_norm": 0.2265625,
      "learning_rate": 0.0018185880250223413,
      "loss": 0.788,
      "num_input_tokens_seen": 475104,
      "step": 815
    },
    {
      "epoch": 0.12213285671730711,
      "grad_norm": 0.28515625,
      "learning_rate": 0.001829758713136729,
      "loss": 0.8167,
      "num_input_tokens_seen": 477952,
      "step": 820
    },
    {
      "epoch": 0.1228775692582663,
      "grad_norm": 0.291015625,
      "learning_rate": 0.001840929401251117,
      "loss": 0.8106,
      "num_input_tokens_seen": 480704,
      "step": 825
    },
    {
      "epoch": 0.1236222817992255,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0018521000893655048,
      "loss": 0.8154,
      "num_input_tokens_seen": 483680,
      "step": 830
    },
    {
      "epoch": 0.1243669943401847,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0018632707774798925,
      "loss": 0.8137,
      "num_input_tokens_seen": 486720,
      "step": 835
    },
    {
      "epoch": 0.12511170688114387,
      "grad_norm": 0.2021484375,
      "learning_rate": 0.0018744414655942805,
      "loss": 0.8004,
      "num_input_tokens_seen": 489536,
      "step": 840
    },
    {
      "epoch": 0.12585641942210307,
      "grad_norm": 0.333984375,
      "learning_rate": 0.0018856121537086685,
      "loss": 0.7934,
      "num_input_tokens_seen": 492256,
      "step": 845
    },
    {
      "epoch": 0.12660113196306225,
      "grad_norm": 0.365234375,
      "learning_rate": 0.0018967828418230562,
      "loss": 0.7985,
      "num_input_tokens_seen": 495168,
      "step": 850
    },
    {
      "epoch": 0.12734584450402145,
      "grad_norm": 0.2392578125,
      "learning_rate": 0.0019079535299374442,
      "loss": 0.8383,
      "num_input_tokens_seen": 498240,
      "step": 855
    },
    {
      "epoch": 0.12809055704498062,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0019191242180518317,
      "loss": 0.7888,
      "num_input_tokens_seen": 501216,
      "step": 860
    },
    {
      "epoch": 0.12883526958593983,
      "grad_norm": 0.419921875,
      "learning_rate": 0.00193029490616622,
      "loss": 0.8048,
      "num_input_tokens_seen": 504352,
      "step": 865
    },
    {
      "epoch": 0.12957998212689903,
      "grad_norm": 0.23828125,
      "learning_rate": 0.0019414655942806074,
      "loss": 0.7945,
      "num_input_tokens_seen": 507168,
      "step": 870
    },
    {
      "epoch": 0.1303246946678582,
      "grad_norm": 0.21484375,
      "learning_rate": 0.0019526362823949954,
      "loss": 0.8243,
      "num_input_tokens_seen": 510112,
      "step": 875
    },
    {
      "epoch": 0.1310694072088174,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0019638069705093836,
      "loss": 0.8002,
      "num_input_tokens_seen": 512896,
      "step": 880
    },
    {
      "epoch": 0.13181411974977658,
      "grad_norm": 0.130859375,
      "learning_rate": 0.001974977658623771,
      "loss": 0.8292,
      "num_input_tokens_seen": 515744,
      "step": 885
    },
    {
      "epoch": 0.13255883229073578,
      "grad_norm": 0.203125,
      "learning_rate": 0.001986148346738159,
      "loss": 0.8324,
      "num_input_tokens_seen": 518624,
      "step": 890
    },
    {
      "epoch": 0.13330354483169496,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0019973190348525466,
      "loss": 0.797,
      "num_input_tokens_seen": 521824,
      "step": 895
    },
    {
      "epoch": 0.13404825737265416,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.002008489722966935,
      "loss": 0.8214,
      "num_input_tokens_seen": 524800,
      "step": 900
    },
    {
      "epoch": 0.13479296991361334,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0020196604110813226,
      "loss": 0.8126,
      "num_input_tokens_seen": 527712,
      "step": 905
    },
    {
      "epoch": 0.13553768245457254,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0020308310991957105,
      "loss": 0.8118,
      "num_input_tokens_seen": 531200,
      "step": 910
    },
    {
      "epoch": 0.1362823949955317,
      "grad_norm": 0.734375,
      "learning_rate": 0.002042001787310098,
      "loss": 0.7957,
      "num_input_tokens_seen": 534080,
      "step": 915
    },
    {
      "epoch": 0.13702710753649092,
      "grad_norm": 0.23828125,
      "learning_rate": 0.002053172475424486,
      "loss": 0.8202,
      "num_input_tokens_seen": 536800,
      "step": 920
    },
    {
      "epoch": 0.1377718200774501,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0020643431635388736,
      "loss": 0.8412,
      "num_input_tokens_seen": 539904,
      "step": 925
    },
    {
      "epoch": 0.1385165326184093,
      "grad_norm": 0.4453125,
      "learning_rate": 0.002075513851653262,
      "loss": 0.794,
      "num_input_tokens_seen": 542560,
      "step": 930
    },
    {
      "epoch": 0.1392612451593685,
      "grad_norm": 3.25,
      "learning_rate": 0.0020866845397676495,
      "loss": 0.9754,
      "num_input_tokens_seen": 545696,
      "step": 935
    },
    {
      "epoch": 0.14000595770032767,
      "grad_norm": 0.4765625,
      "learning_rate": 0.0020978552278820375,
      "loss": 0.8658,
      "num_input_tokens_seen": 548640,
      "step": 940
    },
    {
      "epoch": 0.14075067024128687,
      "grad_norm": 0.431640625,
      "learning_rate": 0.0021090259159964255,
      "loss": 0.801,
      "num_input_tokens_seen": 551296,
      "step": 945
    },
    {
      "epoch": 0.14149538278224605,
      "grad_norm": 0.310546875,
      "learning_rate": 0.002120196604110813,
      "loss": 0.7974,
      "num_input_tokens_seen": 554176,
      "step": 950
    },
    {
      "epoch": 0.14224009532320525,
      "grad_norm": 0.27734375,
      "learning_rate": 0.002131367292225201,
      "loss": 0.8134,
      "num_input_tokens_seen": 556928,
      "step": 955
    },
    {
      "epoch": 0.14298480786416443,
      "grad_norm": 0.5234375,
      "learning_rate": 0.002142537980339589,
      "loss": 0.7909,
      "num_input_tokens_seen": 560096,
      "step": 960
    },
    {
      "epoch": 0.14372952040512363,
      "grad_norm": 0.5546875,
      "learning_rate": 0.002153708668453977,
      "loss": 0.8371,
      "num_input_tokens_seen": 563104,
      "step": 965
    },
    {
      "epoch": 0.1444742329460828,
      "grad_norm": 1.0,
      "learning_rate": 0.0021648793565683644,
      "loss": 0.7764,
      "num_input_tokens_seen": 566048,
      "step": 970
    },
    {
      "epoch": 0.145218945487042,
      "grad_norm": 0.3203125,
      "learning_rate": 0.0021760500446827524,
      "loss": 0.9572,
      "num_input_tokens_seen": 568544,
      "step": 975
    },
    {
      "epoch": 0.14596365802800118,
      "grad_norm": 0.341796875,
      "learning_rate": 0.00218722073279714,
      "loss": 0.7995,
      "num_input_tokens_seen": 571616,
      "step": 980
    },
    {
      "epoch": 0.14670837056896038,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0021983914209115283,
      "loss": 0.811,
      "num_input_tokens_seen": 574496,
      "step": 985
    },
    {
      "epoch": 0.14745308310991956,
      "grad_norm": 0.34765625,
      "learning_rate": 0.002209562109025916,
      "loss": 0.8106,
      "num_input_tokens_seen": 577280,
      "step": 990
    },
    {
      "epoch": 0.14819779565087876,
      "grad_norm": 0.2060546875,
      "learning_rate": 0.002220732797140304,
      "loss": 0.8042,
      "num_input_tokens_seen": 580096,
      "step": 995
    },
    {
      "epoch": 0.14894250819183796,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0022319034852546914,
      "loss": 0.7844,
      "num_input_tokens_seen": 583040,
      "step": 1000
    },
    {
      "epoch": 0.14968722073279714,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0022430741733690794,
      "loss": 0.7881,
      "num_input_tokens_seen": 586016,
      "step": 1005
    },
    {
      "epoch": 0.15043193327375634,
      "grad_norm": 0.171875,
      "learning_rate": 0.0022542448614834673,
      "loss": 0.841,
      "num_input_tokens_seen": 589056,
      "step": 1010
    },
    {
      "epoch": 0.15117664581471552,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0022654155495978553,
      "loss": 0.79,
      "num_input_tokens_seen": 591936,
      "step": 1015
    },
    {
      "epoch": 0.15192135835567472,
      "grad_norm": 0.365234375,
      "learning_rate": 0.0022765862377122433,
      "loss": 0.7988,
      "num_input_tokens_seen": 594848,
      "step": 1020
    },
    {
      "epoch": 0.1526660708966339,
      "grad_norm": 0.220703125,
      "learning_rate": 0.002287756925826631,
      "loss": 0.8258,
      "num_input_tokens_seen": 597792,
      "step": 1025
    },
    {
      "epoch": 0.1534107834375931,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0022989276139410188,
      "loss": 0.8321,
      "num_input_tokens_seen": 600768,
      "step": 1030
    },
    {
      "epoch": 0.15415549597855227,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0023100983020554063,
      "loss": 0.793,
      "num_input_tokens_seen": 603616,
      "step": 1035
    },
    {
      "epoch": 0.15490020851951147,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.0023212689901697943,
      "loss": 0.7979,
      "num_input_tokens_seen": 606432,
      "step": 1040
    },
    {
      "epoch": 0.15564492106047065,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0023324396782841822,
      "loss": 0.8154,
      "num_input_tokens_seen": 609536,
      "step": 1045
    },
    {
      "epoch": 0.15638963360142985,
      "grad_norm": 0.171875,
      "learning_rate": 0.00234361036639857,
      "loss": 0.803,
      "num_input_tokens_seen": 612256,
      "step": 1050
    },
    {
      "epoch": 0.15713434614238903,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0023547810545129578,
      "loss": 0.7985,
      "num_input_tokens_seen": 615136,
      "step": 1055
    },
    {
      "epoch": 0.15787905868334823,
      "grad_norm": 0.244140625,
      "learning_rate": 0.0023659517426273457,
      "loss": 0.8253,
      "num_input_tokens_seen": 618144,
      "step": 1060
    },
    {
      "epoch": 0.15862377122430743,
      "grad_norm": 0.2001953125,
      "learning_rate": 0.0023771224307417333,
      "loss": 0.8004,
      "num_input_tokens_seen": 621440,
      "step": 1065
    },
    {
      "epoch": 0.1593684837652666,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.0023882931188561217,
      "loss": 0.821,
      "num_input_tokens_seen": 624352,
      "step": 1070
    },
    {
      "epoch": 0.1601131963062258,
      "grad_norm": 0.19921875,
      "learning_rate": 0.002399463806970509,
      "loss": 0.789,
      "num_input_tokens_seen": 627392,
      "step": 1075
    },
    {
      "epoch": 0.16085790884718498,
      "grad_norm": 0.24609375,
      "learning_rate": 0.002410634495084897,
      "loss": 0.7933,
      "num_input_tokens_seen": 630368,
      "step": 1080
    },
    {
      "epoch": 0.16160262138814419,
      "grad_norm": 0.13671875,
      "learning_rate": 0.002421805183199285,
      "loss": 0.8103,
      "num_input_tokens_seen": 633504,
      "step": 1085
    },
    {
      "epoch": 0.16234733392910336,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0024329758713136727,
      "loss": 0.828,
      "num_input_tokens_seen": 636640,
      "step": 1090
    },
    {
      "epoch": 0.16309204647006256,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0024441465594280606,
      "loss": 0.781,
      "num_input_tokens_seen": 639744,
      "step": 1095
    },
    {
      "epoch": 0.16383675901102174,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0024553172475424486,
      "loss": 0.8001,
      "num_input_tokens_seen": 642464,
      "step": 1100
    },
    {
      "epoch": 0.16458147155198094,
      "grad_norm": 0.2021484375,
      "learning_rate": 0.0024664879356568366,
      "loss": 0.798,
      "num_input_tokens_seen": 645472,
      "step": 1105
    },
    {
      "epoch": 0.16532618409294011,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.002477658623771224,
      "loss": 0.7871,
      "num_input_tokens_seen": 648320,
      "step": 1110
    },
    {
      "epoch": 0.16607089663389932,
      "grad_norm": 0.5703125,
      "learning_rate": 0.002488829311885612,
      "loss": 0.7895,
      "num_input_tokens_seen": 651168,
      "step": 1115
    },
    {
      "epoch": 0.1668156091748585,
      "grad_norm": 0.25,
      "learning_rate": 0.0024999999999999996,
      "loss": 0.7921,
      "num_input_tokens_seen": 654144,
      "step": 1120
    },
    {
      "epoch": 0.1675603217158177,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.002511170688114388,
      "loss": 0.8011,
      "num_input_tokens_seen": 656928,
      "step": 1125
    },
    {
      "epoch": 0.1683050342567769,
      "grad_norm": 0.34765625,
      "learning_rate": 0.0025223413762287756,
      "loss": 0.8233,
      "num_input_tokens_seen": 659392,
      "step": 1130
    },
    {
      "epoch": 0.16904974679773607,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0025335120643431635,
      "loss": 0.8304,
      "num_input_tokens_seen": 662464,
      "step": 1135
    },
    {
      "epoch": 0.16979445933869527,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.002544682752457551,
      "loss": 0.8154,
      "num_input_tokens_seen": 665408,
      "step": 1140
    },
    {
      "epoch": 0.17053917187965445,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.002555853440571939,
      "loss": 0.8151,
      "num_input_tokens_seen": 668320,
      "step": 1145
    },
    {
      "epoch": 0.17128388442061365,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.002567024128686327,
      "loss": 0.8244,
      "num_input_tokens_seen": 671168,
      "step": 1150
    },
    {
      "epoch": 0.17202859696157283,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.002578194816800715,
      "loss": 0.8037,
      "num_input_tokens_seen": 674368,
      "step": 1155
    },
    {
      "epoch": 0.17277330950253203,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.002589365504915103,
      "loss": 0.8035,
      "num_input_tokens_seen": 677056,
      "step": 1160
    },
    {
      "epoch": 0.1735180220434912,
      "grad_norm": 0.208984375,
      "learning_rate": 0.0026005361930294905,
      "loss": 0.8164,
      "num_input_tokens_seen": 680000,
      "step": 1165
    },
    {
      "epoch": 0.1742627345844504,
      "grad_norm": 0.185546875,
      "learning_rate": 0.0026117068811438784,
      "loss": 0.7988,
      "num_input_tokens_seen": 682752,
      "step": 1170
    },
    {
      "epoch": 0.17500744712540958,
      "grad_norm": 0.171875,
      "learning_rate": 0.002622877569258266,
      "loss": 0.8057,
      "num_input_tokens_seen": 685600,
      "step": 1175
    },
    {
      "epoch": 0.17575215966636878,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.002634048257372654,
      "loss": 0.8157,
      "num_input_tokens_seen": 688448,
      "step": 1180
    },
    {
      "epoch": 0.17649687220732796,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.002645218945487042,
      "loss": 0.8022,
      "num_input_tokens_seen": 691328,
      "step": 1185
    },
    {
      "epoch": 0.17724158474828716,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.00265638963360143,
      "loss": 0.7937,
      "num_input_tokens_seen": 694208,
      "step": 1190
    },
    {
      "epoch": 0.17798629728924636,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0026675603217158174,
      "loss": 0.8153,
      "num_input_tokens_seen": 696928,
      "step": 1195
    },
    {
      "epoch": 0.17873100983020554,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0026787310098302054,
      "loss": 0.7932,
      "num_input_tokens_seen": 699776,
      "step": 1200
    },
    {
      "epoch": 0.17947572237116474,
      "grad_norm": 0.2119140625,
      "learning_rate": 0.002689901697944593,
      "loss": 0.7866,
      "num_input_tokens_seen": 702688,
      "step": 1205
    },
    {
      "epoch": 0.18022043491212392,
      "grad_norm": 0.234375,
      "learning_rate": 0.0027010723860589813,
      "loss": 0.759,
      "num_input_tokens_seen": 705536,
      "step": 1210
    },
    {
      "epoch": 0.18096514745308312,
      "grad_norm": 0.37109375,
      "learning_rate": 0.0027122430741733693,
      "loss": 0.8443,
      "num_input_tokens_seen": 708544,
      "step": 1215
    },
    {
      "epoch": 0.1817098599940423,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.002723413762287757,
      "loss": 0.7938,
      "num_input_tokens_seen": 711200,
      "step": 1220
    },
    {
      "epoch": 0.1824545725350015,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.002734584450402145,
      "loss": 0.8195,
      "num_input_tokens_seen": 714176,
      "step": 1225
    },
    {
      "epoch": 0.18319928507596067,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0027457551385165323,
      "loss": 0.802,
      "num_input_tokens_seen": 716896,
      "step": 1230
    },
    {
      "epoch": 0.18394399761691987,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0027569258266309203,
      "loss": 0.7978,
      "num_input_tokens_seen": 719680,
      "step": 1235
    },
    {
      "epoch": 0.18468871015787905,
      "grad_norm": 0.2236328125,
      "learning_rate": 0.0027680965147453083,
      "loss": 0.8123,
      "num_input_tokens_seen": 722336,
      "step": 1240
    },
    {
      "epoch": 0.18543342269883825,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0027792672028596963,
      "loss": 0.8264,
      "num_input_tokens_seen": 725472,
      "step": 1245
    },
    {
      "epoch": 0.18617813523979743,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.002790437890974084,
      "loss": 0.8186,
      "num_input_tokens_seen": 728512,
      "step": 1250
    },
    {
      "epoch": 0.18692284778075663,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0028016085790884718,
      "loss": 0.8023,
      "num_input_tokens_seen": 731424,
      "step": 1255
    },
    {
      "epoch": 0.1876675603217158,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0028127792672028593,
      "loss": 0.8065,
      "num_input_tokens_seen": 734368,
      "step": 1260
    },
    {
      "epoch": 0.188412272862675,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0028239499553172473,
      "loss": 0.7825,
      "num_input_tokens_seen": 737408,
      "step": 1265
    },
    {
      "epoch": 0.1891569854036342,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0028351206434316352,
      "loss": 0.7678,
      "num_input_tokens_seen": 740576,
      "step": 1270
    },
    {
      "epoch": 0.18990169794459338,
      "grad_norm": 0.173828125,
      "learning_rate": 0.002846291331546023,
      "loss": 0.8316,
      "num_input_tokens_seen": 743200,
      "step": 1275
    },
    {
      "epoch": 0.1906464104855526,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.002857462019660411,
      "loss": 0.8145,
      "num_input_tokens_seen": 746176,
      "step": 1280
    },
    {
      "epoch": 0.19139112302651176,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0028686327077747987,
      "loss": 0.8149,
      "num_input_tokens_seen": 749152,
      "step": 1285
    },
    {
      "epoch": 0.19213583556747096,
      "grad_norm": 0.275390625,
      "learning_rate": 0.0028798033958891867,
      "loss": 0.8109,
      "num_input_tokens_seen": 752128,
      "step": 1290
    },
    {
      "epoch": 0.19288054810843014,
      "grad_norm": 0.416015625,
      "learning_rate": 0.0028909740840035746,
      "loss": 0.7829,
      "num_input_tokens_seen": 755072,
      "step": 1295
    },
    {
      "epoch": 0.19362526064938934,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0029021447721179626,
      "loss": 0.7677,
      "num_input_tokens_seen": 758016,
      "step": 1300
    },
    {
      "epoch": 0.19436997319034852,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00291331546023235,
      "loss": 0.7986,
      "num_input_tokens_seen": 760928,
      "step": 1305
    },
    {
      "epoch": 0.19511468573130772,
      "grad_norm": 0.134765625,
      "learning_rate": 0.002924486148346738,
      "loss": 0.8195,
      "num_input_tokens_seen": 763776,
      "step": 1310
    },
    {
      "epoch": 0.1958593982722669,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0029356568364611257,
      "loss": 0.8154,
      "num_input_tokens_seen": 766432,
      "step": 1315
    },
    {
      "epoch": 0.1966041108132261,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0029468275245755136,
      "loss": 0.7835,
      "num_input_tokens_seen": 769280,
      "step": 1320
    },
    {
      "epoch": 0.19734882335418527,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0029579982126899016,
      "loss": 0.7873,
      "num_input_tokens_seen": 772224,
      "step": 1325
    },
    {
      "epoch": 0.19809353589514447,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0029691689008042896,
      "loss": 0.8426,
      "num_input_tokens_seen": 774944,
      "step": 1330
    },
    {
      "epoch": 0.19883824843610368,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.002980339588918677,
      "loss": 0.8255,
      "num_input_tokens_seen": 777664,
      "step": 1335
    },
    {
      "epoch": 0.19958296097706285,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.002991510277033065,
      "loss": 0.8002,
      "num_input_tokens_seen": 780480,
      "step": 1340
    },
    {
      "epoch": 0.20032767351802205,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.003002680965147453,
      "loss": 0.8231,
      "num_input_tokens_seen": 783296,
      "step": 1345
    },
    {
      "epoch": 0.20107238605898123,
      "grad_norm": 0.2119140625,
      "learning_rate": 0.003013851653261841,
      "loss": 0.8056,
      "num_input_tokens_seen": 786016,
      "step": 1350
    },
    {
      "epoch": 0.20181709859994043,
      "grad_norm": 0.40234375,
      "learning_rate": 0.003025022341376229,
      "loss": 0.8247,
      "num_input_tokens_seen": 788960,
      "step": 1355
    },
    {
      "epoch": 0.2025618111408996,
      "grad_norm": 0.3515625,
      "learning_rate": 0.0030361930294906165,
      "loss": 0.8236,
      "num_input_tokens_seen": 791936,
      "step": 1360
    },
    {
      "epoch": 0.2033065236818588,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0030473637176050045,
      "loss": 0.8269,
      "num_input_tokens_seen": 794720,
      "step": 1365
    },
    {
      "epoch": 0.20405123622281798,
      "grad_norm": 0.158203125,
      "learning_rate": 0.003058534405719392,
      "loss": 0.7759,
      "num_input_tokens_seen": 797376,
      "step": 1370
    },
    {
      "epoch": 0.20479594876377719,
      "grad_norm": 0.15625,
      "learning_rate": 0.00306970509383378,
      "loss": 0.8103,
      "num_input_tokens_seen": 800192,
      "step": 1375
    },
    {
      "epoch": 0.20554066130473636,
      "grad_norm": 0.412109375,
      "learning_rate": 0.003080875781948168,
      "loss": 0.7956,
      "num_input_tokens_seen": 803104,
      "step": 1380
    },
    {
      "epoch": 0.20628537384569556,
      "grad_norm": 0.21484375,
      "learning_rate": 0.003092046470062556,
      "loss": 0.8015,
      "num_input_tokens_seen": 806496,
      "step": 1385
    },
    {
      "epoch": 0.20703008638665474,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.0031032171581769435,
      "loss": 0.8026,
      "num_input_tokens_seen": 809344,
      "step": 1390
    },
    {
      "epoch": 0.20777479892761394,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.0031143878462913314,
      "loss": 0.8065,
      "num_input_tokens_seen": 812256,
      "step": 1395
    },
    {
      "epoch": 0.20851951146857314,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.003125558534405719,
      "loss": 0.845,
      "num_input_tokens_seen": 814816,
      "step": 1400
    },
    {
      "epoch": 0.20926422400953232,
      "grad_norm": 0.19921875,
      "learning_rate": 0.003136729222520107,
      "loss": 0.8042,
      "num_input_tokens_seen": 817440,
      "step": 1405
    },
    {
      "epoch": 0.21000893655049152,
      "grad_norm": 0.388671875,
      "learning_rate": 0.0031478999106344953,
      "loss": 0.7613,
      "num_input_tokens_seen": 820640,
      "step": 1410
    },
    {
      "epoch": 0.2107536490914507,
      "grad_norm": 0.33203125,
      "learning_rate": 0.003159070598748883,
      "loss": 0.8333,
      "num_input_tokens_seen": 823488,
      "step": 1415
    },
    {
      "epoch": 0.2114983616324099,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.003170241286863271,
      "loss": 0.8019,
      "num_input_tokens_seen": 826240,
      "step": 1420
    },
    {
      "epoch": 0.21224307417336907,
      "grad_norm": 0.310546875,
      "learning_rate": 0.0031814119749776584,
      "loss": 0.802,
      "num_input_tokens_seen": 828992,
      "step": 1425
    },
    {
      "epoch": 0.21298778671432828,
      "grad_norm": 0.5546875,
      "learning_rate": 0.0031925826630920463,
      "loss": 0.8035,
      "num_input_tokens_seen": 831648,
      "step": 1430
    },
    {
      "epoch": 0.21373249925528745,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0032037533512064343,
      "loss": 0.786,
      "num_input_tokens_seen": 834336,
      "step": 1435
    },
    {
      "epoch": 0.21447721179624665,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.0032149240393208223,
      "loss": 0.8404,
      "num_input_tokens_seen": 837536,
      "step": 1440
    },
    {
      "epoch": 0.21522192433720583,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00322609472743521,
      "loss": 0.8017,
      "num_input_tokens_seen": 841088,
      "step": 1445
    },
    {
      "epoch": 0.21596663687816503,
      "grad_norm": 0.16796875,
      "learning_rate": 0.003237265415549598,
      "loss": 0.8064,
      "num_input_tokens_seen": 844000,
      "step": 1450
    },
    {
      "epoch": 0.2167113494191242,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.0032484361036639853,
      "loss": 0.7731,
      "num_input_tokens_seen": 846912,
      "step": 1455
    },
    {
      "epoch": 0.2174560619600834,
      "grad_norm": 0.216796875,
      "learning_rate": 0.0032596067917783733,
      "loss": 0.827,
      "num_input_tokens_seen": 849952,
      "step": 1460
    },
    {
      "epoch": 0.2182007745010426,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0032707774798927613,
      "loss": 0.8175,
      "num_input_tokens_seen": 852768,
      "step": 1465
    },
    {
      "epoch": 0.21894548704200179,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.0032819481680071492,
      "loss": 0.8203,
      "num_input_tokens_seen": 855808,
      "step": 1470
    },
    {
      "epoch": 0.219690199582961,
      "grad_norm": 0.25390625,
      "learning_rate": 0.003293118856121537,
      "loss": 0.806,
      "num_input_tokens_seen": 858624,
      "step": 1475
    },
    {
      "epoch": 0.22043491212392016,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0033042895442359247,
      "loss": 0.8093,
      "num_input_tokens_seen": 861600,
      "step": 1480
    },
    {
      "epoch": 0.22117962466487937,
      "grad_norm": 0.2041015625,
      "learning_rate": 0.0033154602323503127,
      "loss": 0.7755,
      "num_input_tokens_seen": 864768,
      "step": 1485
    },
    {
      "epoch": 0.22192433720583854,
      "grad_norm": 0.189453125,
      "learning_rate": 0.0033266309204647002,
      "loss": 0.8289,
      "num_input_tokens_seen": 867776,
      "step": 1490
    },
    {
      "epoch": 0.22266904974679774,
      "grad_norm": 0.2119140625,
      "learning_rate": 0.0033378016085790886,
      "loss": 0.7943,
      "num_input_tokens_seen": 871104,
      "step": 1495
    },
    {
      "epoch": 0.22341376228775692,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.003348972296693476,
      "loss": 0.8039,
      "num_input_tokens_seen": 874304,
      "step": 1500
    },
    {
      "epoch": 0.22415847482871612,
      "grad_norm": 0.12890625,
      "learning_rate": 0.003360142984807864,
      "loss": 0.8083,
      "num_input_tokens_seen": 877184,
      "step": 1505
    },
    {
      "epoch": 0.2249031873696753,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0033713136729222517,
      "loss": 0.8262,
      "num_input_tokens_seen": 880128,
      "step": 1510
    },
    {
      "epoch": 0.2256478999106345,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0033824843610366397,
      "loss": 0.7944,
      "num_input_tokens_seen": 883328,
      "step": 1515
    },
    {
      "epoch": 0.22639261245159367,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.0033936550491510276,
      "loss": 0.8216,
      "num_input_tokens_seen": 886304,
      "step": 1520
    },
    {
      "epoch": 0.22713732499255287,
      "grad_norm": 0.189453125,
      "learning_rate": 0.0034048257372654156,
      "loss": 0.8168,
      "num_input_tokens_seen": 889280,
      "step": 1525
    },
    {
      "epoch": 0.22788203753351208,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.003415996425379803,
      "loss": 0.8008,
      "num_input_tokens_seen": 892000,
      "step": 1530
    },
    {
      "epoch": 0.22862675007447125,
      "grad_norm": 0.23046875,
      "learning_rate": 0.003427167113494191,
      "loss": 0.8287,
      "num_input_tokens_seen": 894912,
      "step": 1535
    },
    {
      "epoch": 0.22937146261543045,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0034383378016085786,
      "loss": 0.7983,
      "num_input_tokens_seen": 897600,
      "step": 1540
    },
    {
      "epoch": 0.23011617515638963,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0034495084897229666,
      "loss": 0.8113,
      "num_input_tokens_seen": 900800,
      "step": 1545
    },
    {
      "epoch": 0.23086088769734883,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.003460679177837355,
      "loss": 0.792,
      "num_input_tokens_seen": 903712,
      "step": 1550
    },
    {
      "epoch": 0.231605600238308,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0034718498659517425,
      "loss": 0.7976,
      "num_input_tokens_seen": 906560,
      "step": 1555
    },
    {
      "epoch": 0.2323503127792672,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0034830205540661305,
      "loss": 0.7974,
      "num_input_tokens_seen": 909344,
      "step": 1560
    },
    {
      "epoch": 0.23309502532022638,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.003494191242180518,
      "loss": 0.7987,
      "num_input_tokens_seen": 911968,
      "step": 1565
    },
    {
      "epoch": 0.2338397378611856,
      "grad_norm": 0.2255859375,
      "learning_rate": 0.003505361930294906,
      "loss": 0.8109,
      "num_input_tokens_seen": 914816,
      "step": 1570
    },
    {
      "epoch": 0.23458445040214476,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0035165326184092936,
      "loss": 0.7841,
      "num_input_tokens_seen": 917696,
      "step": 1575
    },
    {
      "epoch": 0.23532916294310396,
      "grad_norm": 0.2197265625,
      "learning_rate": 0.003527703306523682,
      "loss": 0.8,
      "num_input_tokens_seen": 920992,
      "step": 1580
    },
    {
      "epoch": 0.23607387548406314,
      "grad_norm": 0.2109375,
      "learning_rate": 0.0035388739946380695,
      "loss": 0.7975,
      "num_input_tokens_seen": 923968,
      "step": 1585
    },
    {
      "epoch": 0.23681858802502234,
      "grad_norm": 0.171875,
      "learning_rate": 0.0035500446827524575,
      "loss": 0.8281,
      "num_input_tokens_seen": 926656,
      "step": 1590
    },
    {
      "epoch": 0.23756330056598154,
      "grad_norm": 0.189453125,
      "learning_rate": 0.003561215370866845,
      "loss": 0.8041,
      "num_input_tokens_seen": 929696,
      "step": 1595
    },
    {
      "epoch": 0.23830801310694072,
      "grad_norm": 0.2041015625,
      "learning_rate": 0.003572386058981233,
      "loss": 0.8228,
      "num_input_tokens_seen": 932608,
      "step": 1600
    },
    {
      "epoch": 0.23905272564789992,
      "grad_norm": 0.130859375,
      "learning_rate": 0.003583556747095621,
      "loss": 0.8174,
      "num_input_tokens_seen": 935424,
      "step": 1605
    },
    {
      "epoch": 0.2397974381888591,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.003594727435210009,
      "loss": 0.7908,
      "num_input_tokens_seen": 938240,
      "step": 1610
    },
    {
      "epoch": 0.2405421507298183,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.003605898123324397,
      "loss": 0.806,
      "num_input_tokens_seen": 941152,
      "step": 1615
    },
    {
      "epoch": 0.24128686327077747,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0036170688114387844,
      "loss": 0.79,
      "num_input_tokens_seen": 943936,
      "step": 1620
    },
    {
      "epoch": 0.24203157581173668,
      "grad_norm": 0.185546875,
      "learning_rate": 0.0036282394995531724,
      "loss": 0.7929,
      "num_input_tokens_seen": 946944,
      "step": 1625
    },
    {
      "epoch": 0.24277628835269585,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.00363941018766756,
      "loss": 0.8097,
      "num_input_tokens_seen": 949696,
      "step": 1630
    },
    {
      "epoch": 0.24352100089365505,
      "grad_norm": 0.11328125,
      "learning_rate": 0.0036505808757819483,
      "loss": 0.7953,
      "num_input_tokens_seen": 952608,
      "step": 1635
    },
    {
      "epoch": 0.24426571343461423,
      "grad_norm": 0.142578125,
      "learning_rate": 0.003661751563896336,
      "loss": 0.798,
      "num_input_tokens_seen": 955712,
      "step": 1640
    },
    {
      "epoch": 0.24501042597557343,
      "grad_norm": 0.16015625,
      "learning_rate": 0.003672922252010724,
      "loss": 0.8099,
      "num_input_tokens_seen": 959392,
      "step": 1645
    },
    {
      "epoch": 0.2457551385165326,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0036840929401251114,
      "loss": 0.8173,
      "num_input_tokens_seen": 962464,
      "step": 1650
    },
    {
      "epoch": 0.2464998510574918,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0036952636282394993,
      "loss": 0.8057,
      "num_input_tokens_seen": 965408,
      "step": 1655
    },
    {
      "epoch": 0.247244563598451,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0037064343163538873,
      "loss": 0.7866,
      "num_input_tokens_seen": 968288,
      "step": 1660
    },
    {
      "epoch": 0.2479892761394102,
      "grad_norm": 0.1875,
      "learning_rate": 0.0037176050044682753,
      "loss": 0.8133,
      "num_input_tokens_seen": 971168,
      "step": 1665
    },
    {
      "epoch": 0.2487339886803694,
      "grad_norm": 0.16015625,
      "learning_rate": 0.003728775692582663,
      "loss": 0.8129,
      "num_input_tokens_seen": 974016,
      "step": 1670
    },
    {
      "epoch": 0.24947870122132856,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0037399463806970508,
      "loss": 0.8012,
      "num_input_tokens_seen": 977056,
      "step": 1675
    },
    {
      "epoch": 0.25022341376228774,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0037511170688114387,
      "loss": 0.8194,
      "num_input_tokens_seen": 980384,
      "step": 1680
    },
    {
      "epoch": 0.25096812630324694,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0037622877569258267,
      "loss": 0.8221,
      "num_input_tokens_seen": 983264,
      "step": 1685
    },
    {
      "epoch": 0.25171283884420614,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0037734584450402142,
      "loss": 0.803,
      "num_input_tokens_seen": 986496,
      "step": 1690
    },
    {
      "epoch": 0.25245755138516535,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0037846291331546022,
      "loss": 0.7886,
      "num_input_tokens_seen": 989248,
      "step": 1695
    },
    {
      "epoch": 0.2532022639261245,
      "grad_norm": 0.185546875,
      "learning_rate": 0.00379579982126899,
      "loss": 0.8152,
      "num_input_tokens_seen": 992000,
      "step": 1700
    },
    {
      "epoch": 0.2539469764670837,
      "grad_norm": 0.18359375,
      "learning_rate": 0.003806970509383378,
      "loss": 0.8078,
      "num_input_tokens_seen": 994784,
      "step": 1705
    },
    {
      "epoch": 0.2546916890080429,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0038181411974977653,
      "loss": 0.7974,
      "num_input_tokens_seen": 997696,
      "step": 1710
    },
    {
      "epoch": 0.2554364015490021,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0038293118856121532,
      "loss": 0.8187,
      "num_input_tokens_seen": 1000416,
      "step": 1715
    },
    {
      "epoch": 0.25618111408996125,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0038404825737265416,
      "loss": 0.7855,
      "num_input_tokens_seen": 1002976,
      "step": 1720
    },
    {
      "epoch": 0.25692582663092045,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0038516532618409296,
      "loss": 0.819,
      "num_input_tokens_seen": 1006048,
      "step": 1725
    },
    {
      "epoch": 0.25767053917187965,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0038628239499553167,
      "loss": 0.7978,
      "num_input_tokens_seen": 1009184,
      "step": 1730
    },
    {
      "epoch": 0.25841525171283886,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0038739946380697047,
      "loss": 0.7983,
      "num_input_tokens_seen": 1011904,
      "step": 1735
    },
    {
      "epoch": 0.25915996425379806,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.0038851653261840926,
      "loss": 0.7865,
      "num_input_tokens_seen": 1014432,
      "step": 1740
    },
    {
      "epoch": 0.2599046767947572,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.003896336014298481,
      "loss": 0.8264,
      "num_input_tokens_seen": 1017376,
      "step": 1745
    },
    {
      "epoch": 0.2606493893357164,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.003907506702412869,
      "loss": 0.818,
      "num_input_tokens_seen": 1020192,
      "step": 1750
    },
    {
      "epoch": 0.2613941018766756,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.003918677390527256,
      "loss": 0.811,
      "num_input_tokens_seen": 1023360,
      "step": 1755
    },
    {
      "epoch": 0.2621388144176348,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.003929848078641644,
      "loss": 0.8079,
      "num_input_tokens_seen": 1026208,
      "step": 1760
    },
    {
      "epoch": 0.26288352695859396,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.003941018766756032,
      "loss": 0.7813,
      "num_input_tokens_seen": 1029024,
      "step": 1765
    },
    {
      "epoch": 0.26362823949955316,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.00395218945487042,
      "loss": 0.8197,
      "num_input_tokens_seen": 1032064,
      "step": 1770
    },
    {
      "epoch": 0.26437295204051237,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.003963360142984807,
      "loss": 0.8026,
      "num_input_tokens_seen": 1034880,
      "step": 1775
    },
    {
      "epoch": 0.26511766458147157,
      "grad_norm": 0.07421875,
      "learning_rate": 0.003974530831099195,
      "loss": 0.7915,
      "num_input_tokens_seen": 1037696,
      "step": 1780
    },
    {
      "epoch": 0.2658623771224307,
      "grad_norm": 0.1328125,
      "learning_rate": 0.003985701519213583,
      "loss": 0.8252,
      "num_input_tokens_seen": 1040448,
      "step": 1785
    },
    {
      "epoch": 0.2666070896633899,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.003996872207327972,
      "loss": 0.8003,
      "num_input_tokens_seen": 1043104,
      "step": 1790
    },
    {
      "epoch": 0.2673518022043491,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.004008042895442359,
      "loss": 0.7904,
      "num_input_tokens_seen": 1045824,
      "step": 1795
    },
    {
      "epoch": 0.2680965147453083,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.004019213583556747,
      "loss": 0.7982,
      "num_input_tokens_seen": 1048672,
      "step": 1800
    },
    {
      "epoch": 0.2688412272862675,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.004030384271671135,
      "loss": 0.802,
      "num_input_tokens_seen": 1051360,
      "step": 1805
    },
    {
      "epoch": 0.2695859398272267,
      "grad_norm": 0.185546875,
      "learning_rate": 0.004041554959785523,
      "loss": 0.8205,
      "num_input_tokens_seen": 1054304,
      "step": 1810
    },
    {
      "epoch": 0.2703306523681859,
      "grad_norm": 0.267578125,
      "learning_rate": 0.004052725647899911,
      "loss": 0.8175,
      "num_input_tokens_seen": 1057952,
      "step": 1815
    },
    {
      "epoch": 0.2710753649091451,
      "grad_norm": 0.150390625,
      "learning_rate": 0.004063896336014298,
      "loss": 0.8011,
      "num_input_tokens_seen": 1060704,
      "step": 1820
    },
    {
      "epoch": 0.2718200774501043,
      "grad_norm": 0.2060546875,
      "learning_rate": 0.004075067024128686,
      "loss": 0.7827,
      "num_input_tokens_seen": 1063360,
      "step": 1825
    },
    {
      "epoch": 0.2725647899910634,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.004086237712243074,
      "loss": 0.792,
      "num_input_tokens_seen": 1066112,
      "step": 1830
    },
    {
      "epoch": 0.27330950253202263,
      "grad_norm": 45.75,
      "learning_rate": 0.004097408400357462,
      "loss": 1.167,
      "num_input_tokens_seen": 1068992,
      "step": 1835
    },
    {
      "epoch": 0.27405421507298183,
      "grad_norm": 0.216796875,
      "learning_rate": 0.004108579088471849,
      "loss": 0.8448,
      "num_input_tokens_seen": 1071872,
      "step": 1840
    },
    {
      "epoch": 0.27479892761394104,
      "grad_norm": 0.921875,
      "learning_rate": 0.004119749776586238,
      "loss": 0.8134,
      "num_input_tokens_seen": 1074752,
      "step": 1845
    },
    {
      "epoch": 0.2755436401549002,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.004130920464700626,
      "loss": 0.8569,
      "num_input_tokens_seen": 1077728,
      "step": 1850
    },
    {
      "epoch": 0.2762883526958594,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.004142091152815014,
      "loss": 0.7924,
      "num_input_tokens_seen": 1080512,
      "step": 1855
    },
    {
      "epoch": 0.2770330652368186,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.004153261840929401,
      "loss": 0.7996,
      "num_input_tokens_seen": 1083424,
      "step": 1860
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.004164432529043789,
      "loss": 0.7974,
      "num_input_tokens_seen": 1086208,
      "step": 1865
    },
    {
      "epoch": 0.278522490318737,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.004175603217158177,
      "loss": 0.8234,
      "num_input_tokens_seen": 1089152,
      "step": 1870
    },
    {
      "epoch": 0.27926720285969614,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.004186773905272565,
      "loss": 0.7978,
      "num_input_tokens_seen": 1092000,
      "step": 1875
    },
    {
      "epoch": 0.28001191540065534,
      "grad_norm": 0.1171875,
      "learning_rate": 0.004197944593386953,
      "loss": 0.7903,
      "num_input_tokens_seen": 1095008,
      "step": 1880
    },
    {
      "epoch": 0.28075662794161454,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.00420911528150134,
      "loss": 0.8021,
      "num_input_tokens_seen": 1098080,
      "step": 1885
    },
    {
      "epoch": 0.28150134048257375,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.004220285969615728,
      "loss": 0.8237,
      "num_input_tokens_seen": 1100736,
      "step": 1890
    },
    {
      "epoch": 0.2822460530235329,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.004231456657730116,
      "loss": 0.8198,
      "num_input_tokens_seen": 1103744,
      "step": 1895
    },
    {
      "epoch": 0.2829907655644921,
      "grad_norm": 0.0859375,
      "learning_rate": 0.004242627345844504,
      "loss": 0.7883,
      "num_input_tokens_seen": 1106304,
      "step": 1900
    },
    {
      "epoch": 0.2837354781054513,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.004253798033958892,
      "loss": 0.8205,
      "num_input_tokens_seen": 1109472,
      "step": 1905
    },
    {
      "epoch": 0.2844801906464105,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.00426496872207328,
      "loss": 0.8084,
      "num_input_tokens_seen": 1112192,
      "step": 1910
    },
    {
      "epoch": 0.28522490318736965,
      "grad_norm": 0.25,
      "learning_rate": 0.004276139410187668,
      "loss": 0.8026,
      "num_input_tokens_seen": 1115168,
      "step": 1915
    },
    {
      "epoch": 0.28596961572832885,
      "grad_norm": 0.173828125,
      "learning_rate": 0.004287310098302056,
      "loss": 0.8083,
      "num_input_tokens_seen": 1118016,
      "step": 1920
    },
    {
      "epoch": 0.28671432826928805,
      "grad_norm": 0.28125,
      "learning_rate": 0.004298480786416443,
      "loss": 0.8222,
      "num_input_tokens_seen": 1120704,
      "step": 1925
    },
    {
      "epoch": 0.28745904081024726,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.004309651474530831,
      "loss": 0.8029,
      "num_input_tokens_seen": 1123552,
      "step": 1930
    },
    {
      "epoch": 0.28820375335120646,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.004320822162645219,
      "loss": 0.8159,
      "num_input_tokens_seen": 1126368,
      "step": 1935
    },
    {
      "epoch": 0.2889484658921656,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.004331992850759607,
      "loss": 0.7952,
      "num_input_tokens_seen": 1129312,
      "step": 1940
    },
    {
      "epoch": 0.2896931784331248,
      "grad_norm": 0.2099609375,
      "learning_rate": 0.004343163538873995,
      "loss": 0.7949,
      "num_input_tokens_seen": 1132256,
      "step": 1945
    },
    {
      "epoch": 0.290437890974084,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.004354334226988382,
      "loss": 0.8148,
      "num_input_tokens_seen": 1135008,
      "step": 1950
    },
    {
      "epoch": 0.2911826035150432,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.00436550491510277,
      "loss": 0.7817,
      "num_input_tokens_seen": 1137728,
      "step": 1955
    },
    {
      "epoch": 0.29192731605600236,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0043766756032171585,
      "loss": 0.7769,
      "num_input_tokens_seen": 1140512,
      "step": 1960
    },
    {
      "epoch": 0.29267202859696156,
      "grad_norm": 0.2119140625,
      "learning_rate": 0.0043878462913315465,
      "loss": 0.8068,
      "num_input_tokens_seen": 1143616,
      "step": 1965
    },
    {
      "epoch": 0.29341674113792077,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.004399016979445934,
      "loss": 0.7804,
      "num_input_tokens_seen": 1146496,
      "step": 1970
    },
    {
      "epoch": 0.29416145367887997,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0044101876675603216,
      "loss": 0.774,
      "num_input_tokens_seen": 1149600,
      "step": 1975
    },
    {
      "epoch": 0.2949061662198391,
      "grad_norm": 0.376953125,
      "learning_rate": 0.0044213583556747095,
      "loss": 0.8425,
      "num_input_tokens_seen": 1152320,
      "step": 1980
    },
    {
      "epoch": 0.2956508787607983,
      "grad_norm": 0.2275390625,
      "learning_rate": 0.0044325290437890975,
      "loss": 0.7277,
      "num_input_tokens_seen": 1155648,
      "step": 1985
    },
    {
      "epoch": 0.2963955913017575,
      "grad_norm": 0.310546875,
      "learning_rate": 0.004443699731903485,
      "loss": 1.0019,
      "num_input_tokens_seen": 1158336,
      "step": 1990
    },
    {
      "epoch": 0.2971403038427167,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.004454870420017873,
      "loss": 0.8634,
      "num_input_tokens_seen": 1161280,
      "step": 1995
    },
    {
      "epoch": 0.2978850163836759,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0044660411081322605,
      "loss": 0.8086,
      "num_input_tokens_seen": 1164064,
      "step": 2000
    },
    {
      "epoch": 0.2986297289246351,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0044772117962466485,
      "loss": 0.7784,
      "num_input_tokens_seen": 1166880,
      "step": 2005
    },
    {
      "epoch": 0.2993744414655943,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0044883824843610365,
      "loss": 0.8299,
      "num_input_tokens_seen": 1169792,
      "step": 2010
    },
    {
      "epoch": 0.3001191540065535,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.0044995531724754245,
      "loss": 0.7999,
      "num_input_tokens_seen": 1172608,
      "step": 2015
    },
    {
      "epoch": 0.3008638665475127,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.004510723860589812,
      "loss": 0.8082,
      "num_input_tokens_seen": 1175168,
      "step": 2020
    },
    {
      "epoch": 0.30160857908847183,
      "grad_norm": 0.0625,
      "learning_rate": 0.0045218945487042,
      "loss": 0.8174,
      "num_input_tokens_seen": 1178016,
      "step": 2025
    },
    {
      "epoch": 0.30235329162943103,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.004533065236818588,
      "loss": 0.7824,
      "num_input_tokens_seen": 1180800,
      "step": 2030
    },
    {
      "epoch": 0.30309800417039023,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0045442359249329755,
      "loss": 0.7757,
      "num_input_tokens_seen": 1183776,
      "step": 2035
    },
    {
      "epoch": 0.30384271671134944,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.004555406613047363,
      "loss": 0.8222,
      "num_input_tokens_seen": 1186880,
      "step": 2040
    },
    {
      "epoch": 0.3045874292523086,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.004566577301161751,
      "loss": 0.7931,
      "num_input_tokens_seen": 1189856,
      "step": 2045
    },
    {
      "epoch": 0.3053321417932678,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.004577747989276139,
      "loss": 0.8225,
      "num_input_tokens_seen": 1192736,
      "step": 2050
    },
    {
      "epoch": 0.306076854334227,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0045889186773905265,
      "loss": 0.7805,
      "num_input_tokens_seen": 1195808,
      "step": 2055
    },
    {
      "epoch": 0.3068215668751862,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0046000893655049144,
      "loss": 0.7865,
      "num_input_tokens_seen": 1198656,
      "step": 2060
    },
    {
      "epoch": 0.3075662794161454,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.004611260053619302,
      "loss": 0.8075,
      "num_input_tokens_seen": 1201664,
      "step": 2065
    },
    {
      "epoch": 0.30831099195710454,
      "grad_norm": 0.119140625,
      "learning_rate": 0.004622430741733691,
      "loss": 0.8101,
      "num_input_tokens_seen": 1204384,
      "step": 2070
    },
    {
      "epoch": 0.30905570449806374,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.004633601429848079,
      "loss": 0.791,
      "num_input_tokens_seen": 1207360,
      "step": 2075
    },
    {
      "epoch": 0.30980041703902295,
      "grad_norm": 0.076171875,
      "learning_rate": 0.004644772117962466,
      "loss": 0.838,
      "num_input_tokens_seen": 1210144,
      "step": 2080
    },
    {
      "epoch": 0.31054512957998215,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.004655942806076854,
      "loss": 0.7827,
      "num_input_tokens_seen": 1213088,
      "step": 2085
    },
    {
      "epoch": 0.3112898421209413,
      "grad_norm": 0.1015625,
      "learning_rate": 0.004667113494191242,
      "loss": 0.8126,
      "num_input_tokens_seen": 1216416,
      "step": 2090
    },
    {
      "epoch": 0.3120345546619005,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.00467828418230563,
      "loss": 0.8275,
      "num_input_tokens_seen": 1219424,
      "step": 2095
    },
    {
      "epoch": 0.3127792672028597,
      "grad_norm": 0.21484375,
      "learning_rate": 0.004689454870420017,
      "loss": 0.8199,
      "num_input_tokens_seen": 1222528,
      "step": 2100
    },
    {
      "epoch": 0.3135239797438189,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.004700625558534405,
      "loss": 0.7786,
      "num_input_tokens_seen": 1225408,
      "step": 2105
    },
    {
      "epoch": 0.31426869228477805,
      "grad_norm": 0.076171875,
      "learning_rate": 0.004711796246648793,
      "loss": 0.8157,
      "num_input_tokens_seen": 1228384,
      "step": 2110
    },
    {
      "epoch": 0.31501340482573725,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.004722966934763181,
      "loss": 0.8157,
      "num_input_tokens_seen": 1230880,
      "step": 2115
    },
    {
      "epoch": 0.31575811736669646,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.004734137622877568,
      "loss": 0.799,
      "num_input_tokens_seen": 1233920,
      "step": 2120
    },
    {
      "epoch": 0.31650282990765566,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.004745308310991957,
      "loss": 0.7912,
      "num_input_tokens_seen": 1236704,
      "step": 2125
    },
    {
      "epoch": 0.31724754244861486,
      "grad_norm": 0.2236328125,
      "learning_rate": 0.004756478999106345,
      "loss": 0.8166,
      "num_input_tokens_seen": 1239616,
      "step": 2130
    },
    {
      "epoch": 0.317992254989574,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.004767649687220733,
      "loss": 0.8064,
      "num_input_tokens_seen": 1242464,
      "step": 2135
    },
    {
      "epoch": 0.3187369675305332,
      "grad_norm": 0.2314453125,
      "learning_rate": 0.004778820375335121,
      "loss": 0.7963,
      "num_input_tokens_seen": 1245440,
      "step": 2140
    },
    {
      "epoch": 0.3194816800714924,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.004789991063449508,
      "loss": 0.7968,
      "num_input_tokens_seen": 1248416,
      "step": 2145
    },
    {
      "epoch": 0.3202263926124516,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.004801161751563896,
      "loss": 0.8293,
      "num_input_tokens_seen": 1251008,
      "step": 2150
    },
    {
      "epoch": 0.32097110515341076,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.004812332439678284,
      "loss": 0.7915,
      "num_input_tokens_seen": 1253888,
      "step": 2155
    },
    {
      "epoch": 0.32171581769436997,
      "grad_norm": 0.13671875,
      "learning_rate": 0.004823503127792672,
      "loss": 0.7886,
      "num_input_tokens_seen": 1256512,
      "step": 2160
    },
    {
      "epoch": 0.32246053023532917,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.004834673815907059,
      "loss": 0.808,
      "num_input_tokens_seen": 1259232,
      "step": 2165
    },
    {
      "epoch": 0.32320524277628837,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.004845844504021447,
      "loss": 0.8232,
      "num_input_tokens_seen": 1261952,
      "step": 2170
    },
    {
      "epoch": 0.3239499553172475,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.004857015192135835,
      "loss": 0.8055,
      "num_input_tokens_seen": 1264704,
      "step": 2175
    },
    {
      "epoch": 0.3246946678582067,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.004868185880250223,
      "loss": 0.8108,
      "num_input_tokens_seen": 1267424,
      "step": 2180
    },
    {
      "epoch": 0.3254393803991659,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.004879356568364611,
      "loss": 0.794,
      "num_input_tokens_seen": 1270496,
      "step": 2185
    },
    {
      "epoch": 0.3261840929401251,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.004890527256478999,
      "loss": 0.7929,
      "num_input_tokens_seen": 1273248,
      "step": 2190
    },
    {
      "epoch": 0.32692880548108433,
      "grad_norm": 0.09765625,
      "learning_rate": 0.004901697944593387,
      "loss": 0.7683,
      "num_input_tokens_seen": 1276288,
      "step": 2195
    },
    {
      "epoch": 0.3276735180220435,
      "grad_norm": 0.09375,
      "learning_rate": 0.004912868632707775,
      "loss": 0.7998,
      "num_input_tokens_seen": 1279584,
      "step": 2200
    },
    {
      "epoch": 0.3284182305630027,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.004924039320822163,
      "loss": 0.7985,
      "num_input_tokens_seen": 1282656,
      "step": 2205
    },
    {
      "epoch": 0.3291629431039619,
      "grad_norm": 0.078125,
      "learning_rate": 0.00493521000893655,
      "loss": 0.7859,
      "num_input_tokens_seen": 1285568,
      "step": 2210
    },
    {
      "epoch": 0.3299076556449211,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.004946380697050938,
      "loss": 0.8076,
      "num_input_tokens_seen": 1288448,
      "step": 2215
    },
    {
      "epoch": 0.33065236818588023,
      "grad_norm": 0.173828125,
      "learning_rate": 0.004957551385165326,
      "loss": 0.7768,
      "num_input_tokens_seen": 1291264,
      "step": 2220
    },
    {
      "epoch": 0.33139708072683943,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.004968722073279714,
      "loss": 0.7992,
      "num_input_tokens_seen": 1294240,
      "step": 2225
    },
    {
      "epoch": 0.33214179326779864,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.004979892761394101,
      "loss": 0.7924,
      "num_input_tokens_seen": 1296992,
      "step": 2230
    },
    {
      "epoch": 0.33288650580875784,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.004991063449508489,
      "loss": 0.8205,
      "num_input_tokens_seen": 1299808,
      "step": 2235
    },
    {
      "epoch": 0.333631218349717,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.005002234137622878,
      "loss": 0.8069,
      "num_input_tokens_seen": 1302816,
      "step": 2240
    },
    {
      "epoch": 0.3343759308906762,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.005013404825737266,
      "loss": 0.7866,
      "num_input_tokens_seen": 1305600,
      "step": 2245
    },
    {
      "epoch": 0.3351206434316354,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.005024575513851653,
      "loss": 0.813,
      "num_input_tokens_seen": 1309088,
      "step": 2250
    },
    {
      "epoch": 0.3358653559725946,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.005035746201966041,
      "loss": 0.8387,
      "num_input_tokens_seen": 1311968,
      "step": 2255
    },
    {
      "epoch": 0.3366100685135538,
      "grad_norm": 0.1171875,
      "learning_rate": 0.005046916890080429,
      "loss": 0.7961,
      "num_input_tokens_seen": 1315040,
      "step": 2260
    },
    {
      "epoch": 0.33735478105451294,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.005058087578194817,
      "loss": 0.8141,
      "num_input_tokens_seen": 1318304,
      "step": 2265
    },
    {
      "epoch": 0.33809949359547214,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.005069258266309205,
      "loss": 0.8023,
      "num_input_tokens_seen": 1320960,
      "step": 2270
    },
    {
      "epoch": 0.33884420613643135,
      "grad_norm": 0.103515625,
      "learning_rate": 0.005080428954423592,
      "loss": 0.846,
      "num_input_tokens_seen": 1323712,
      "step": 2275
    },
    {
      "epoch": 0.33958891867739055,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.00509159964253798,
      "loss": 0.8243,
      "num_input_tokens_seen": 1326752,
      "step": 2280
    },
    {
      "epoch": 0.3403336312183497,
      "grad_norm": 0.0625,
      "learning_rate": 0.005102770330652368,
      "loss": 0.7878,
      "num_input_tokens_seen": 1329600,
      "step": 2285
    },
    {
      "epoch": 0.3410783437593089,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.005113941018766756,
      "loss": 0.8195,
      "num_input_tokens_seen": 1332448,
      "step": 2290
    },
    {
      "epoch": 0.3418230563002681,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.005125111706881144,
      "loss": 0.8056,
      "num_input_tokens_seen": 1335072,
      "step": 2295
    },
    {
      "epoch": 0.3425677688412273,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.005136282394995532,
      "loss": 0.8155,
      "num_input_tokens_seen": 1337856,
      "step": 2300
    },
    {
      "epoch": 0.34331248138218645,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.00514745308310992,
      "loss": 0.7939,
      "num_input_tokens_seen": 1340896,
      "step": 2305
    },
    {
      "epoch": 0.34405719392314565,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.005158623771224308,
      "loss": 0.7912,
      "num_input_tokens_seen": 1343840,
      "step": 2310
    },
    {
      "epoch": 0.34480190646410486,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.005169794459338695,
      "loss": 0.824,
      "num_input_tokens_seen": 1346688,
      "step": 2315
    },
    {
      "epoch": 0.34554661900506406,
      "grad_norm": 0.078125,
      "learning_rate": 0.005180965147453083,
      "loss": 0.8175,
      "num_input_tokens_seen": 1349312,
      "step": 2320
    },
    {
      "epoch": 0.34629133154602326,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.005192135835567471,
      "loss": 0.7999,
      "num_input_tokens_seen": 1352320,
      "step": 2325
    },
    {
      "epoch": 0.3470360440869824,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.005203306523681859,
      "loss": 0.8015,
      "num_input_tokens_seen": 1355168,
      "step": 2330
    },
    {
      "epoch": 0.3477807566279416,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.005214477211796246,
      "loss": 0.7909,
      "num_input_tokens_seen": 1358112,
      "step": 2335
    },
    {
      "epoch": 0.3485254691689008,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.005225647899910634,
      "loss": 0.8202,
      "num_input_tokens_seen": 1360992,
      "step": 2340
    },
    {
      "epoch": 0.34927018170986,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.005236818588025022,
      "loss": 0.7869,
      "num_input_tokens_seen": 1363744,
      "step": 2345
    },
    {
      "epoch": 0.35001489425081916,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.00524798927613941,
      "loss": 0.7978,
      "num_input_tokens_seen": 1366848,
      "step": 2350
    },
    {
      "epoch": 0.35075960679177837,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0052591599642537986,
      "loss": 0.8153,
      "num_input_tokens_seen": 1369536,
      "step": 2355
    },
    {
      "epoch": 0.35150431933273757,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.005270330652368186,
      "loss": 0.805,
      "num_input_tokens_seen": 1372064,
      "step": 2360
    },
    {
      "epoch": 0.35224903187369677,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.005281501340482574,
      "loss": 0.7884,
      "num_input_tokens_seen": 1375136,
      "step": 2365
    },
    {
      "epoch": 0.3529937444146559,
      "grad_norm": 0.080078125,
      "learning_rate": 0.005292672028596962,
      "loss": 0.7968,
      "num_input_tokens_seen": 1377952,
      "step": 2370
    },
    {
      "epoch": 0.3537384569556151,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.00530384271671135,
      "loss": 0.7839,
      "num_input_tokens_seen": 1380864,
      "step": 2375
    },
    {
      "epoch": 0.3544831694965743,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.005315013404825737,
      "loss": 0.7975,
      "num_input_tokens_seen": 1384000,
      "step": 2380
    },
    {
      "epoch": 0.3552278820375335,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.005326184092940125,
      "loss": 0.8095,
      "num_input_tokens_seen": 1387232,
      "step": 2385
    },
    {
      "epoch": 0.35597259457849273,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.005337354781054513,
      "loss": 0.817,
      "num_input_tokens_seen": 1390112,
      "step": 2390
    },
    {
      "epoch": 0.3567173071194519,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.005348525469168901,
      "loss": 0.8174,
      "num_input_tokens_seen": 1393472,
      "step": 2395
    },
    {
      "epoch": 0.3574620196604111,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.005359696157283288,
      "loss": 0.7856,
      "num_input_tokens_seen": 1396288,
      "step": 2400
    },
    {
      "epoch": 0.3582067322013703,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.005370866845397676,
      "loss": 0.8144,
      "num_input_tokens_seen": 1399360,
      "step": 2405
    },
    {
      "epoch": 0.3589514447423295,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0053820375335120645,
      "loss": 0.8188,
      "num_input_tokens_seen": 1402528,
      "step": 2410
    },
    {
      "epoch": 0.35969615728328863,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.0053932082216264525,
      "loss": 0.8356,
      "num_input_tokens_seen": 1405472,
      "step": 2415
    },
    {
      "epoch": 0.36044086982424783,
      "grad_norm": 0.078125,
      "learning_rate": 0.00540437890974084,
      "loss": 0.8073,
      "num_input_tokens_seen": 1408512,
      "step": 2420
    },
    {
      "epoch": 0.36118558236520704,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0054155495978552275,
      "loss": 0.7966,
      "num_input_tokens_seen": 1411264,
      "step": 2425
    },
    {
      "epoch": 0.36193029490616624,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0054267202859696155,
      "loss": 0.805,
      "num_input_tokens_seen": 1413792,
      "step": 2430
    },
    {
      "epoch": 0.3626750074471254,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0054378909740840035,
      "loss": 0.8097,
      "num_input_tokens_seen": 1416800,
      "step": 2435
    },
    {
      "epoch": 0.3634197199880846,
      "grad_norm": 0.09375,
      "learning_rate": 0.0054490616621983914,
      "loss": 0.773,
      "num_input_tokens_seen": 1419552,
      "step": 2440
    },
    {
      "epoch": 0.3641644325290438,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0054602323503127785,
      "loss": 0.8012,
      "num_input_tokens_seen": 1422272,
      "step": 2445
    },
    {
      "epoch": 0.364909145070003,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.0054714030384271665,
      "loss": 0.8058,
      "num_input_tokens_seen": 1425120,
      "step": 2450
    },
    {
      "epoch": 0.3656538576109622,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0054825737265415545,
      "loss": 0.7963,
      "num_input_tokens_seen": 1427840,
      "step": 2455
    },
    {
      "epoch": 0.36639857015192134,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0054937444146559425,
      "loss": 0.8081,
      "num_input_tokens_seen": 1430944,
      "step": 2460
    },
    {
      "epoch": 0.36714328269288055,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00550491510277033,
      "loss": 0.8009,
      "num_input_tokens_seen": 1433792,
      "step": 2465
    },
    {
      "epoch": 0.36788799523383975,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.005516085790884718,
      "loss": 0.7863,
      "num_input_tokens_seen": 1436704,
      "step": 2470
    },
    {
      "epoch": 0.36863270777479895,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.005527256478999106,
      "loss": 0.7968,
      "num_input_tokens_seen": 1439360,
      "step": 2475
    },
    {
      "epoch": 0.3693774203157581,
      "grad_norm": 0.1796875,
      "learning_rate": 0.005538427167113494,
      "loss": 0.8603,
      "num_input_tokens_seen": 1442016,
      "step": 2480
    },
    {
      "epoch": 0.3701221328567173,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.005549597855227882,
      "loss": 0.8337,
      "num_input_tokens_seen": 1444800,
      "step": 2485
    },
    {
      "epoch": 0.3708668453976765,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.005560768543342269,
      "loss": 0.8224,
      "num_input_tokens_seen": 1447488,
      "step": 2490
    },
    {
      "epoch": 0.3716115579386357,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.005571939231456657,
      "loss": 0.8146,
      "num_input_tokens_seen": 1450144,
      "step": 2495
    },
    {
      "epoch": 0.37235627047959485,
      "grad_norm": 0.091796875,
      "learning_rate": 0.005583109919571045,
      "loss": 0.7983,
      "num_input_tokens_seen": 1453184,
      "step": 2500
    },
    {
      "epoch": 0.37310098302055406,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.005594280607685433,
      "loss": 0.7993,
      "num_input_tokens_seen": 1456128,
      "step": 2505
    },
    {
      "epoch": 0.37384569556151326,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.00560545129579982,
      "loss": 0.7952,
      "num_input_tokens_seen": 1459168,
      "step": 2510
    },
    {
      "epoch": 0.37459040810247246,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.005616621983914208,
      "loss": 0.7952,
      "num_input_tokens_seen": 1462080,
      "step": 2515
    },
    {
      "epoch": 0.3753351206434316,
      "grad_norm": 0.109375,
      "learning_rate": 0.005627792672028597,
      "loss": 0.823,
      "num_input_tokens_seen": 1464896,
      "step": 2520
    },
    {
      "epoch": 0.3760798331843908,
      "grad_norm": 0.08984375,
      "learning_rate": 0.005638963360142985,
      "loss": 0.8251,
      "num_input_tokens_seen": 1467648,
      "step": 2525
    },
    {
      "epoch": 0.37682454572535,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.005650134048257372,
      "loss": 0.8115,
      "num_input_tokens_seen": 1470464,
      "step": 2530
    },
    {
      "epoch": 0.3775692582663092,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00566130473637176,
      "loss": 0.7964,
      "num_input_tokens_seen": 1473248,
      "step": 2535
    },
    {
      "epoch": 0.3783139708072684,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.005672475424486148,
      "loss": 0.7992,
      "num_input_tokens_seen": 1476256,
      "step": 2540
    },
    {
      "epoch": 0.37905868334822757,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.005683646112600536,
      "loss": 0.7868,
      "num_input_tokens_seen": 1479136,
      "step": 2545
    },
    {
      "epoch": 0.37980339588918677,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.005694816800714924,
      "loss": 0.8043,
      "num_input_tokens_seen": 1481696,
      "step": 2550
    },
    {
      "epoch": 0.38054810843014597,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.005705987488829311,
      "loss": 0.8233,
      "num_input_tokens_seen": 1484640,
      "step": 2555
    },
    {
      "epoch": 0.3812928209711052,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.005717158176943699,
      "loss": 0.8172,
      "num_input_tokens_seen": 1487296,
      "step": 2560
    },
    {
      "epoch": 0.3820375335120643,
      "grad_norm": 0.076171875,
      "learning_rate": 0.005728328865058087,
      "loss": 0.8236,
      "num_input_tokens_seen": 1490208,
      "step": 2565
    },
    {
      "epoch": 0.3827822460530235,
      "grad_norm": 0.083984375,
      "learning_rate": 0.005739499553172475,
      "loss": 0.8056,
      "num_input_tokens_seen": 1493152,
      "step": 2570
    },
    {
      "epoch": 0.3835269585939827,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.005750670241286863,
      "loss": 0.797,
      "num_input_tokens_seen": 1496032,
      "step": 2575
    },
    {
      "epoch": 0.38427167113494193,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.005761840929401251,
      "loss": 0.8061,
      "num_input_tokens_seen": 1498880,
      "step": 2580
    },
    {
      "epoch": 0.3850163836759011,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.005773011617515639,
      "loss": 0.7983,
      "num_input_tokens_seen": 1501696,
      "step": 2585
    },
    {
      "epoch": 0.3857610962168603,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.005784182305630027,
      "loss": 0.7959,
      "num_input_tokens_seen": 1504448,
      "step": 2590
    },
    {
      "epoch": 0.3865058087578195,
      "grad_norm": 0.072265625,
      "learning_rate": 0.005795352993744414,
      "loss": 0.7863,
      "num_input_tokens_seen": 1507520,
      "step": 2595
    },
    {
      "epoch": 0.3872505212987787,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.005806523681858802,
      "loss": 0.7973,
      "num_input_tokens_seen": 1510176,
      "step": 2600
    },
    {
      "epoch": 0.3879952338397379,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.00581769436997319,
      "loss": 0.799,
      "num_input_tokens_seen": 1512992,
      "step": 2605
    },
    {
      "epoch": 0.38873994638069703,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.005828865058087578,
      "loss": 0.808,
      "num_input_tokens_seen": 1515680,
      "step": 2610
    },
    {
      "epoch": 0.38948465892165623,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.005840035746201966,
      "loss": 0.8053,
      "num_input_tokens_seen": 1518400,
      "step": 2615
    },
    {
      "epoch": 0.39022937146261544,
      "grad_norm": 0.1015625,
      "learning_rate": 0.005851206434316353,
      "loss": 0.8225,
      "num_input_tokens_seen": 1522560,
      "step": 2620
    },
    {
      "epoch": 0.39097408400357464,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.005862377122430741,
      "loss": 0.8285,
      "num_input_tokens_seen": 1525408,
      "step": 2625
    },
    {
      "epoch": 0.3917187965445338,
      "grad_norm": 0.095703125,
      "learning_rate": 0.005873547810545129,
      "loss": 0.8163,
      "num_input_tokens_seen": 1528256,
      "step": 2630
    },
    {
      "epoch": 0.392463509085493,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.005884718498659518,
      "loss": 0.7951,
      "num_input_tokens_seen": 1531264,
      "step": 2635
    },
    {
      "epoch": 0.3932082216264522,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.005895889186773905,
      "loss": 0.8046,
      "num_input_tokens_seen": 1534272,
      "step": 2640
    },
    {
      "epoch": 0.3939529341674114,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.005907059874888293,
      "loss": 0.8192,
      "num_input_tokens_seen": 1537472,
      "step": 2645
    },
    {
      "epoch": 0.39469764670837054,
      "grad_norm": 0.091796875,
      "learning_rate": 0.005918230563002681,
      "loss": 0.8019,
      "num_input_tokens_seen": 1540352,
      "step": 2650
    },
    {
      "epoch": 0.39544235924932974,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.005929401251117069,
      "loss": 0.7914,
      "num_input_tokens_seen": 1543328,
      "step": 2655
    },
    {
      "epoch": 0.39618707179028895,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.005940571939231456,
      "loss": 0.8237,
      "num_input_tokens_seen": 1546272,
      "step": 2660
    },
    {
      "epoch": 0.39693178433124815,
      "grad_norm": 0.0859375,
      "learning_rate": 0.005951742627345844,
      "loss": 0.8177,
      "num_input_tokens_seen": 1549280,
      "step": 2665
    },
    {
      "epoch": 0.39767649687220735,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.005962913315460232,
      "loss": 0.8201,
      "num_input_tokens_seen": 1552256,
      "step": 2670
    },
    {
      "epoch": 0.3984212094131665,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.00597408400357462,
      "loss": 0.8072,
      "num_input_tokens_seen": 1555168,
      "step": 2675
    },
    {
      "epoch": 0.3991659219541257,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.005985254691689008,
      "loss": 0.8045,
      "num_input_tokens_seen": 1558112,
      "step": 2680
    },
    {
      "epoch": 0.3999106344950849,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.005996425379803395,
      "loss": 0.7813,
      "num_input_tokens_seen": 1561184,
      "step": 2685
    },
    {
      "epoch": 0.4006553470360441,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.006007596067917784,
      "loss": 0.8273,
      "num_input_tokens_seen": 1563936,
      "step": 2690
    },
    {
      "epoch": 0.40140005957700325,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.006018766756032172,
      "loss": 0.8008,
      "num_input_tokens_seen": 1566592,
      "step": 2695
    },
    {
      "epoch": 0.40214477211796246,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00602993744414656,
      "loss": 0.8193,
      "num_input_tokens_seen": 1569568,
      "step": 2700
    },
    {
      "epoch": 0.40288948465892166,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.006041108132260947,
      "loss": 0.7988,
      "num_input_tokens_seen": 1572384,
      "step": 2705
    },
    {
      "epoch": 0.40363419719988086,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.006052278820375335,
      "loss": 0.7884,
      "num_input_tokens_seen": 1575296,
      "step": 2710
    },
    {
      "epoch": 0.40437890974084,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.006063449508489723,
      "loss": 0.802,
      "num_input_tokens_seen": 1578208,
      "step": 2715
    },
    {
      "epoch": 0.4051236222817992,
      "grad_norm": 0.1328125,
      "learning_rate": 0.006074620196604111,
      "loss": 0.8236,
      "num_input_tokens_seen": 1581408,
      "step": 2720
    },
    {
      "epoch": 0.4058683348227584,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.006085790884718498,
      "loss": 0.8018,
      "num_input_tokens_seen": 1584096,
      "step": 2725
    },
    {
      "epoch": 0.4066130473637176,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.006096961572832886,
      "loss": 0.8073,
      "num_input_tokens_seen": 1587008,
      "step": 2730
    },
    {
      "epoch": 0.4073577599046768,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.006108132260947274,
      "loss": 0.7903,
      "num_input_tokens_seen": 1589824,
      "step": 2735
    },
    {
      "epoch": 0.40810247244563597,
      "grad_norm": 0.15625,
      "learning_rate": 0.006119302949061662,
      "loss": 0.8195,
      "num_input_tokens_seen": 1592800,
      "step": 2740
    },
    {
      "epoch": 0.40884718498659517,
      "grad_norm": 0.107421875,
      "learning_rate": 0.00613047363717605,
      "loss": 0.7999,
      "num_input_tokens_seen": 1595808,
      "step": 2745
    },
    {
      "epoch": 0.40959189752755437,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.006141644325290438,
      "loss": 0.8034,
      "num_input_tokens_seen": 1598592,
      "step": 2750
    },
    {
      "epoch": 0.4103366100685136,
      "grad_norm": 0.142578125,
      "learning_rate": 0.006152815013404826,
      "loss": 0.8029,
      "num_input_tokens_seen": 1601824,
      "step": 2755
    },
    {
      "epoch": 0.4110813226094727,
      "grad_norm": 0.142578125,
      "learning_rate": 0.006163985701519214,
      "loss": 0.8185,
      "num_input_tokens_seen": 1604672,
      "step": 2760
    },
    {
      "epoch": 0.4118260351504319,
      "grad_norm": 0.080078125,
      "learning_rate": 0.006175156389633602,
      "loss": 0.8116,
      "num_input_tokens_seen": 1607584,
      "step": 2765
    },
    {
      "epoch": 0.4125707476913911,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.006186327077747989,
      "loss": 0.8112,
      "num_input_tokens_seen": 1610496,
      "step": 2770
    },
    {
      "epoch": 0.41331546023235033,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.006197497765862377,
      "loss": 0.8162,
      "num_input_tokens_seen": 1613472,
      "step": 2775
    },
    {
      "epoch": 0.4140601727733095,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.006208668453976765,
      "loss": 0.7906,
      "num_input_tokens_seen": 1616416,
      "step": 2780
    },
    {
      "epoch": 0.4148048853142687,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.006219839142091153,
      "loss": 0.8043,
      "num_input_tokens_seen": 1619456,
      "step": 2785
    },
    {
      "epoch": 0.4155495978552279,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.00623100983020554,
      "loss": 0.7984,
      "num_input_tokens_seen": 1622528,
      "step": 2790
    },
    {
      "epoch": 0.4162943103961871,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.006242180518319928,
      "loss": 0.8035,
      "num_input_tokens_seen": 1625280,
      "step": 2795
    },
    {
      "epoch": 0.4170390229371463,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.006253351206434316,
      "loss": 0.7976,
      "num_input_tokens_seen": 1628192,
      "step": 2800
    },
    {
      "epoch": 0.41778373547810543,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0062645218945487045,
      "loss": 0.7845,
      "num_input_tokens_seen": 1631264,
      "step": 2805
    },
    {
      "epoch": 0.41852844801906464,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.0062756925826630925,
      "loss": 0.8,
      "num_input_tokens_seen": 1634080,
      "step": 2810
    },
    {
      "epoch": 0.41927316056002384,
      "grad_norm": 0.255859375,
      "learning_rate": 0.00628686327077748,
      "loss": 0.818,
      "num_input_tokens_seen": 1637056,
      "step": 2815
    },
    {
      "epoch": 0.42001787310098304,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.006298033958891868,
      "loss": 0.8472,
      "num_input_tokens_seen": 1640064,
      "step": 2820
    },
    {
      "epoch": 0.4207625856419422,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0063092046470062555,
      "loss": 0.8414,
      "num_input_tokens_seen": 1642784,
      "step": 2825
    },
    {
      "epoch": 0.4215072981829014,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0063203753351206435,
      "loss": 0.7909,
      "num_input_tokens_seen": 1645536,
      "step": 2830
    },
    {
      "epoch": 0.4222520107238606,
      "grad_norm": 0.06640625,
      "learning_rate": 0.006331546023235031,
      "loss": 0.7871,
      "num_input_tokens_seen": 1648128,
      "step": 2835
    },
    {
      "epoch": 0.4229967232648198,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.006342716711349419,
      "loss": 0.7893,
      "num_input_tokens_seen": 1650880,
      "step": 2840
    },
    {
      "epoch": 0.42374143580577894,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0063538873994638066,
      "loss": 0.806,
      "num_input_tokens_seen": 1653664,
      "step": 2845
    },
    {
      "epoch": 0.42448614834673815,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0063650580875781945,
      "loss": 0.7705,
      "num_input_tokens_seen": 1656640,
      "step": 2850
    },
    {
      "epoch": 0.42523086088769735,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.006376228775692582,
      "loss": 0.8076,
      "num_input_tokens_seen": 1659296,
      "step": 2855
    },
    {
      "epoch": 0.42597557342865655,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0063873994638069705,
      "loss": 0.819,
      "num_input_tokens_seen": 1662112,
      "step": 2860
    },
    {
      "epoch": 0.42672028596961575,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.006398570151921358,
      "loss": 0.8228,
      "num_input_tokens_seen": 1664832,
      "step": 2865
    },
    {
      "epoch": 0.4274649985105749,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.006409740840035746,
      "loss": 0.7994,
      "num_input_tokens_seen": 1667872,
      "step": 2870
    },
    {
      "epoch": 0.4282097110515341,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.006420911528150134,
      "loss": 0.8089,
      "num_input_tokens_seen": 1670592,
      "step": 2875
    },
    {
      "epoch": 0.4289544235924933,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0064320822162645215,
      "loss": 0.7864,
      "num_input_tokens_seen": 1673472,
      "step": 2880
    },
    {
      "epoch": 0.4296991361334525,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0064432529043789094,
      "loss": 0.7938,
      "num_input_tokens_seen": 1676352,
      "step": 2885
    },
    {
      "epoch": 0.43044384867441166,
      "grad_norm": 0.091796875,
      "learning_rate": 0.006454423592493297,
      "loss": 0.8229,
      "num_input_tokens_seen": 1679328,
      "step": 2890
    },
    {
      "epoch": 0.43118856121537086,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.006465594280607685,
      "loss": 0.7961,
      "num_input_tokens_seen": 1682240,
      "step": 2895
    },
    {
      "epoch": 0.43193327375633006,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.0064767649687220725,
      "loss": 0.8511,
      "num_input_tokens_seen": 1685120,
      "step": 2900
    },
    {
      "epoch": 0.43267798629728926,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0064879356568364605,
      "loss": 0.8197,
      "num_input_tokens_seen": 1688000,
      "step": 2905
    },
    {
      "epoch": 0.4334226988382484,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.006499106344950848,
      "loss": 0.7966,
      "num_input_tokens_seen": 1690816,
      "step": 2910
    },
    {
      "epoch": 0.4341674113792076,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.006510277033065237,
      "loss": 0.8212,
      "num_input_tokens_seen": 1693600,
      "step": 2915
    },
    {
      "epoch": 0.4349121239201668,
      "grad_norm": 0.212890625,
      "learning_rate": 0.006521447721179624,
      "loss": 0.8079,
      "num_input_tokens_seen": 1696544,
      "step": 2920
    },
    {
      "epoch": 0.435656836461126,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.006532618409294012,
      "loss": 0.7943,
      "num_input_tokens_seen": 1699200,
      "step": 2925
    },
    {
      "epoch": 0.4364015490020852,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0065437890974084,
      "loss": 0.8185,
      "num_input_tokens_seen": 1702080,
      "step": 2930
    },
    {
      "epoch": 0.43714626154304437,
      "grad_norm": 0.068359375,
      "learning_rate": 0.006554959785522788,
      "loss": 0.7901,
      "num_input_tokens_seen": 1705088,
      "step": 2935
    },
    {
      "epoch": 0.43789097408400357,
      "grad_norm": 0.06640625,
      "learning_rate": 0.006566130473637176,
      "loss": 0.8147,
      "num_input_tokens_seen": 1707808,
      "step": 2940
    },
    {
      "epoch": 0.4386356866249628,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.006577301161751563,
      "loss": 0.8102,
      "num_input_tokens_seen": 1710592,
      "step": 2945
    },
    {
      "epoch": 0.439380399165922,
      "grad_norm": 0.068359375,
      "learning_rate": 0.006588471849865951,
      "loss": 0.7644,
      "num_input_tokens_seen": 1713664,
      "step": 2950
    },
    {
      "epoch": 0.4401251117068811,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.006599642537980339,
      "loss": 0.798,
      "num_input_tokens_seen": 1716512,
      "step": 2955
    },
    {
      "epoch": 0.4408698242478403,
      "grad_norm": 0.09375,
      "learning_rate": 0.006610813226094727,
      "loss": 0.7991,
      "num_input_tokens_seen": 1719424,
      "step": 2960
    },
    {
      "epoch": 0.4416145367887995,
      "grad_norm": 0.109375,
      "learning_rate": 0.006621983914209114,
      "loss": 0.7762,
      "num_input_tokens_seen": 1722208,
      "step": 2965
    },
    {
      "epoch": 0.44235924932975873,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.006633154602323503,
      "loss": 0.8036,
      "num_input_tokens_seen": 1725248,
      "step": 2970
    },
    {
      "epoch": 0.4431039618707179,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.006644325290437891,
      "loss": 0.8234,
      "num_input_tokens_seen": 1728192,
      "step": 2975
    },
    {
      "epoch": 0.4438486744116771,
      "grad_norm": 0.103515625,
      "learning_rate": 0.006655495978552279,
      "loss": 0.8216,
      "num_input_tokens_seen": 1731328,
      "step": 2980
    },
    {
      "epoch": 0.4445933869526363,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.006666666666666666,
      "loss": 0.7941,
      "num_input_tokens_seen": 1734304,
      "step": 2985
    },
    {
      "epoch": 0.4453380994935955,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.006677837354781054,
      "loss": 0.7999,
      "num_input_tokens_seen": 1736896,
      "step": 2990
    },
    {
      "epoch": 0.4460828120345547,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.006689008042895442,
      "loss": 0.7549,
      "num_input_tokens_seen": 1739712,
      "step": 2995
    },
    {
      "epoch": 0.44682752457551383,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.00670017873100983,
      "loss": 0.8194,
      "num_input_tokens_seen": 1742528,
      "step": 3000
    },
    {
      "epoch": 0.44757223711647304,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.006711349419124218,
      "loss": 0.8239,
      "num_input_tokens_seen": 1745536,
      "step": 3005
    },
    {
      "epoch": 0.44831694965743224,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.006722520107238605,
      "loss": 0.8207,
      "num_input_tokens_seen": 1748448,
      "step": 3010
    },
    {
      "epoch": 0.44906166219839144,
      "grad_norm": 0.181640625,
      "learning_rate": 0.006733690795352993,
      "loss": 0.7745,
      "num_input_tokens_seen": 1751168,
      "step": 3015
    },
    {
      "epoch": 0.4498063747393506,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.006744861483467381,
      "loss": 0.7886,
      "num_input_tokens_seen": 1753760,
      "step": 3020
    },
    {
      "epoch": 0.4505510872803098,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.006756032171581769,
      "loss": 0.8137,
      "num_input_tokens_seen": 1756832,
      "step": 3025
    },
    {
      "epoch": 0.451295799821269,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.006767202859696157,
      "loss": 0.8085,
      "num_input_tokens_seen": 1759840,
      "step": 3030
    },
    {
      "epoch": 0.4520405123622282,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.006778373547810545,
      "loss": 0.8062,
      "num_input_tokens_seen": 1762752,
      "step": 3035
    },
    {
      "epoch": 0.45278522490318734,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.006789544235924933,
      "loss": 0.8184,
      "num_input_tokens_seen": 1765568,
      "step": 3040
    },
    {
      "epoch": 0.45352993744414655,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.006800714924039321,
      "loss": 0.8258,
      "num_input_tokens_seen": 1768160,
      "step": 3045
    },
    {
      "epoch": 0.45427464998510575,
      "grad_norm": 0.095703125,
      "learning_rate": 0.006811885612153708,
      "loss": 0.8131,
      "num_input_tokens_seen": 1771072,
      "step": 3050
    },
    {
      "epoch": 0.45501936252606495,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.006823056300268096,
      "loss": 0.7809,
      "num_input_tokens_seen": 1774016,
      "step": 3055
    },
    {
      "epoch": 0.45576407506702415,
      "grad_norm": 0.0703125,
      "learning_rate": 0.006834226988382484,
      "loss": 0.7916,
      "num_input_tokens_seen": 1777088,
      "step": 3060
    },
    {
      "epoch": 0.4565087876079833,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.006845397676496872,
      "loss": 0.7893,
      "num_input_tokens_seen": 1779840,
      "step": 3065
    },
    {
      "epoch": 0.4572535001489425,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.00685656836461126,
      "loss": 0.8032,
      "num_input_tokens_seen": 1782656,
      "step": 3070
    },
    {
      "epoch": 0.4579982126899017,
      "grad_norm": 0.0703125,
      "learning_rate": 0.006867739052725647,
      "loss": 0.7977,
      "num_input_tokens_seen": 1785792,
      "step": 3075
    },
    {
      "epoch": 0.4587429252308609,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.006878909740840035,
      "loss": 0.814,
      "num_input_tokens_seen": 1788896,
      "step": 3080
    },
    {
      "epoch": 0.45948763777182006,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.006890080428954424,
      "loss": 0.7915,
      "num_input_tokens_seen": 1791904,
      "step": 3085
    },
    {
      "epoch": 0.46023235031277926,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.006901251117068812,
      "loss": 0.8377,
      "num_input_tokens_seen": 1794816,
      "step": 3090
    },
    {
      "epoch": 0.46097706285373846,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.006912421805183199,
      "loss": 0.786,
      "num_input_tokens_seen": 1797536,
      "step": 3095
    },
    {
      "epoch": 0.46172177539469766,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.006923592493297587,
      "loss": 0.8338,
      "num_input_tokens_seen": 1800512,
      "step": 3100
    },
    {
      "epoch": 0.4624664879356568,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.006934763181411975,
      "loss": 0.8189,
      "num_input_tokens_seen": 1803328,
      "step": 3105
    },
    {
      "epoch": 0.463211200476616,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.006945933869526363,
      "loss": 0.824,
      "num_input_tokens_seen": 1806176,
      "step": 3110
    },
    {
      "epoch": 0.4639559130175752,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00695710455764075,
      "loss": 0.8025,
      "num_input_tokens_seen": 1809056,
      "step": 3115
    },
    {
      "epoch": 0.4647006255585344,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.006968275245755138,
      "loss": 0.8,
      "num_input_tokens_seen": 1812032,
      "step": 3120
    },
    {
      "epoch": 0.4654453380994936,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.006979445933869526,
      "loss": 0.8116,
      "num_input_tokens_seen": 1814976,
      "step": 3125
    },
    {
      "epoch": 0.46619005064045277,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.006990616621983914,
      "loss": 0.7993,
      "num_input_tokens_seen": 1818080,
      "step": 3130
    },
    {
      "epoch": 0.46693476318141197,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.007001787310098302,
      "loss": 0.7906,
      "num_input_tokens_seen": 1821216,
      "step": 3135
    },
    {
      "epoch": 0.4676794757223712,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00701295799821269,
      "loss": 0.8016,
      "num_input_tokens_seen": 1824000,
      "step": 3140
    },
    {
      "epoch": 0.4684241882633304,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.007024128686327078,
      "loss": 0.8138,
      "num_input_tokens_seen": 1826560,
      "step": 3145
    },
    {
      "epoch": 0.4691689008042895,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.007035299374441466,
      "loss": 0.8105,
      "num_input_tokens_seen": 1829248,
      "step": 3150
    },
    {
      "epoch": 0.4699136133452487,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.007046470062555854,
      "loss": 0.8002,
      "num_input_tokens_seen": 1832032,
      "step": 3155
    },
    {
      "epoch": 0.47065832588620793,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.007057640750670241,
      "loss": 0.7858,
      "num_input_tokens_seen": 1835328,
      "step": 3160
    },
    {
      "epoch": 0.47140303842716713,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.007068811438784629,
      "loss": 0.8007,
      "num_input_tokens_seen": 1838176,
      "step": 3165
    },
    {
      "epoch": 0.4721477509681263,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.007079982126899017,
      "loss": 0.8187,
      "num_input_tokens_seen": 1841824,
      "step": 3170
    },
    {
      "epoch": 0.4728924635090855,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.007091152815013405,
      "loss": 0.8047,
      "num_input_tokens_seen": 1845024,
      "step": 3175
    },
    {
      "epoch": 0.4736371760500447,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.007102323503127792,
      "loss": 0.8032,
      "num_input_tokens_seen": 1847744,
      "step": 3180
    },
    {
      "epoch": 0.4743818885910039,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.00711349419124218,
      "loss": 0.7917,
      "num_input_tokens_seen": 1850560,
      "step": 3185
    },
    {
      "epoch": 0.4751266011319631,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.007124664879356568,
      "loss": 0.7921,
      "num_input_tokens_seen": 1853504,
      "step": 3190
    },
    {
      "epoch": 0.47587131367292224,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.007135835567470956,
      "loss": 0.7934,
      "num_input_tokens_seen": 1856480,
      "step": 3195
    },
    {
      "epoch": 0.47661602621388144,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.007147006255585345,
      "loss": 0.8105,
      "num_input_tokens_seen": 1859392,
      "step": 3200
    },
    {
      "epoch": 0.47736073875484064,
      "grad_norm": 0.072265625,
      "learning_rate": 0.007158176943699732,
      "loss": 0.7797,
      "num_input_tokens_seen": 1862240,
      "step": 3205
    },
    {
      "epoch": 0.47810545129579984,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.00716934763181412,
      "loss": 0.7975,
      "num_input_tokens_seen": 1865024,
      "step": 3210
    },
    {
      "epoch": 0.478850163836759,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.007180518319928508,
      "loss": 0.7932,
      "num_input_tokens_seen": 1867776,
      "step": 3215
    },
    {
      "epoch": 0.4795948763777182,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.007191689008042896,
      "loss": 0.8089,
      "num_input_tokens_seen": 1870752,
      "step": 3220
    },
    {
      "epoch": 0.4803395889186774,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.007202859696157283,
      "loss": 0.7776,
      "num_input_tokens_seen": 1873888,
      "step": 3225
    },
    {
      "epoch": 0.4810843014596366,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.007214030384271671,
      "loss": 0.7947,
      "num_input_tokens_seen": 1876896,
      "step": 3230
    },
    {
      "epoch": 0.48182901400059575,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.007225201072386059,
      "loss": 0.7909,
      "num_input_tokens_seen": 1879360,
      "step": 3235
    },
    {
      "epoch": 0.48257372654155495,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.007236371760500447,
      "loss": 0.8397,
      "num_input_tokens_seen": 1882592,
      "step": 3240
    },
    {
      "epoch": 0.48331843908251415,
      "grad_norm": 0.091796875,
      "learning_rate": 0.007247542448614834,
      "loss": 0.8302,
      "num_input_tokens_seen": 1885728,
      "step": 3245
    },
    {
      "epoch": 0.48406315162347335,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.007258713136729222,
      "loss": 0.7873,
      "num_input_tokens_seen": 1888896,
      "step": 3250
    },
    {
      "epoch": 0.48480786416443256,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0072698838248436105,
      "loss": 0.814,
      "num_input_tokens_seen": 1892000,
      "step": 3255
    },
    {
      "epoch": 0.4855525767053917,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0072810545129579985,
      "loss": 0.8294,
      "num_input_tokens_seen": 1894784,
      "step": 3260
    },
    {
      "epoch": 0.4862972892463509,
      "grad_norm": 0.056640625,
      "learning_rate": 0.007292225201072386,
      "loss": 0.8205,
      "num_input_tokens_seen": 1897440,
      "step": 3265
    },
    {
      "epoch": 0.4870420017873101,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0073033958891867735,
      "loss": 0.8331,
      "num_input_tokens_seen": 1900352,
      "step": 3270
    },
    {
      "epoch": 0.4877867143282693,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0073145665773011615,
      "loss": 0.8037,
      "num_input_tokens_seen": 1903392,
      "step": 3275
    },
    {
      "epoch": 0.48853142686922846,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0073257372654155495,
      "loss": 0.8034,
      "num_input_tokens_seen": 1906176,
      "step": 3280
    },
    {
      "epoch": 0.48927613941018766,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0073369079535299375,
      "loss": 0.8142,
      "num_input_tokens_seen": 1909024,
      "step": 3285
    },
    {
      "epoch": 0.49002085195114686,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0073480786416443246,
      "loss": 0.8086,
      "num_input_tokens_seen": 1912160,
      "step": 3290
    },
    {
      "epoch": 0.49076556449210607,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0073592493297587125,
      "loss": 0.8198,
      "num_input_tokens_seen": 1915104,
      "step": 3295
    },
    {
      "epoch": 0.4915102770330652,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0073704200178731005,
      "loss": 0.7962,
      "num_input_tokens_seen": 1917984,
      "step": 3300
    },
    {
      "epoch": 0.4922549895740244,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0073815907059874885,
      "loss": 0.8198,
      "num_input_tokens_seen": 1921152,
      "step": 3305
    },
    {
      "epoch": 0.4929997021149836,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.007392761394101876,
      "loss": 0.8,
      "num_input_tokens_seen": 1923872,
      "step": 3310
    },
    {
      "epoch": 0.4937444146559428,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.007403932082216264,
      "loss": 0.765,
      "num_input_tokens_seen": 1926848,
      "step": 3315
    },
    {
      "epoch": 0.494489127196902,
      "grad_norm": 0.23828125,
      "learning_rate": 0.007415102770330652,
      "loss": 0.7882,
      "num_input_tokens_seen": 1929920,
      "step": 3320
    },
    {
      "epoch": 0.49523383973786117,
      "grad_norm": 0.083984375,
      "learning_rate": 0.00742627345844504,
      "loss": 0.8464,
      "num_input_tokens_seen": 1932736,
      "step": 3325
    },
    {
      "epoch": 0.4959785522788204,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0074374441465594274,
      "loss": 0.8641,
      "num_input_tokens_seen": 1935392,
      "step": 3330
    },
    {
      "epoch": 0.4967232648197796,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.007448614834673815,
      "loss": 0.8144,
      "num_input_tokens_seen": 1938176,
      "step": 3335
    },
    {
      "epoch": 0.4974679773607388,
      "grad_norm": 0.091796875,
      "learning_rate": 0.007459785522788203,
      "loss": 0.8023,
      "num_input_tokens_seen": 1941056,
      "step": 3340
    },
    {
      "epoch": 0.4982126899016979,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.007470956210902591,
      "loss": 0.8095,
      "num_input_tokens_seen": 1943936,
      "step": 3345
    },
    {
      "epoch": 0.4989574024426571,
      "grad_norm": 0.080078125,
      "learning_rate": 0.007482126899016979,
      "loss": 0.7966,
      "num_input_tokens_seen": 1946784,
      "step": 3350
    },
    {
      "epoch": 0.49970211498361633,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.007493297587131366,
      "loss": 0.7973,
      "num_input_tokens_seen": 1949760,
      "step": 3355
    },
    {
      "epoch": 0.5004468275245755,
      "grad_norm": 0.091796875,
      "learning_rate": 0.007504468275245755,
      "loss": 0.7992,
      "num_input_tokens_seen": 1952352,
      "step": 3360
    },
    {
      "epoch": 0.5011915400655347,
      "grad_norm": 0.0546875,
      "learning_rate": 0.007515638963360142,
      "loss": 0.7846,
      "num_input_tokens_seen": 1955072,
      "step": 3365
    },
    {
      "epoch": 0.5019362526064939,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.00752680965147453,
      "loss": 0.8269,
      "num_input_tokens_seen": 1957696,
      "step": 3370
    },
    {
      "epoch": 0.5026809651474531,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.007537980339588918,
      "loss": 0.8188,
      "num_input_tokens_seen": 1960736,
      "step": 3375
    },
    {
      "epoch": 0.5034256776884123,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.007549151027703306,
      "loss": 0.8032,
      "num_input_tokens_seen": 1963616,
      "step": 3380
    },
    {
      "epoch": 0.5041703902293715,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.007560321715817694,
      "loss": 0.7796,
      "num_input_tokens_seen": 1966528,
      "step": 3385
    },
    {
      "epoch": 0.5049151027703307,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.007571492403932082,
      "loss": 0.8188,
      "num_input_tokens_seen": 1969312,
      "step": 3390
    },
    {
      "epoch": 0.5056598153112899,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00758266309204647,
      "loss": 0.8104,
      "num_input_tokens_seen": 1972096,
      "step": 3395
    },
    {
      "epoch": 0.506404527852249,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.007593833780160858,
      "loss": 0.8005,
      "num_input_tokens_seen": 1975040,
      "step": 3400
    },
    {
      "epoch": 0.5071492403932082,
      "grad_norm": 0.083984375,
      "learning_rate": 0.007605004468275246,
      "loss": 0.7993,
      "num_input_tokens_seen": 1977856,
      "step": 3405
    },
    {
      "epoch": 0.5078939529341674,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.007616175156389632,
      "loss": 0.7996,
      "num_input_tokens_seen": 1980608,
      "step": 3410
    },
    {
      "epoch": 0.5086386654751266,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00762734584450402,
      "loss": 0.8088,
      "num_input_tokens_seen": 1983328,
      "step": 3415
    },
    {
      "epoch": 0.5093833780160858,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.007638516532618409,
      "loss": 0.8219,
      "num_input_tokens_seen": 1986400,
      "step": 3420
    },
    {
      "epoch": 0.510128090557045,
      "grad_norm": 0.0859375,
      "learning_rate": 0.007649687220732797,
      "loss": 0.8118,
      "num_input_tokens_seen": 1989248,
      "step": 3425
    },
    {
      "epoch": 0.5108728030980042,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.007660857908847185,
      "loss": 0.8146,
      "num_input_tokens_seen": 1992224,
      "step": 3430
    },
    {
      "epoch": 0.5116175156389634,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.007672028596961573,
      "loss": 0.8196,
      "num_input_tokens_seen": 1995136,
      "step": 3435
    },
    {
      "epoch": 0.5123622281799225,
      "grad_norm": 0.103515625,
      "learning_rate": 0.007683199285075961,
      "loss": 0.8153,
      "num_input_tokens_seen": 1997792,
      "step": 3440
    },
    {
      "epoch": 0.5131069407208817,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.007694369973190349,
      "loss": 0.79,
      "num_input_tokens_seen": 2000896,
      "step": 3445
    },
    {
      "epoch": 0.5138516532618409,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.007705540661304735,
      "loss": 0.8089,
      "num_input_tokens_seen": 2003840,
      "step": 3450
    },
    {
      "epoch": 0.5145963658028001,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.007716711349419123,
      "loss": 0.813,
      "num_input_tokens_seen": 2006496,
      "step": 3455
    },
    {
      "epoch": 0.5153410783437593,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.007727882037533511,
      "loss": 0.8309,
      "num_input_tokens_seen": 2009568,
      "step": 3460
    },
    {
      "epoch": 0.5160857908847185,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.007739052725647899,
      "loss": 0.8067,
      "num_input_tokens_seen": 2012384,
      "step": 3465
    },
    {
      "epoch": 0.5168305034256777,
      "grad_norm": 0.072265625,
      "learning_rate": 0.007750223413762287,
      "loss": 0.796,
      "num_input_tokens_seen": 2015360,
      "step": 3470
    },
    {
      "epoch": 0.5175752159666369,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.007761394101876675,
      "loss": 0.8089,
      "num_input_tokens_seen": 2018048,
      "step": 3475
    },
    {
      "epoch": 0.5183199285075961,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.007772564789991064,
      "loss": 0.7924,
      "num_input_tokens_seen": 2020960,
      "step": 3480
    },
    {
      "epoch": 0.5190646410485552,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.007783735478105452,
      "loss": 0.8008,
      "num_input_tokens_seen": 2023744,
      "step": 3485
    },
    {
      "epoch": 0.5198093535895144,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.00779490616621984,
      "loss": 0.7945,
      "num_input_tokens_seen": 2026592,
      "step": 3490
    },
    {
      "epoch": 0.5205540661304736,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.007806076854334226,
      "loss": 0.8172,
      "num_input_tokens_seen": 2029728,
      "step": 3495
    },
    {
      "epoch": 0.5212987786714328,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.007817247542448613,
      "loss": 0.8178,
      "num_input_tokens_seen": 2032672,
      "step": 3500
    },
    {
      "epoch": 0.522043491212392,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.007828418230563003,
      "loss": 0.796,
      "num_input_tokens_seen": 2035808,
      "step": 3505
    },
    {
      "epoch": 0.5227882037533512,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.007839588918677391,
      "loss": 0.8049,
      "num_input_tokens_seen": 2038720,
      "step": 3510
    },
    {
      "epoch": 0.5235329162943104,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.007850759606791779,
      "loss": 0.7857,
      "num_input_tokens_seen": 2041664,
      "step": 3515
    },
    {
      "epoch": 0.5242776288352696,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.007861930294906167,
      "loss": 0.8076,
      "num_input_tokens_seen": 2044800,
      "step": 3520
    },
    {
      "epoch": 0.5250223413762288,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.007873100983020555,
      "loss": 0.8136,
      "num_input_tokens_seen": 2047392,
      "step": 3525
    },
    {
      "epoch": 0.5257670539171879,
      "grad_norm": 0.078125,
      "learning_rate": 0.007884271671134943,
      "loss": 0.816,
      "num_input_tokens_seen": 2050496,
      "step": 3530
    },
    {
      "epoch": 0.5265117664581471,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00789544235924933,
      "loss": 0.8097,
      "num_input_tokens_seen": 2053984,
      "step": 3535
    },
    {
      "epoch": 0.5272564789991063,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.007906613047363717,
      "loss": 0.7966,
      "num_input_tokens_seen": 2057120,
      "step": 3540
    },
    {
      "epoch": 0.5280011915400655,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.007917783735478105,
      "loss": 0.7939,
      "num_input_tokens_seen": 2059680,
      "step": 3545
    },
    {
      "epoch": 0.5287459040810247,
      "grad_norm": 0.078125,
      "learning_rate": 0.007928954423592493,
      "loss": 0.8063,
      "num_input_tokens_seen": 2062560,
      "step": 3550
    },
    {
      "epoch": 0.5294906166219839,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00794012511170688,
      "loss": 0.8119,
      "num_input_tokens_seen": 2065536,
      "step": 3555
    },
    {
      "epoch": 0.5302353291629431,
      "grad_norm": 0.115234375,
      "learning_rate": 0.007951295799821269,
      "loss": 0.7921,
      "num_input_tokens_seen": 2068416,
      "step": 3560
    },
    {
      "epoch": 0.5309800417039023,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.007962466487935657,
      "loss": 0.7863,
      "num_input_tokens_seen": 2071328,
      "step": 3565
    },
    {
      "epoch": 0.5317247542448614,
      "grad_norm": 0.07421875,
      "learning_rate": 0.007973637176050045,
      "loss": 0.8061,
      "num_input_tokens_seen": 2074016,
      "step": 3570
    },
    {
      "epoch": 0.5324694667858206,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.007984807864164433,
      "loss": 0.7856,
      "num_input_tokens_seen": 2076832,
      "step": 3575
    },
    {
      "epoch": 0.5332141793267798,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.007995978552278819,
      "loss": 0.8238,
      "num_input_tokens_seen": 2079648,
      "step": 3580
    },
    {
      "epoch": 0.533958891867739,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.008007149240393207,
      "loss": 0.8098,
      "num_input_tokens_seen": 2082656,
      "step": 3585
    },
    {
      "epoch": 0.5347036044086982,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.008018319928507595,
      "loss": 0.7879,
      "num_input_tokens_seen": 2085664,
      "step": 3590
    },
    {
      "epoch": 0.5354483169496574,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.008029490616621983,
      "loss": 0.8279,
      "num_input_tokens_seen": 2088768,
      "step": 3595
    },
    {
      "epoch": 0.5361930294906166,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.00804066130473637,
      "loss": 0.7891,
      "num_input_tokens_seen": 2091648,
      "step": 3600
    },
    {
      "epoch": 0.5369377420315758,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.008051831992850759,
      "loss": 0.8175,
      "num_input_tokens_seen": 2094816,
      "step": 3605
    },
    {
      "epoch": 0.537682454572535,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.008063002680965147,
      "loss": 0.7905,
      "num_input_tokens_seen": 2097696,
      "step": 3610
    },
    {
      "epoch": 0.5384271671134941,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.008074173369079535,
      "loss": 0.8021,
      "num_input_tokens_seen": 2100320,
      "step": 3615
    },
    {
      "epoch": 0.5391718796544533,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.008085344057193924,
      "loss": 0.8194,
      "num_input_tokens_seen": 2103328,
      "step": 3620
    },
    {
      "epoch": 0.5399165921954125,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.00809651474530831,
      "loss": 0.7852,
      "num_input_tokens_seen": 2106240,
      "step": 3625
    },
    {
      "epoch": 0.5406613047363718,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.008107685433422699,
      "loss": 0.7832,
      "num_input_tokens_seen": 2109280,
      "step": 3630
    },
    {
      "epoch": 0.541406017277331,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.008118856121537087,
      "loss": 0.8327,
      "num_input_tokens_seen": 2111936,
      "step": 3635
    },
    {
      "epoch": 0.5421507298182902,
      "grad_norm": 0.08984375,
      "learning_rate": 0.008130026809651475,
      "loss": 0.7872,
      "num_input_tokens_seen": 2114624,
      "step": 3640
    },
    {
      "epoch": 0.5428954423592494,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.008141197497765863,
      "loss": 0.7982,
      "num_input_tokens_seen": 2117664,
      "step": 3645
    },
    {
      "epoch": 0.5436401549002086,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00815236818588025,
      "loss": 0.8244,
      "num_input_tokens_seen": 2120448,
      "step": 3650
    },
    {
      "epoch": 0.5443848674411678,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.008163538873994639,
      "loss": 0.8037,
      "num_input_tokens_seen": 2123552,
      "step": 3655
    },
    {
      "epoch": 0.5451295799821269,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.008174709562109026,
      "loss": 0.8228,
      "num_input_tokens_seen": 2126688,
      "step": 3660
    },
    {
      "epoch": 0.5458742925230861,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.008185880250223414,
      "loss": 0.7878,
      "num_input_tokens_seen": 2130432,
      "step": 3665
    },
    {
      "epoch": 0.5466190050640453,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0081970509383378,
      "loss": 0.8284,
      "num_input_tokens_seen": 2133376,
      "step": 3670
    },
    {
      "epoch": 0.5473637176050045,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.008208221626452189,
      "loss": 0.8404,
      "num_input_tokens_seen": 2136576,
      "step": 3675
    },
    {
      "epoch": 0.5481084301459637,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.008219392314566577,
      "loss": 0.7991,
      "num_input_tokens_seen": 2139776,
      "step": 3680
    },
    {
      "epoch": 0.5488531426869229,
      "grad_norm": 0.1328125,
      "learning_rate": 0.008230563002680965,
      "loss": 0.7935,
      "num_input_tokens_seen": 2142528,
      "step": 3685
    },
    {
      "epoch": 0.5495978552278821,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.008241733690795353,
      "loss": 0.8062,
      "num_input_tokens_seen": 2145472,
      "step": 3690
    },
    {
      "epoch": 0.5503425677688413,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.00825290437890974,
      "loss": 0.7994,
      "num_input_tokens_seen": 2148384,
      "step": 3695
    },
    {
      "epoch": 0.5510872803098004,
      "grad_norm": 0.064453125,
      "learning_rate": 0.008264075067024129,
      "loss": 0.8154,
      "num_input_tokens_seen": 2151136,
      "step": 3700
    },
    {
      "epoch": 0.5518319928507596,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.008275245755138516,
      "loss": 0.8141,
      "num_input_tokens_seen": 2153952,
      "step": 3705
    },
    {
      "epoch": 0.5525767053917188,
      "grad_norm": 0.05859375,
      "learning_rate": 0.008286416443252903,
      "loss": 0.7987,
      "num_input_tokens_seen": 2156736,
      "step": 3710
    },
    {
      "epoch": 0.553321417932678,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00829758713136729,
      "loss": 0.7933,
      "num_input_tokens_seen": 2159776,
      "step": 3715
    },
    {
      "epoch": 0.5540661304736372,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.008308757819481679,
      "loss": 0.8128,
      "num_input_tokens_seen": 2162656,
      "step": 3720
    },
    {
      "epoch": 0.5548108430145964,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.008319928507596067,
      "loss": 0.8011,
      "num_input_tokens_seen": 2165792,
      "step": 3725
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.008331099195710456,
      "loss": 0.7935,
      "num_input_tokens_seen": 2168704,
      "step": 3730
    },
    {
      "epoch": 0.5563002680965148,
      "grad_norm": 0.056640625,
      "learning_rate": 0.008342269883824844,
      "loss": 0.8027,
      "num_input_tokens_seen": 2171808,
      "step": 3735
    },
    {
      "epoch": 0.557044980637474,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.008353440571939232,
      "loss": 0.7986,
      "num_input_tokens_seen": 2174816,
      "step": 3740
    },
    {
      "epoch": 0.5577896931784331,
      "grad_norm": 0.06640625,
      "learning_rate": 0.00836461126005362,
      "loss": 0.775,
      "num_input_tokens_seen": 2177536,
      "step": 3745
    },
    {
      "epoch": 0.5585344057193923,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.008375781948168008,
      "loss": 0.8222,
      "num_input_tokens_seen": 2180416,
      "step": 3750
    },
    {
      "epoch": 0.5592791182603515,
      "grad_norm": 0.037109375,
      "learning_rate": 0.008386952636282394,
      "loss": 0.8019,
      "num_input_tokens_seen": 2183520,
      "step": 3755
    },
    {
      "epoch": 0.5600238308013107,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.008398123324396782,
      "loss": 0.7982,
      "num_input_tokens_seen": 2186016,
      "step": 3760
    },
    {
      "epoch": 0.5607685433422699,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00840929401251117,
      "loss": 0.8087,
      "num_input_tokens_seen": 2189376,
      "step": 3765
    },
    {
      "epoch": 0.5615132558832291,
      "grad_norm": 0.083984375,
      "learning_rate": 0.008420464700625558,
      "loss": 0.7916,
      "num_input_tokens_seen": 2192160,
      "step": 3770
    },
    {
      "epoch": 0.5622579684241883,
      "grad_norm": 0.05078125,
      "learning_rate": 0.008431635388739946,
      "loss": 0.846,
      "num_input_tokens_seen": 2195200,
      "step": 3775
    },
    {
      "epoch": 0.5630026809651475,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.008442806076854334,
      "loss": 0.7751,
      "num_input_tokens_seen": 2198048,
      "step": 3780
    },
    {
      "epoch": 0.5637473935061067,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.008453976764968722,
      "loss": 0.7933,
      "num_input_tokens_seen": 2201024,
      "step": 3785
    },
    {
      "epoch": 0.5644921060470658,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.00846514745308311,
      "loss": 0.8075,
      "num_input_tokens_seen": 2203712,
      "step": 3790
    },
    {
      "epoch": 0.565236818588025,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.008476318141197496,
      "loss": 0.7996,
      "num_input_tokens_seen": 2206464,
      "step": 3795
    },
    {
      "epoch": 0.5659815311289842,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.008487488829311884,
      "loss": 0.794,
      "num_input_tokens_seen": 2209504,
      "step": 3800
    },
    {
      "epoch": 0.5667262436699434,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.008498659517426272,
      "loss": 0.8085,
      "num_input_tokens_seen": 2213088,
      "step": 3805
    },
    {
      "epoch": 0.5674709562109026,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.00850983020554066,
      "loss": 0.8204,
      "num_input_tokens_seen": 2215904,
      "step": 3810
    },
    {
      "epoch": 0.5682156687518618,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.008521000893655048,
      "loss": 0.7623,
      "num_input_tokens_seen": 2218848,
      "step": 3815
    },
    {
      "epoch": 0.568960381292821,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.008532171581769436,
      "loss": 0.8212,
      "num_input_tokens_seen": 2221920,
      "step": 3820
    },
    {
      "epoch": 0.5697050938337802,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.008543342269883824,
      "loss": 0.8109,
      "num_input_tokens_seen": 2224992,
      "step": 3825
    },
    {
      "epoch": 0.5704498063747393,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.008554512957998212,
      "loss": 0.8075,
      "num_input_tokens_seen": 2227872,
      "step": 3830
    },
    {
      "epoch": 0.5711945189156985,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0085656836461126,
      "loss": 0.7684,
      "num_input_tokens_seen": 2230272,
      "step": 3835
    },
    {
      "epoch": 0.5719392314566577,
      "grad_norm": 0.0625,
      "learning_rate": 0.008576854334226986,
      "loss": 0.7674,
      "num_input_tokens_seen": 2233216,
      "step": 3840
    },
    {
      "epoch": 0.5726839439976169,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.008588025022341376,
      "loss": 0.8482,
      "num_input_tokens_seen": 2236192,
      "step": 3845
    },
    {
      "epoch": 0.5734286565385761,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.008599195710455764,
      "loss": 0.8017,
      "num_input_tokens_seen": 2239296,
      "step": 3850
    },
    {
      "epoch": 0.5741733690795353,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.008610366398570152,
      "loss": 0.8253,
      "num_input_tokens_seen": 2242272,
      "step": 3855
    },
    {
      "epoch": 0.5749180816204945,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.00862153708668454,
      "loss": 0.7887,
      "num_input_tokens_seen": 2245024,
      "step": 3860
    },
    {
      "epoch": 0.5756627941614537,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.008632707774798928,
      "loss": 0.801,
      "num_input_tokens_seen": 2247744,
      "step": 3865
    },
    {
      "epoch": 0.5764075067024129,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.008643878462913316,
      "loss": 0.7703,
      "num_input_tokens_seen": 2250624,
      "step": 3870
    },
    {
      "epoch": 0.577152219243372,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.008655049151027704,
      "loss": 0.8087,
      "num_input_tokens_seen": 2253440,
      "step": 3875
    },
    {
      "epoch": 0.5778969317843312,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.008666219839142092,
      "loss": 0.8111,
      "num_input_tokens_seen": 2256288,
      "step": 3880
    },
    {
      "epoch": 0.5786416443252904,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.008677390527256478,
      "loss": 0.8094,
      "num_input_tokens_seen": 2259232,
      "step": 3885
    },
    {
      "epoch": 0.5793863568662496,
      "grad_norm": 0.03515625,
      "learning_rate": 0.008688561215370866,
      "loss": 0.8219,
      "num_input_tokens_seen": 2262400,
      "step": 3890
    },
    {
      "epoch": 0.5801310694072088,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.008699731903485254,
      "loss": 0.7938,
      "num_input_tokens_seen": 2265056,
      "step": 3895
    },
    {
      "epoch": 0.580875781948168,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.008710902591599642,
      "loss": 0.7859,
      "num_input_tokens_seen": 2268096,
      "step": 3900
    },
    {
      "epoch": 0.5816204944891272,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.00872207327971403,
      "loss": 0.782,
      "num_input_tokens_seen": 2271136,
      "step": 3905
    },
    {
      "epoch": 0.5823652070300864,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.008733243967828418,
      "loss": 0.8139,
      "num_input_tokens_seen": 2274048,
      "step": 3910
    },
    {
      "epoch": 0.5831099195710456,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.008744414655942806,
      "loss": 0.7835,
      "num_input_tokens_seen": 2276736,
      "step": 3915
    },
    {
      "epoch": 0.5838546321120047,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.008755585344057194,
      "loss": 0.8034,
      "num_input_tokens_seen": 2279584,
      "step": 3920
    },
    {
      "epoch": 0.5845993446529639,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.00876675603217158,
      "loss": 0.8201,
      "num_input_tokens_seen": 2282592,
      "step": 3925
    },
    {
      "epoch": 0.5853440571939231,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.008777926720285968,
      "loss": 0.7893,
      "num_input_tokens_seen": 2285824,
      "step": 3930
    },
    {
      "epoch": 0.5860887697348823,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.008789097408400356,
      "loss": 0.8064,
      "num_input_tokens_seen": 2288672,
      "step": 3935
    },
    {
      "epoch": 0.5868334822758415,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.008800268096514744,
      "loss": 0.7908,
      "num_input_tokens_seen": 2291520,
      "step": 3940
    },
    {
      "epoch": 0.5875781948168007,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.008811438784629132,
      "loss": 0.7738,
      "num_input_tokens_seen": 2294048,
      "step": 3945
    },
    {
      "epoch": 0.5883229073577599,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00882260947274352,
      "loss": 0.7884,
      "num_input_tokens_seen": 2296832,
      "step": 3950
    },
    {
      "epoch": 0.5890676198987191,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.008833780160857908,
      "loss": 0.8101,
      "num_input_tokens_seen": 2299808,
      "step": 3955
    },
    {
      "epoch": 0.5898123324396782,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.008844950848972298,
      "loss": 0.8054,
      "num_input_tokens_seen": 2302592,
      "step": 3960
    },
    {
      "epoch": 0.5905570449806374,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.008856121537086686,
      "loss": 0.7896,
      "num_input_tokens_seen": 2305344,
      "step": 3965
    },
    {
      "epoch": 0.5913017575215966,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.008867292225201072,
      "loss": 0.7936,
      "num_input_tokens_seen": 2308096,
      "step": 3970
    },
    {
      "epoch": 0.5920464700625558,
      "grad_norm": 0.052734375,
      "learning_rate": 0.00887846291331546,
      "loss": 0.7878,
      "num_input_tokens_seen": 2311104,
      "step": 3975
    },
    {
      "epoch": 0.592791182603515,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.008889633601429848,
      "loss": 0.7957,
      "num_input_tokens_seen": 2314080,
      "step": 3980
    },
    {
      "epoch": 0.5935358951444742,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.008900804289544236,
      "loss": 0.7883,
      "num_input_tokens_seen": 2317120,
      "step": 3985
    },
    {
      "epoch": 0.5942806076854334,
      "grad_norm": 0.0703125,
      "learning_rate": 0.008911974977658624,
      "loss": 0.8281,
      "num_input_tokens_seen": 2319968,
      "step": 3990
    },
    {
      "epoch": 0.5950253202263927,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.008923145665773012,
      "loss": 0.795,
      "num_input_tokens_seen": 2322944,
      "step": 3995
    },
    {
      "epoch": 0.5957700327673519,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0089343163538874,
      "loss": 0.7703,
      "num_input_tokens_seen": 2325920,
      "step": 4000
    },
    {
      "epoch": 0.596514745308311,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.008945487042001788,
      "loss": 0.7966,
      "num_input_tokens_seen": 2328768,
      "step": 4005
    },
    {
      "epoch": 0.5972594578492701,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.008956657730116176,
      "loss": 0.8396,
      "num_input_tokens_seen": 2331616,
      "step": 4010
    },
    {
      "epoch": 0.5980041703902294,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.008967828418230562,
      "loss": 0.8076,
      "num_input_tokens_seen": 2334592,
      "step": 4015
    },
    {
      "epoch": 0.5987488829311886,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.00897899910634495,
      "loss": 0.802,
      "num_input_tokens_seen": 2337408,
      "step": 4020
    },
    {
      "epoch": 0.5994935954721478,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.008990169794459338,
      "loss": 0.798,
      "num_input_tokens_seen": 2340320,
      "step": 4025
    },
    {
      "epoch": 0.600238308013107,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.009001340482573726,
      "loss": 0.7859,
      "num_input_tokens_seen": 2343136,
      "step": 4030
    },
    {
      "epoch": 0.6009830205540662,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.009012511170688114,
      "loss": 0.7966,
      "num_input_tokens_seen": 2346208,
      "step": 4035
    },
    {
      "epoch": 0.6017277330950254,
      "grad_norm": 0.03515625,
      "learning_rate": 0.009023681858802502,
      "loss": 0.7902,
      "num_input_tokens_seen": 2348864,
      "step": 4040
    },
    {
      "epoch": 0.6024724456359845,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.00903485254691689,
      "loss": 0.7895,
      "num_input_tokens_seen": 2351520,
      "step": 4045
    },
    {
      "epoch": 0.6032171581769437,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.009046023235031278,
      "loss": 0.7987,
      "num_input_tokens_seen": 2354592,
      "step": 4050
    },
    {
      "epoch": 0.6039618707179029,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.009057193923145664,
      "loss": 0.7909,
      "num_input_tokens_seen": 2357376,
      "step": 4055
    },
    {
      "epoch": 0.6047065832588621,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.009068364611260052,
      "loss": 0.7862,
      "num_input_tokens_seen": 2360128,
      "step": 4060
    },
    {
      "epoch": 0.6054512957998213,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.00907953529937444,
      "loss": 0.78,
      "num_input_tokens_seen": 2363136,
      "step": 4065
    },
    {
      "epoch": 0.6061960083407805,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00909070598748883,
      "loss": 0.7879,
      "num_input_tokens_seen": 2365984,
      "step": 4070
    },
    {
      "epoch": 0.6069407208817397,
      "grad_norm": 0.0546875,
      "learning_rate": 0.009101876675603218,
      "loss": 0.7751,
      "num_input_tokens_seen": 2368640,
      "step": 4075
    },
    {
      "epoch": 0.6076854334226989,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.009113047363717606,
      "loss": 0.7679,
      "num_input_tokens_seen": 2371424,
      "step": 4080
    },
    {
      "epoch": 0.6084301459636581,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.009124218051831993,
      "loss": 0.8093,
      "num_input_tokens_seen": 2374240,
      "step": 4085
    },
    {
      "epoch": 0.6091748585046172,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.009135388739946381,
      "loss": 0.7717,
      "num_input_tokens_seen": 2377376,
      "step": 4090
    },
    {
      "epoch": 0.6099195710455764,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.00914655942806077,
      "loss": 0.8076,
      "num_input_tokens_seen": 2380128,
      "step": 4095
    },
    {
      "epoch": 0.6106642835865356,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.009157730116175156,
      "loss": 0.7631,
      "num_input_tokens_seen": 2382944,
      "step": 4100
    },
    {
      "epoch": 0.6114089961274948,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.009168900804289544,
      "loss": 0.7851,
      "num_input_tokens_seen": 2385824,
      "step": 4105
    },
    {
      "epoch": 0.612153708668454,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.009180071492403932,
      "loss": 0.9143,
      "num_input_tokens_seen": 2388736,
      "step": 4110
    },
    {
      "epoch": 0.6128984212094132,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.00919124218051832,
      "loss": 0.8635,
      "num_input_tokens_seen": 2391584,
      "step": 4115
    },
    {
      "epoch": 0.6136431337503724,
      "grad_norm": 0.046875,
      "learning_rate": 0.009202412868632708,
      "loss": 0.7921,
      "num_input_tokens_seen": 2394368,
      "step": 4120
    },
    {
      "epoch": 0.6143878462913316,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.009213583556747095,
      "loss": 0.8151,
      "num_input_tokens_seen": 2397568,
      "step": 4125
    },
    {
      "epoch": 0.6151325588322908,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.009224754244861483,
      "loss": 0.8137,
      "num_input_tokens_seen": 2400416,
      "step": 4130
    },
    {
      "epoch": 0.6158772713732499,
      "grad_norm": 0.044921875,
      "learning_rate": 0.009235924932975871,
      "loss": 0.8102,
      "num_input_tokens_seen": 2403360,
      "step": 4135
    },
    {
      "epoch": 0.6166219839142091,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00924709562109026,
      "loss": 0.7946,
      "num_input_tokens_seen": 2406112,
      "step": 4140
    },
    {
      "epoch": 0.6173666964551683,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.009258266309204646,
      "loss": 0.8033,
      "num_input_tokens_seen": 2409120,
      "step": 4145
    },
    {
      "epoch": 0.6181114089961275,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.009269436997319034,
      "loss": 0.7955,
      "num_input_tokens_seen": 2412000,
      "step": 4150
    },
    {
      "epoch": 0.6188561215370867,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.009280607685433422,
      "loss": 0.7905,
      "num_input_tokens_seen": 2414624,
      "step": 4155
    },
    {
      "epoch": 0.6196008340780459,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.00929177837354781,
      "loss": 0.7798,
      "num_input_tokens_seen": 2417280,
      "step": 4160
    },
    {
      "epoch": 0.6203455466190051,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.009302949061662198,
      "loss": 0.7811,
      "num_input_tokens_seen": 2420128,
      "step": 4165
    },
    {
      "epoch": 0.6210902591599643,
      "grad_norm": 0.0625,
      "learning_rate": 0.009314119749776585,
      "loss": 0.7986,
      "num_input_tokens_seen": 2423136,
      "step": 4170
    },
    {
      "epoch": 0.6218349717009234,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.009325290437890973,
      "loss": 0.8025,
      "num_input_tokens_seen": 2426176,
      "step": 4175
    },
    {
      "epoch": 0.6225796842418826,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.009336461126005361,
      "loss": 0.8118,
      "num_input_tokens_seen": 2429216,
      "step": 4180
    },
    {
      "epoch": 0.6233243967828418,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.00934763181411975,
      "loss": 0.8074,
      "num_input_tokens_seen": 2432416,
      "step": 4185
    },
    {
      "epoch": 0.624069109323801,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.009358802502234137,
      "loss": 0.7722,
      "num_input_tokens_seen": 2435264,
      "step": 4190
    },
    {
      "epoch": 0.6248138218647602,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.009369973190348525,
      "loss": 0.8144,
      "num_input_tokens_seen": 2438336,
      "step": 4195
    },
    {
      "epoch": 0.6255585344057194,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.009381143878462913,
      "loss": 0.8083,
      "num_input_tokens_seen": 2441376,
      "step": 4200
    },
    {
      "epoch": 0.6263032469466786,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.009392314566577301,
      "loss": 0.7892,
      "num_input_tokens_seen": 2444544,
      "step": 4205
    },
    {
      "epoch": 0.6270479594876378,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.00940348525469169,
      "loss": 0.804,
      "num_input_tokens_seen": 2447328,
      "step": 4210
    },
    {
      "epoch": 0.627792672028597,
      "grad_norm": 0.103515625,
      "learning_rate": 0.009414655942806077,
      "loss": 0.8164,
      "num_input_tokens_seen": 2450464,
      "step": 4215
    },
    {
      "epoch": 0.6285373845695561,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.009425826630920465,
      "loss": 0.8972,
      "num_input_tokens_seen": 2453280,
      "step": 4220
    },
    {
      "epoch": 0.6292820971105153,
      "grad_norm": 0.11328125,
      "learning_rate": 0.009436997319034853,
      "loss": 0.8273,
      "num_input_tokens_seen": 2456256,
      "step": 4225
    },
    {
      "epoch": 0.6300268096514745,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.00944816800714924,
      "loss": 0.793,
      "num_input_tokens_seen": 2459136,
      "step": 4230
    },
    {
      "epoch": 0.6307715221924337,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.009459338695263627,
      "loss": 0.7977,
      "num_input_tokens_seen": 2461792,
      "step": 4235
    },
    {
      "epoch": 0.6315162347333929,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.009470509383378015,
      "loss": 0.8183,
      "num_input_tokens_seen": 2464768,
      "step": 4240
    },
    {
      "epoch": 0.6322609472743521,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.009481680071492403,
      "loss": 0.7966,
      "num_input_tokens_seen": 2467680,
      "step": 4245
    },
    {
      "epoch": 0.6330056598153113,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.009492850759606791,
      "loss": 0.7944,
      "num_input_tokens_seen": 2470464,
      "step": 4250
    },
    {
      "epoch": 0.6337503723562705,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00950402144772118,
      "loss": 0.7828,
      "num_input_tokens_seen": 2473376,
      "step": 4255
    },
    {
      "epoch": 0.6344950848972297,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.009515192135835567,
      "loss": 0.7957,
      "num_input_tokens_seen": 2476640,
      "step": 4260
    },
    {
      "epoch": 0.6352397974381888,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.009526362823949955,
      "loss": 0.7918,
      "num_input_tokens_seen": 2479968,
      "step": 4265
    },
    {
      "epoch": 0.635984509979148,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.009537533512064343,
      "loss": 0.7856,
      "num_input_tokens_seen": 2482816,
      "step": 4270
    },
    {
      "epoch": 0.6367292225201072,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.00954870420017873,
      "loss": 0.8232,
      "num_input_tokens_seen": 2485568,
      "step": 4275
    },
    {
      "epoch": 0.6374739350610664,
      "grad_norm": 0.142578125,
      "learning_rate": 0.009559874888293117,
      "loss": 0.7984,
      "num_input_tokens_seen": 2488480,
      "step": 4280
    },
    {
      "epoch": 0.6382186476020256,
      "grad_norm": 0.13671875,
      "learning_rate": 0.009571045576407505,
      "loss": 0.8232,
      "num_input_tokens_seen": 2491360,
      "step": 4285
    },
    {
      "epoch": 0.6389633601429848,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.009582216264521893,
      "loss": 0.8204,
      "num_input_tokens_seen": 2494272,
      "step": 4290
    },
    {
      "epoch": 0.639708072683944,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.009593386952636283,
      "loss": 0.798,
      "num_input_tokens_seen": 2496928,
      "step": 4295
    },
    {
      "epoch": 0.6404527852249032,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.009604557640750671,
      "loss": 0.8097,
      "num_input_tokens_seen": 2499680,
      "step": 4300
    },
    {
      "epoch": 0.6411974977658623,
      "grad_norm": 0.10546875,
      "learning_rate": 0.009615728328865059,
      "loss": 0.805,
      "num_input_tokens_seen": 2502400,
      "step": 4305
    },
    {
      "epoch": 0.6419422103068215,
      "grad_norm": 0.251953125,
      "learning_rate": 0.009626899016979447,
      "loss": 0.8043,
      "num_input_tokens_seen": 2505184,
      "step": 4310
    },
    {
      "epoch": 0.6426869228477807,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.009638069705093833,
      "loss": 0.8273,
      "num_input_tokens_seen": 2508160,
      "step": 4315
    },
    {
      "epoch": 0.6434316353887399,
      "grad_norm": 0.09375,
      "learning_rate": 0.009649240393208221,
      "loss": 0.7963,
      "num_input_tokens_seen": 2510944,
      "step": 4320
    },
    {
      "epoch": 0.6441763479296991,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.009660411081322609,
      "loss": 0.9333,
      "num_input_tokens_seen": 2513856,
      "step": 4325
    },
    {
      "epoch": 0.6449210604706583,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.009671581769436997,
      "loss": 0.7956,
      "num_input_tokens_seen": 2516512,
      "step": 4330
    },
    {
      "epoch": 0.6456657730116175,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.009682752457551385,
      "loss": 0.8271,
      "num_input_tokens_seen": 2519200,
      "step": 4335
    },
    {
      "epoch": 0.6464104855525767,
      "grad_norm": 0.06640625,
      "learning_rate": 0.009693923145665773,
      "loss": 0.8016,
      "num_input_tokens_seen": 2522016,
      "step": 4340
    },
    {
      "epoch": 0.6471551980935359,
      "grad_norm": 0.0703125,
      "learning_rate": 0.009705093833780161,
      "loss": 0.8055,
      "num_input_tokens_seen": 2524768,
      "step": 4345
    },
    {
      "epoch": 0.647899910634495,
      "grad_norm": 0.142578125,
      "learning_rate": 0.009716264521894549,
      "loss": 0.7904,
      "num_input_tokens_seen": 2527328,
      "step": 4350
    },
    {
      "epoch": 0.6486446231754542,
      "grad_norm": 0.140625,
      "learning_rate": 0.009727435210008937,
      "loss": 0.7911,
      "num_input_tokens_seen": 2530304,
      "step": 4355
    },
    {
      "epoch": 0.6493893357164134,
      "grad_norm": 0.16796875,
      "learning_rate": 0.009738605898123323,
      "loss": 0.8286,
      "num_input_tokens_seen": 2533536,
      "step": 4360
    },
    {
      "epoch": 0.6501340482573726,
      "grad_norm": 0.158203125,
      "learning_rate": 0.009749776586237711,
      "loss": 0.7785,
      "num_input_tokens_seen": 2536544,
      "step": 4365
    },
    {
      "epoch": 0.6508787607983318,
      "grad_norm": 0.1640625,
      "learning_rate": 0.009760947274352099,
      "loss": 0.8489,
      "num_input_tokens_seen": 2539456,
      "step": 4370
    },
    {
      "epoch": 0.651623473339291,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.009772117962466487,
      "loss": 0.815,
      "num_input_tokens_seen": 2542528,
      "step": 4375
    },
    {
      "epoch": 0.6523681858802503,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.009783288650580875,
      "loss": 0.8119,
      "num_input_tokens_seen": 2545664,
      "step": 4380
    },
    {
      "epoch": 0.6531128984212095,
      "grad_norm": 0.162109375,
      "learning_rate": 0.009794459338695263,
      "loss": 0.7876,
      "num_input_tokens_seen": 2548544,
      "step": 4385
    },
    {
      "epoch": 0.6538576109621687,
      "grad_norm": 0.064453125,
      "learning_rate": 0.009805630026809651,
      "loss": 0.8323,
      "num_input_tokens_seen": 2551392,
      "step": 4390
    },
    {
      "epoch": 0.6546023235031277,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.009816800714924039,
      "loss": 0.7707,
      "num_input_tokens_seen": 2554336,
      "step": 4395
    },
    {
      "epoch": 0.655347036044087,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.009827971403038427,
      "loss": 0.7909,
      "num_input_tokens_seen": 2557056,
      "step": 4400
    },
    {
      "epoch": 0.6560917485850462,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.009839142091152815,
      "loss": 0.8402,
      "num_input_tokens_seen": 2559904,
      "step": 4405
    },
    {
      "epoch": 0.6568364611260054,
      "grad_norm": 0.07421875,
      "learning_rate": 0.009850312779267203,
      "loss": 0.8202,
      "num_input_tokens_seen": 2562752,
      "step": 4410
    },
    {
      "epoch": 0.6575811736669646,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.00986148346738159,
      "loss": 0.7745,
      "num_input_tokens_seen": 2565536,
      "step": 4415
    },
    {
      "epoch": 0.6583258862079238,
      "grad_norm": 0.171875,
      "learning_rate": 0.009872654155495979,
      "loss": 0.7915,
      "num_input_tokens_seen": 2568512,
      "step": 4420
    },
    {
      "epoch": 0.659070598748883,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.009883824843610367,
      "loss": 0.7708,
      "num_input_tokens_seen": 2571392,
      "step": 4425
    },
    {
      "epoch": 0.6598153112898422,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.009894995531724755,
      "loss": 0.7957,
      "num_input_tokens_seen": 2574368,
      "step": 4430
    },
    {
      "epoch": 0.6605600238308013,
      "grad_norm": 0.115234375,
      "learning_rate": 0.009906166219839143,
      "loss": 0.8209,
      "num_input_tokens_seen": 2577440,
      "step": 4435
    },
    {
      "epoch": 0.6613047363717605,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00991733690795353,
      "loss": 0.8521,
      "num_input_tokens_seen": 2580608,
      "step": 4440
    },
    {
      "epoch": 0.6620494489127197,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.009928507596067917,
      "loss": 0.8234,
      "num_input_tokens_seen": 2583232,
      "step": 4445
    },
    {
      "epoch": 0.6627941614536789,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.009939678284182305,
      "loss": 0.7927,
      "num_input_tokens_seen": 2586208,
      "step": 4450
    },
    {
      "epoch": 0.6635388739946381,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.009950848972296693,
      "loss": 0.8258,
      "num_input_tokens_seen": 2589120,
      "step": 4455
    },
    {
      "epoch": 0.6642835865355973,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.00996201966041108,
      "loss": 0.828,
      "num_input_tokens_seen": 2591872,
      "step": 4460
    },
    {
      "epoch": 0.6650282990765565,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.009973190348525469,
      "loss": 0.8193,
      "num_input_tokens_seen": 2594944,
      "step": 4465
    },
    {
      "epoch": 0.6657730116175157,
      "grad_norm": 0.095703125,
      "learning_rate": 0.009984361036639857,
      "loss": 0.8049,
      "num_input_tokens_seen": 2597760,
      "step": 4470
    },
    {
      "epoch": 0.6665177241584749,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.009995531724754245,
      "loss": 0.8179,
      "num_input_tokens_seen": 2600672,
      "step": 4475
    },
    {
      "epoch": 0.667262436699434,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.010006702412868633,
      "loss": 0.7972,
      "num_input_tokens_seen": 2603456,
      "step": 4480
    },
    {
      "epoch": 0.6680071492403932,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.01001787310098302,
      "loss": 0.8243,
      "num_input_tokens_seen": 2606496,
      "step": 4485
    },
    {
      "epoch": 0.6687518617813524,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.010029043789097407,
      "loss": 0.8134,
      "num_input_tokens_seen": 2609216,
      "step": 4490
    },
    {
      "epoch": 0.6694965743223116,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.010040214477211795,
      "loss": 0.8137,
      "num_input_tokens_seen": 2611936,
      "step": 4495
    },
    {
      "epoch": 0.6702412868632708,
      "grad_norm": 0.083984375,
      "learning_rate": 0.010051385165326183,
      "loss": 0.7941,
      "num_input_tokens_seen": 2614848,
      "step": 4500
    },
    {
      "epoch": 0.67098599940423,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.01006255585344057,
      "loss": 0.8299,
      "num_input_tokens_seen": 2617856,
      "step": 4505
    },
    {
      "epoch": 0.6717307119451892,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.010073726541554959,
      "loss": 0.8129,
      "num_input_tokens_seen": 2620960,
      "step": 4510
    },
    {
      "epoch": 0.6724754244861484,
      "grad_norm": 0.044921875,
      "learning_rate": 0.010084897229669347,
      "loss": 0.81,
      "num_input_tokens_seen": 2623936,
      "step": 4515
    },
    {
      "epoch": 0.6732201370271076,
      "grad_norm": 0.0546875,
      "learning_rate": 0.010096067917783736,
      "loss": 0.8098,
      "num_input_tokens_seen": 2626816,
      "step": 4520
    },
    {
      "epoch": 0.6739648495680667,
      "grad_norm": 0.072265625,
      "learning_rate": 0.010107238605898124,
      "loss": 0.8034,
      "num_input_tokens_seen": 2629504,
      "step": 4525
    },
    {
      "epoch": 0.6747095621090259,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.010118409294012512,
      "loss": 0.7997,
      "num_input_tokens_seen": 2632032,
      "step": 4530
    },
    {
      "epoch": 0.6754542746499851,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.010129579982126899,
      "loss": 0.8466,
      "num_input_tokens_seen": 2635104,
      "step": 4535
    },
    {
      "epoch": 0.6761989871909443,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.010140750670241287,
      "loss": 0.8073,
      "num_input_tokens_seen": 2637888,
      "step": 4540
    },
    {
      "epoch": 0.6769436997319035,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.010151921358355675,
      "loss": 0.8084,
      "num_input_tokens_seen": 2640608,
      "step": 4545
    },
    {
      "epoch": 0.6776884122728627,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.010163092046470062,
      "loss": 0.811,
      "num_input_tokens_seen": 2643616,
      "step": 4550
    },
    {
      "epoch": 0.6784331248138219,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.01017426273458445,
      "loss": 0.8017,
      "num_input_tokens_seen": 2646464,
      "step": 4555
    },
    {
      "epoch": 0.6791778373547811,
      "grad_norm": 0.0390625,
      "learning_rate": 0.010185433422698838,
      "loss": 0.8059,
      "num_input_tokens_seen": 2649280,
      "step": 4560
    },
    {
      "epoch": 0.6799225498957402,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.010196604110813226,
      "loss": 0.7891,
      "num_input_tokens_seen": 2652288,
      "step": 4565
    },
    {
      "epoch": 0.6806672624366994,
      "grad_norm": 0.033203125,
      "learning_rate": 0.010207774798927614,
      "loss": 0.7952,
      "num_input_tokens_seen": 2655520,
      "step": 4570
    },
    {
      "epoch": 0.6814119749776586,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.010218945487042,
      "loss": 0.8198,
      "num_input_tokens_seen": 2658336,
      "step": 4575
    },
    {
      "epoch": 0.6821566875186178,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.010230116175156389,
      "loss": 0.7843,
      "num_input_tokens_seen": 2661568,
      "step": 4580
    },
    {
      "epoch": 0.682901400059577,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.010241286863270777,
      "loss": 0.8044,
      "num_input_tokens_seen": 2664576,
      "step": 4585
    },
    {
      "epoch": 0.6836461126005362,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.010252457551385164,
      "loss": 0.8178,
      "num_input_tokens_seen": 2667584,
      "step": 4590
    },
    {
      "epoch": 0.6843908251414954,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.010263628239499552,
      "loss": 0.7923,
      "num_input_tokens_seen": 2670272,
      "step": 4595
    },
    {
      "epoch": 0.6851355376824546,
      "grad_norm": 0.0546875,
      "learning_rate": 0.01027479892761394,
      "loss": 0.7942,
      "num_input_tokens_seen": 2673312,
      "step": 4600
    },
    {
      "epoch": 0.6858802502234138,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.010285969615728328,
      "loss": 0.8352,
      "num_input_tokens_seen": 2676512,
      "step": 4605
    },
    {
      "epoch": 0.6866249627643729,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.010297140303842716,
      "loss": 0.8083,
      "num_input_tokens_seen": 2679488,
      "step": 4610
    },
    {
      "epoch": 0.6873696753053321,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.010308310991957104,
      "loss": 0.812,
      "num_input_tokens_seen": 2682208,
      "step": 4615
    },
    {
      "epoch": 0.6881143878462913,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.01031948168007149,
      "loss": 0.7994,
      "num_input_tokens_seen": 2684992,
      "step": 4620
    },
    {
      "epoch": 0.6888591003872505,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.010330652368185879,
      "loss": 0.803,
      "num_input_tokens_seen": 2687840,
      "step": 4625
    },
    {
      "epoch": 0.6896038129282097,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.010341823056300267,
      "loss": 0.8278,
      "num_input_tokens_seen": 2690592,
      "step": 4630
    },
    {
      "epoch": 0.6903485254691689,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.010352993744414656,
      "loss": 0.8109,
      "num_input_tokens_seen": 2693472,
      "step": 4635
    },
    {
      "epoch": 0.6910932380101281,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.010364164432529044,
      "loss": 0.7867,
      "num_input_tokens_seen": 2696256,
      "step": 4640
    },
    {
      "epoch": 0.6918379505510873,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.010375335120643432,
      "loss": 0.8002,
      "num_input_tokens_seen": 2699104,
      "step": 4645
    },
    {
      "epoch": 0.6925826630920465,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.01038650580875782,
      "loss": 0.7941,
      "num_input_tokens_seen": 2702240,
      "step": 4650
    },
    {
      "epoch": 0.6933273756330056,
      "grad_norm": 0.046875,
      "learning_rate": 0.010397676496872208,
      "loss": 0.7886,
      "num_input_tokens_seen": 2704864,
      "step": 4655
    },
    {
      "epoch": 0.6940720881739648,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.010408847184986596,
      "loss": 0.7907,
      "num_input_tokens_seen": 2707552,
      "step": 4660
    },
    {
      "epoch": 0.694816800714924,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.010420017873100982,
      "loss": 0.8159,
      "num_input_tokens_seen": 2710592,
      "step": 4665
    },
    {
      "epoch": 0.6955615132558832,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01043118856121537,
      "loss": 0.8271,
      "num_input_tokens_seen": 2713504,
      "step": 4670
    },
    {
      "epoch": 0.6963062257968424,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.010442359249329758,
      "loss": 0.8205,
      "num_input_tokens_seen": 2716448,
      "step": 4675
    },
    {
      "epoch": 0.6970509383378016,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.010453529937444146,
      "loss": 0.7938,
      "num_input_tokens_seen": 2719296,
      "step": 4680
    },
    {
      "epoch": 0.6977956508787608,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.010464700625558534,
      "loss": 0.814,
      "num_input_tokens_seen": 2722112,
      "step": 4685
    },
    {
      "epoch": 0.69854036341972,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.010475871313672922,
      "loss": 0.8062,
      "num_input_tokens_seen": 2725056,
      "step": 4690
    },
    {
      "epoch": 0.6992850759606791,
      "grad_norm": 0.0859375,
      "learning_rate": 0.01048704200178731,
      "loss": 0.7832,
      "num_input_tokens_seen": 2727936,
      "step": 4695
    },
    {
      "epoch": 0.7000297885016383,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.010498212689901698,
      "loss": 0.8186,
      "num_input_tokens_seen": 2730656,
      "step": 4700
    },
    {
      "epoch": 0.7007745010425975,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.010509383378016084,
      "loss": 0.8121,
      "num_input_tokens_seen": 2733888,
      "step": 4705
    },
    {
      "epoch": 0.7015192135835567,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.010520554066130472,
      "loss": 0.8147,
      "num_input_tokens_seen": 2736640,
      "step": 4710
    },
    {
      "epoch": 0.7022639261245159,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.01053172475424486,
      "loss": 0.7913,
      "num_input_tokens_seen": 2739360,
      "step": 4715
    },
    {
      "epoch": 0.7030086386654751,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.010542895442359248,
      "loss": 0.7983,
      "num_input_tokens_seen": 2742080,
      "step": 4720
    },
    {
      "epoch": 0.7037533512064343,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.010554066130473636,
      "loss": 0.7878,
      "num_input_tokens_seen": 2744736,
      "step": 4725
    },
    {
      "epoch": 0.7044980637473935,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.010565236818588024,
      "loss": 0.8143,
      "num_input_tokens_seen": 2747712,
      "step": 4730
    },
    {
      "epoch": 0.7052427762883527,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.010576407506702412,
      "loss": 0.8126,
      "num_input_tokens_seen": 2750688,
      "step": 4735
    },
    {
      "epoch": 0.7059874888293118,
      "grad_norm": 0.078125,
      "learning_rate": 0.0105875781948168,
      "loss": 0.788,
      "num_input_tokens_seen": 2753568,
      "step": 4740
    },
    {
      "epoch": 0.706732201370271,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.010598748882931188,
      "loss": 0.8227,
      "num_input_tokens_seen": 2756416,
      "step": 4745
    },
    {
      "epoch": 0.7074769139112302,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.010609919571045576,
      "loss": 0.7904,
      "num_input_tokens_seen": 2759488,
      "step": 4750
    },
    {
      "epoch": 0.7082216264521894,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.010621090259159964,
      "loss": 0.7927,
      "num_input_tokens_seen": 2762432,
      "step": 4755
    },
    {
      "epoch": 0.7089663389931486,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.010632260947274352,
      "loss": 0.8189,
      "num_input_tokens_seen": 2765504,
      "step": 4760
    },
    {
      "epoch": 0.7097110515341079,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.01064343163538874,
      "loss": 0.8001,
      "num_input_tokens_seen": 2768416,
      "step": 4765
    },
    {
      "epoch": 0.710455764075067,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.010654602323503128,
      "loss": 0.7854,
      "num_input_tokens_seen": 2771456,
      "step": 4770
    },
    {
      "epoch": 0.7112004766160263,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.010665773011617516,
      "loss": 0.7999,
      "num_input_tokens_seen": 2774720,
      "step": 4775
    },
    {
      "epoch": 0.7119451891569855,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.010676943699731904,
      "loss": 0.813,
      "num_input_tokens_seen": 2777600,
      "step": 4780
    },
    {
      "epoch": 0.7126899016979446,
      "grad_norm": 0.05078125,
      "learning_rate": 0.010688114387846292,
      "loss": 0.8085,
      "num_input_tokens_seen": 2780608,
      "step": 4785
    },
    {
      "epoch": 0.7134346142389038,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.010699285075960678,
      "loss": 0.7913,
      "num_input_tokens_seen": 2783328,
      "step": 4790
    },
    {
      "epoch": 0.714179326779863,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.010710455764075066,
      "loss": 0.7852,
      "num_input_tokens_seen": 2786336,
      "step": 4795
    },
    {
      "epoch": 0.7149240393208222,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.010721626452189454,
      "loss": 0.7872,
      "num_input_tokens_seen": 2789024,
      "step": 4800
    },
    {
      "epoch": 0.7156687518617814,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.010732797140303842,
      "loss": 0.7805,
      "num_input_tokens_seen": 2792064,
      "step": 4805
    },
    {
      "epoch": 0.7164134644027406,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.01074396782841823,
      "loss": 0.7826,
      "num_input_tokens_seen": 2794848,
      "step": 4810
    },
    {
      "epoch": 0.7171581769436998,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.010755138516532618,
      "loss": 0.7861,
      "num_input_tokens_seen": 2797728,
      "step": 4815
    },
    {
      "epoch": 0.717902889484659,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.010766309204647006,
      "loss": 0.8036,
      "num_input_tokens_seen": 2800832,
      "step": 4820
    },
    {
      "epoch": 0.7186476020256181,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.010777479892761394,
      "loss": 0.8138,
      "num_input_tokens_seen": 2803712,
      "step": 4825
    },
    {
      "epoch": 0.7193923145665773,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.010788650580875782,
      "loss": 0.7826,
      "num_input_tokens_seen": 2806528,
      "step": 4830
    },
    {
      "epoch": 0.7201370271075365,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.010799821268990168,
      "loss": 0.8322,
      "num_input_tokens_seen": 2809216,
      "step": 4835
    },
    {
      "epoch": 0.7208817396484957,
      "grad_norm": 0.037109375,
      "learning_rate": 0.010810991957104556,
      "loss": 0.8075,
      "num_input_tokens_seen": 2812224,
      "step": 4840
    },
    {
      "epoch": 0.7216264521894549,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.010822162645218944,
      "loss": 0.8076,
      "num_input_tokens_seen": 2815008,
      "step": 4845
    },
    {
      "epoch": 0.7223711647304141,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.010833333333333332,
      "loss": 0.8223,
      "num_input_tokens_seen": 2817632,
      "step": 4850
    },
    {
      "epoch": 0.7231158772713733,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.01084450402144772,
      "loss": 0.7809,
      "num_input_tokens_seen": 2820352,
      "step": 4855
    },
    {
      "epoch": 0.7238605898123325,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01085567470956211,
      "loss": 0.8073,
      "num_input_tokens_seen": 2823328,
      "step": 4860
    },
    {
      "epoch": 0.7246053023532917,
      "grad_norm": 0.046875,
      "learning_rate": 0.010866845397676498,
      "loss": 0.8025,
      "num_input_tokens_seen": 2826368,
      "step": 4865
    },
    {
      "epoch": 0.7253500148942508,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.010878016085790886,
      "loss": 0.8311,
      "num_input_tokens_seen": 2829504,
      "step": 4870
    },
    {
      "epoch": 0.72609472743521,
      "grad_norm": 0.064453125,
      "learning_rate": 0.010889186773905274,
      "loss": 0.7888,
      "num_input_tokens_seen": 2832384,
      "step": 4875
    },
    {
      "epoch": 0.7268394399761692,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01090035746201966,
      "loss": 0.7739,
      "num_input_tokens_seen": 2834976,
      "step": 4880
    },
    {
      "epoch": 0.7275841525171284,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.010911528150134048,
      "loss": 0.8229,
      "num_input_tokens_seen": 2837952,
      "step": 4885
    },
    {
      "epoch": 0.7283288650580876,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.010922698838248436,
      "loss": 0.7819,
      "num_input_tokens_seen": 2840736,
      "step": 4890
    },
    {
      "epoch": 0.7290735775990468,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.010933869526362824,
      "loss": 0.7933,
      "num_input_tokens_seen": 2843552,
      "step": 4895
    },
    {
      "epoch": 0.729818290140006,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.010945040214477212,
      "loss": 0.8024,
      "num_input_tokens_seen": 2846592,
      "step": 4900
    },
    {
      "epoch": 0.7305630026809652,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.0109562109025916,
      "loss": 0.8262,
      "num_input_tokens_seen": 2849376,
      "step": 4905
    },
    {
      "epoch": 0.7313077152219244,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.010967381590705988,
      "loss": 0.7848,
      "num_input_tokens_seen": 2852032,
      "step": 4910
    },
    {
      "epoch": 0.7320524277628835,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.010978552278820376,
      "loss": 0.7767,
      "num_input_tokens_seen": 2855360,
      "step": 4915
    },
    {
      "epoch": 0.7327971403038427,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.010989722966934762,
      "loss": 0.7981,
      "num_input_tokens_seen": 2858080,
      "step": 4920
    },
    {
      "epoch": 0.7335418528448019,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.01100089365504915,
      "loss": 0.7916,
      "num_input_tokens_seen": 2861024,
      "step": 4925
    },
    {
      "epoch": 0.7342865653857611,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.011012064343163538,
      "loss": 0.8535,
      "num_input_tokens_seen": 2864128,
      "step": 4930
    },
    {
      "epoch": 0.7350312779267203,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.011023235031277926,
      "loss": 0.8038,
      "num_input_tokens_seen": 2866912,
      "step": 4935
    },
    {
      "epoch": 0.7357759904676795,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.011034405719392314,
      "loss": 0.8031,
      "num_input_tokens_seen": 2869824,
      "step": 4940
    },
    {
      "epoch": 0.7365207030086387,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.011045576407506702,
      "loss": 0.8035,
      "num_input_tokens_seen": 2872576,
      "step": 4945
    },
    {
      "epoch": 0.7372654155495979,
      "grad_norm": 0.02734375,
      "learning_rate": 0.01105674709562109,
      "loss": 0.8126,
      "num_input_tokens_seen": 2875552,
      "step": 4950
    },
    {
      "epoch": 0.738010128090557,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.011067917783735478,
      "loss": 0.8082,
      "num_input_tokens_seen": 2878560,
      "step": 4955
    },
    {
      "epoch": 0.7387548406315162,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.011079088471849866,
      "loss": 0.7934,
      "num_input_tokens_seen": 2881344,
      "step": 4960
    },
    {
      "epoch": 0.7394995531724754,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.011090259159964252,
      "loss": 0.8208,
      "num_input_tokens_seen": 2884416,
      "step": 4965
    },
    {
      "epoch": 0.7402442657134346,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.011101429848078641,
      "loss": 0.7721,
      "num_input_tokens_seen": 2887264,
      "step": 4970
    },
    {
      "epoch": 0.7409889782543938,
      "grad_norm": 0.07421875,
      "learning_rate": 0.01111260053619303,
      "loss": 0.8015,
      "num_input_tokens_seen": 2890016,
      "step": 4975
    },
    {
      "epoch": 0.741733690795353,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.011123771224307417,
      "loss": 0.8046,
      "num_input_tokens_seen": 2892832,
      "step": 4980
    },
    {
      "epoch": 0.7424784033363122,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.011134941912421805,
      "loss": 0.8097,
      "num_input_tokens_seen": 2895904,
      "step": 4985
    },
    {
      "epoch": 0.7432231158772714,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.011146112600536193,
      "loss": 0.8006,
      "num_input_tokens_seen": 2898848,
      "step": 4990
    },
    {
      "epoch": 0.7439678284182306,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.011157283288650581,
      "loss": 0.79,
      "num_input_tokens_seen": 2901696,
      "step": 4995
    },
    {
      "epoch": 0.7447125409591897,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.01116845397676497,
      "loss": 0.8128,
      "num_input_tokens_seen": 2904224,
      "step": 5000
    },
    {
      "epoch": 0.7454572535001489,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.011179624664879357,
      "loss": 0.823,
      "num_input_tokens_seen": 2906848,
      "step": 5005
    },
    {
      "epoch": 0.7462019660411081,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.011190795352993744,
      "loss": 0.8122,
      "num_input_tokens_seen": 2909568,
      "step": 5010
    },
    {
      "epoch": 0.7469466785820673,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.011201966041108131,
      "loss": 0.8033,
      "num_input_tokens_seen": 2912352,
      "step": 5015
    },
    {
      "epoch": 0.7476913911230265,
      "grad_norm": 0.0390625,
      "learning_rate": 0.01121313672922252,
      "loss": 0.7918,
      "num_input_tokens_seen": 2914720,
      "step": 5020
    },
    {
      "epoch": 0.7484361036639857,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.011224307417336907,
      "loss": 0.8046,
      "num_input_tokens_seen": 2917536,
      "step": 5025
    },
    {
      "epoch": 0.7491808162049449,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.011235478105451295,
      "loss": 0.8188,
      "num_input_tokens_seen": 2920448,
      "step": 5030
    },
    {
      "epoch": 0.7499255287459041,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.011246648793565683,
      "loss": 0.8006,
      "num_input_tokens_seen": 2923712,
      "step": 5035
    },
    {
      "epoch": 0.7506702412868632,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.011257819481680071,
      "loss": 0.8031,
      "num_input_tokens_seen": 2926528,
      "step": 5040
    },
    {
      "epoch": 0.7514149538278224,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.01126899016979446,
      "loss": 0.7961,
      "num_input_tokens_seen": 2929280,
      "step": 5045
    },
    {
      "epoch": 0.7521596663687816,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.011280160857908846,
      "loss": 0.8013,
      "num_input_tokens_seen": 2932288,
      "step": 5050
    },
    {
      "epoch": 0.7529043789097408,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.011291331546023234,
      "loss": 0.8097,
      "num_input_tokens_seen": 2934976,
      "step": 5055
    },
    {
      "epoch": 0.7536490914507,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.011302502234137621,
      "loss": 0.8029,
      "num_input_tokens_seen": 2937856,
      "step": 5060
    },
    {
      "epoch": 0.7543938039916592,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.01131367292225201,
      "loss": 0.7936,
      "num_input_tokens_seen": 2940384,
      "step": 5065
    },
    {
      "epoch": 0.7551385165326184,
      "grad_norm": 0.037109375,
      "learning_rate": 0.011324843610366397,
      "loss": 0.8093,
      "num_input_tokens_seen": 2943232,
      "step": 5070
    },
    {
      "epoch": 0.7558832290735776,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.011336014298480785,
      "loss": 0.8153,
      "num_input_tokens_seen": 2946208,
      "step": 5075
    },
    {
      "epoch": 0.7566279416145368,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.011347184986595173,
      "loss": 0.7927,
      "num_input_tokens_seen": 2949280,
      "step": 5080
    },
    {
      "epoch": 0.7573726541554959,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.011358355674709563,
      "loss": 0.8057,
      "num_input_tokens_seen": 2951808,
      "step": 5085
    },
    {
      "epoch": 0.7581173666964551,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.011369526362823951,
      "loss": 0.8003,
      "num_input_tokens_seen": 2954560,
      "step": 5090
    },
    {
      "epoch": 0.7588620792374143,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.011380697050938337,
      "loss": 0.8073,
      "num_input_tokens_seen": 2957376,
      "step": 5095
    },
    {
      "epoch": 0.7596067917783735,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.011391867739052725,
      "loss": 0.7684,
      "num_input_tokens_seen": 2960352,
      "step": 5100
    },
    {
      "epoch": 0.7603515043193327,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.011403038427167113,
      "loss": 0.8232,
      "num_input_tokens_seen": 2963328,
      "step": 5105
    },
    {
      "epoch": 0.7610962168602919,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.011414209115281501,
      "loss": 0.7832,
      "num_input_tokens_seen": 2966208,
      "step": 5110
    },
    {
      "epoch": 0.7618409294012511,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.01142537980339589,
      "loss": 0.7985,
      "num_input_tokens_seen": 2969120,
      "step": 5115
    },
    {
      "epoch": 0.7625856419422103,
      "grad_norm": 0.06640625,
      "learning_rate": 0.011436550491510277,
      "loss": 0.8185,
      "num_input_tokens_seen": 2971776,
      "step": 5120
    },
    {
      "epoch": 0.7633303544831695,
      "grad_norm": 0.052734375,
      "learning_rate": 0.011447721179624665,
      "loss": 0.7989,
      "num_input_tokens_seen": 2974560,
      "step": 5125
    },
    {
      "epoch": 0.7640750670241286,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.011458891867739053,
      "loss": 0.8193,
      "num_input_tokens_seen": 2977376,
      "step": 5130
    },
    {
      "epoch": 0.7648197795650878,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.011470062555853441,
      "loss": 0.7911,
      "num_input_tokens_seen": 2980192,
      "step": 5135
    },
    {
      "epoch": 0.765564492106047,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.011481233243967827,
      "loss": 0.8178,
      "num_input_tokens_seen": 2982848,
      "step": 5140
    },
    {
      "epoch": 0.7663092046470062,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.011492403932082215,
      "loss": 0.8119,
      "num_input_tokens_seen": 2985344,
      "step": 5145
    },
    {
      "epoch": 0.7670539171879655,
      "grad_norm": 0.046875,
      "learning_rate": 0.011503574620196603,
      "loss": 0.8021,
      "num_input_tokens_seen": 2988128,
      "step": 5150
    },
    {
      "epoch": 0.7677986297289247,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.011514745308310991,
      "loss": 0.8009,
      "num_input_tokens_seen": 2991040,
      "step": 5155
    },
    {
      "epoch": 0.7685433422698839,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.011525915996425379,
      "loss": 0.7944,
      "num_input_tokens_seen": 2994112,
      "step": 5160
    },
    {
      "epoch": 0.7692880548108431,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.011537086684539767,
      "loss": 0.7979,
      "num_input_tokens_seen": 2996960,
      "step": 5165
    },
    {
      "epoch": 0.7700327673518021,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.011548257372654155,
      "loss": 0.7947,
      "num_input_tokens_seen": 2999680,
      "step": 5170
    },
    {
      "epoch": 0.7707774798927614,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.011559428060768543,
      "loss": 0.7796,
      "num_input_tokens_seen": 3002336,
      "step": 5175
    },
    {
      "epoch": 0.7715221924337206,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.01157059874888293,
      "loss": 0.8133,
      "num_input_tokens_seen": 3005216,
      "step": 5180
    },
    {
      "epoch": 0.7722669049746798,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.011581769436997317,
      "loss": 0.771,
      "num_input_tokens_seen": 3008032,
      "step": 5185
    },
    {
      "epoch": 0.773011617515639,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.011592940125111705,
      "loss": 0.7818,
      "num_input_tokens_seen": 3010848,
      "step": 5190
    },
    {
      "epoch": 0.7737563300565982,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.011604110813226095,
      "loss": 0.8213,
      "num_input_tokens_seen": 3013568,
      "step": 5195
    },
    {
      "epoch": 0.7745010425975574,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.011615281501340483,
      "loss": 0.8254,
      "num_input_tokens_seen": 3016512,
      "step": 5200
    },
    {
      "epoch": 0.7752457551385166,
      "grad_norm": 0.08203125,
      "learning_rate": 0.01162645218945487,
      "loss": 0.86,
      "num_input_tokens_seen": 3019424,
      "step": 5205
    },
    {
      "epoch": 0.7759904676794758,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.011637622877569259,
      "loss": 0.7721,
      "num_input_tokens_seen": 3022368,
      "step": 5210
    },
    {
      "epoch": 0.7767351802204349,
      "grad_norm": 0.044921875,
      "learning_rate": 0.011648793565683647,
      "loss": 0.7979,
      "num_input_tokens_seen": 3025120,
      "step": 5215
    },
    {
      "epoch": 0.7774798927613941,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.011659964253798035,
      "loss": 0.7989,
      "num_input_tokens_seen": 3028032,
      "step": 5220
    },
    {
      "epoch": 0.7782246053023533,
      "grad_norm": 0.041015625,
      "learning_rate": 0.011671134941912421,
      "loss": 0.7838,
      "num_input_tokens_seen": 3030624,
      "step": 5225
    },
    {
      "epoch": 0.7789693178433125,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.011682305630026809,
      "loss": 0.8213,
      "num_input_tokens_seen": 3033472,
      "step": 5230
    },
    {
      "epoch": 0.7797140303842717,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.011693476318141197,
      "loss": 0.7751,
      "num_input_tokens_seen": 3036384,
      "step": 5235
    },
    {
      "epoch": 0.7804587429252309,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.011704647006255585,
      "loss": 0.7908,
      "num_input_tokens_seen": 3039200,
      "step": 5240
    },
    {
      "epoch": 0.7812034554661901,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.011715817694369973,
      "loss": 0.8105,
      "num_input_tokens_seen": 3042208,
      "step": 5245
    },
    {
      "epoch": 0.7819481680071493,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.01172698838248436,
      "loss": 0.8146,
      "num_input_tokens_seen": 3044992,
      "step": 5250
    },
    {
      "epoch": 0.7826928805481085,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.011738159070598749,
      "loss": 0.7683,
      "num_input_tokens_seen": 3048288,
      "step": 5255
    },
    {
      "epoch": 0.7834375930890676,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.011749329758713137,
      "loss": 0.8038,
      "num_input_tokens_seen": 3051072,
      "step": 5260
    },
    {
      "epoch": 0.7841823056300268,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.011760500446827525,
      "loss": 0.804,
      "num_input_tokens_seen": 3054016,
      "step": 5265
    },
    {
      "epoch": 0.784927018170986,
      "grad_norm": 0.06640625,
      "learning_rate": 0.011771671134941911,
      "loss": 0.7972,
      "num_input_tokens_seen": 3057088,
      "step": 5270
    },
    {
      "epoch": 0.7856717307119452,
      "grad_norm": 0.046875,
      "learning_rate": 0.011782841823056299,
      "loss": 0.8222,
      "num_input_tokens_seen": 3060096,
      "step": 5275
    },
    {
      "epoch": 0.7864164432529044,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.011794012511170687,
      "loss": 0.8099,
      "num_input_tokens_seen": 3063040,
      "step": 5280
    },
    {
      "epoch": 0.7871611557938636,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.011805183199285075,
      "loss": 0.7993,
      "num_input_tokens_seen": 3065824,
      "step": 5285
    },
    {
      "epoch": 0.7879058683348228,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.011816353887399463,
      "loss": 0.8205,
      "num_input_tokens_seen": 3068672,
      "step": 5290
    },
    {
      "epoch": 0.788650580875782,
      "grad_norm": 0.0546875,
      "learning_rate": 0.01182752457551385,
      "loss": 0.8064,
      "num_input_tokens_seen": 3071616,
      "step": 5295
    },
    {
      "epoch": 0.7893952934167411,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.011838695263628239,
      "loss": 0.8012,
      "num_input_tokens_seen": 3074432,
      "step": 5300
    },
    {
      "epoch": 0.7901400059577003,
      "grad_norm": 0.359375,
      "learning_rate": 0.011849865951742627,
      "loss": 0.8249,
      "num_input_tokens_seen": 3077472,
      "step": 5305
    },
    {
      "epoch": 0.7908847184986595,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.011861036639857015,
      "loss": 0.7426,
      "num_input_tokens_seen": 3080544,
      "step": 5310
    },
    {
      "epoch": 0.7916294310396187,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.011872207327971403,
      "loss": 0.8176,
      "num_input_tokens_seen": 3083584,
      "step": 5315
    },
    {
      "epoch": 0.7923741435805779,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.01188337801608579,
      "loss": 0.8499,
      "num_input_tokens_seen": 3086208,
      "step": 5320
    },
    {
      "epoch": 0.7931188561215371,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.011894548704200179,
      "loss": 0.7876,
      "num_input_tokens_seen": 3089152,
      "step": 5325
    },
    {
      "epoch": 0.7938635686624963,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.011905719392314567,
      "loss": 0.804,
      "num_input_tokens_seen": 3091968,
      "step": 5330
    },
    {
      "epoch": 0.7946082812034555,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.011916890080428955,
      "loss": 0.8121,
      "num_input_tokens_seen": 3095072,
      "step": 5335
    },
    {
      "epoch": 0.7953529937444147,
      "grad_norm": 0.048828125,
      "learning_rate": 0.011928060768543343,
      "loss": 0.7974,
      "num_input_tokens_seen": 3097920,
      "step": 5340
    },
    {
      "epoch": 0.7960977062853738,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.01193923145665773,
      "loss": 0.81,
      "num_input_tokens_seen": 3101056,
      "step": 5345
    },
    {
      "epoch": 0.796842418826333,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.011950402144772118,
      "loss": 0.8107,
      "num_input_tokens_seen": 3103904,
      "step": 5350
    },
    {
      "epoch": 0.7975871313672922,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.011961572832886505,
      "loss": 0.81,
      "num_input_tokens_seen": 3106912,
      "step": 5355
    },
    {
      "epoch": 0.7983318439082514,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.011972743521000893,
      "loss": 0.8053,
      "num_input_tokens_seen": 3109408,
      "step": 5360
    },
    {
      "epoch": 0.7990765564492106,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01198391420911528,
      "loss": 0.8013,
      "num_input_tokens_seen": 3112352,
      "step": 5365
    },
    {
      "epoch": 0.7998212689901698,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.011995084897229669,
      "loss": 0.7944,
      "num_input_tokens_seen": 3115328,
      "step": 5370
    },
    {
      "epoch": 0.800565981531129,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.012006255585344057,
      "loss": 0.81,
      "num_input_tokens_seen": 3118112,
      "step": 5375
    },
    {
      "epoch": 0.8013106940720882,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.012017426273458445,
      "loss": 0.8168,
      "num_input_tokens_seen": 3121280,
      "step": 5380
    },
    {
      "epoch": 0.8020554066130474,
      "grad_norm": 0.044921875,
      "learning_rate": 0.012028596961572833,
      "loss": 0.7833,
      "num_input_tokens_seen": 3124352,
      "step": 5385
    },
    {
      "epoch": 0.8028001191540065,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.01203976764968722,
      "loss": 0.7945,
      "num_input_tokens_seen": 3127296,
      "step": 5390
    },
    {
      "epoch": 0.8035448316949657,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.012050938337801608,
      "loss": 0.82,
      "num_input_tokens_seen": 3130336,
      "step": 5395
    },
    {
      "epoch": 0.8042895442359249,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.012062109025915995,
      "loss": 0.8261,
      "num_input_tokens_seen": 3133408,
      "step": 5400
    },
    {
      "epoch": 0.8050342567768841,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.012073279714030383,
      "loss": 0.8018,
      "num_input_tokens_seen": 3136352,
      "step": 5405
    },
    {
      "epoch": 0.8057789693178433,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.01208445040214477,
      "loss": 0.7927,
      "num_input_tokens_seen": 3139232,
      "step": 5410
    },
    {
      "epoch": 0.8065236818588025,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.012095621090259159,
      "loss": 0.8014,
      "num_input_tokens_seen": 3142080,
      "step": 5415
    },
    {
      "epoch": 0.8072683943997617,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.012106791778373548,
      "loss": 0.8158,
      "num_input_tokens_seen": 3144864,
      "step": 5420
    },
    {
      "epoch": 0.8080131069407209,
      "grad_norm": 0.044921875,
      "learning_rate": 0.012117962466487936,
      "loss": 0.8133,
      "num_input_tokens_seen": 3147712,
      "step": 5425
    },
    {
      "epoch": 0.80875781948168,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.012129133154602324,
      "loss": 0.8073,
      "num_input_tokens_seen": 3150688,
      "step": 5430
    },
    {
      "epoch": 0.8095025320226392,
      "grad_norm": 0.025390625,
      "learning_rate": 0.012140303842716712,
      "loss": 0.8118,
      "num_input_tokens_seen": 3153760,
      "step": 5435
    },
    {
      "epoch": 0.8102472445635984,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.012151474530831098,
      "loss": 0.7649,
      "num_input_tokens_seen": 3156896,
      "step": 5440
    },
    {
      "epoch": 0.8109919571045576,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.012162645218945486,
      "loss": 0.8234,
      "num_input_tokens_seen": 3159648,
      "step": 5445
    },
    {
      "epoch": 0.8117366696455168,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.012173815907059874,
      "loss": 0.7999,
      "num_input_tokens_seen": 3162432,
      "step": 5450
    },
    {
      "epoch": 0.812481382186476,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.012184986595174262,
      "loss": 0.8002,
      "num_input_tokens_seen": 3165088,
      "step": 5455
    },
    {
      "epoch": 0.8132260947274352,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.01219615728328865,
      "loss": 0.8089,
      "num_input_tokens_seen": 3167936,
      "step": 5460
    },
    {
      "epoch": 0.8139708072683944,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.012207327971403038,
      "loss": 0.7996,
      "num_input_tokens_seen": 3170848,
      "step": 5465
    },
    {
      "epoch": 0.8147155198093536,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.012218498659517426,
      "loss": 0.8074,
      "num_input_tokens_seen": 3173760,
      "step": 5470
    },
    {
      "epoch": 0.8154602323503127,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.012229669347631814,
      "loss": 0.7984,
      "num_input_tokens_seen": 3176288,
      "step": 5475
    },
    {
      "epoch": 0.8162049448912719,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.012240840035746202,
      "loss": 0.8244,
      "num_input_tokens_seen": 3179392,
      "step": 5480
    },
    {
      "epoch": 0.8169496574322311,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.012252010723860588,
      "loss": 0.8104,
      "num_input_tokens_seen": 3182080,
      "step": 5485
    },
    {
      "epoch": 0.8176943699731903,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.012263181411974976,
      "loss": 0.7917,
      "num_input_tokens_seen": 3184480,
      "step": 5490
    },
    {
      "epoch": 0.8184390825141495,
      "grad_norm": 0.02734375,
      "learning_rate": 0.012274352100089364,
      "loss": 0.8091,
      "num_input_tokens_seen": 3187424,
      "step": 5495
    },
    {
      "epoch": 0.8191837950551087,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.012285522788203752,
      "loss": 0.8054,
      "num_input_tokens_seen": 3190304,
      "step": 5500
    },
    {
      "epoch": 0.819928507596068,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.01229669347631814,
      "loss": 0.8003,
      "num_input_tokens_seen": 3193280,
      "step": 5505
    },
    {
      "epoch": 0.8206732201370271,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.012307864164432528,
      "loss": 0.8141,
      "num_input_tokens_seen": 3196160,
      "step": 5510
    },
    {
      "epoch": 0.8214179326779864,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.012319034852546916,
      "loss": 0.7999,
      "num_input_tokens_seen": 3199072,
      "step": 5515
    },
    {
      "epoch": 0.8221626452189454,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.012330205540661304,
      "loss": 0.811,
      "num_input_tokens_seen": 3201856,
      "step": 5520
    },
    {
      "epoch": 0.8229073577599046,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.012341376228775692,
      "loss": 0.818,
      "num_input_tokens_seen": 3204672,
      "step": 5525
    },
    {
      "epoch": 0.8236520703008638,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.012352546916890078,
      "loss": 0.7992,
      "num_input_tokens_seen": 3207392,
      "step": 5530
    },
    {
      "epoch": 0.824396782841823,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.012363717605004468,
      "loss": 0.8103,
      "num_input_tokens_seen": 3210208,
      "step": 5535
    },
    {
      "epoch": 0.8251414953827823,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.012374888293118856,
      "loss": 0.8085,
      "num_input_tokens_seen": 3212832,
      "step": 5540
    },
    {
      "epoch": 0.8258862079237415,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.012386058981233244,
      "loss": 0.8143,
      "num_input_tokens_seen": 3215360,
      "step": 5545
    },
    {
      "epoch": 0.8266309204647007,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.012397229669347632,
      "loss": 0.7987,
      "num_input_tokens_seen": 3218208,
      "step": 5550
    },
    {
      "epoch": 0.8273756330056599,
      "grad_norm": 0.05078125,
      "learning_rate": 0.01240840035746202,
      "loss": 0.813,
      "num_input_tokens_seen": 3221248,
      "step": 5555
    },
    {
      "epoch": 0.828120345546619,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.012419571045576408,
      "loss": 0.8108,
      "num_input_tokens_seen": 3224416,
      "step": 5560
    },
    {
      "epoch": 0.8288650580875782,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.012430741733690796,
      "loss": 0.8077,
      "num_input_tokens_seen": 3227104,
      "step": 5565
    },
    {
      "epoch": 0.8296097706285374,
      "grad_norm": 0.064453125,
      "learning_rate": 0.012441912421805182,
      "loss": 0.7919,
      "num_input_tokens_seen": 3229856,
      "step": 5570
    },
    {
      "epoch": 0.8303544831694966,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.01245308310991957,
      "loss": 0.818,
      "num_input_tokens_seen": 3232960,
      "step": 5575
    },
    {
      "epoch": 0.8310991957104558,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.012464253798033958,
      "loss": 0.8051,
      "num_input_tokens_seen": 3235840,
      "step": 5580
    },
    {
      "epoch": 0.831843908251415,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.012475424486148346,
      "loss": 0.8106,
      "num_input_tokens_seen": 3238496,
      "step": 5585
    },
    {
      "epoch": 0.8325886207923742,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.012486595174262734,
      "loss": 0.7981,
      "num_input_tokens_seen": 3241088,
      "step": 5590
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.012497765862377122,
      "loss": 0.8031,
      "num_input_tokens_seen": 3244032,
      "step": 5595
    },
    {
      "epoch": 0.8340780458742926,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.01250893655049151,
      "loss": 0.8035,
      "num_input_tokens_seen": 3247040,
      "step": 5600
    },
    {
      "epoch": 0.8348227584152517,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.012520107238605898,
      "loss": 0.7781,
      "num_input_tokens_seen": 3250080,
      "step": 5605
    },
    {
      "epoch": 0.8355674709562109,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.012531277926720286,
      "loss": 0.8418,
      "num_input_tokens_seen": 3252992,
      "step": 5610
    },
    {
      "epoch": 0.8363121834971701,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.012542448614834672,
      "loss": 0.7999,
      "num_input_tokens_seen": 3256032,
      "step": 5615
    },
    {
      "epoch": 0.8370568960381293,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.01255361930294906,
      "loss": 0.7991,
      "num_input_tokens_seen": 3258656,
      "step": 5620
    },
    {
      "epoch": 0.8378016085790885,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.012564789991063448,
      "loss": 0.82,
      "num_input_tokens_seen": 3261472,
      "step": 5625
    },
    {
      "epoch": 0.8385463211200477,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.012575960679177836,
      "loss": 0.7853,
      "num_input_tokens_seen": 3264576,
      "step": 5630
    },
    {
      "epoch": 0.8392910336610069,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.012587131367292224,
      "loss": 0.795,
      "num_input_tokens_seen": 3267488,
      "step": 5635
    },
    {
      "epoch": 0.8400357462019661,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.012598302055406612,
      "loss": 0.8036,
      "num_input_tokens_seen": 3270432,
      "step": 5640
    },
    {
      "epoch": 0.8407804587429253,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.012609472743521,
      "loss": 0.7911,
      "num_input_tokens_seen": 3273216,
      "step": 5645
    },
    {
      "epoch": 0.8415251712838844,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01262064343163539,
      "loss": 0.7837,
      "num_input_tokens_seen": 3276448,
      "step": 5650
    },
    {
      "epoch": 0.8422698838248436,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.012631814119749778,
      "loss": 0.8102,
      "num_input_tokens_seen": 3279296,
      "step": 5655
    },
    {
      "epoch": 0.8430145963658028,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.012642984807864164,
      "loss": 0.8369,
      "num_input_tokens_seen": 3282112,
      "step": 5660
    },
    {
      "epoch": 0.843759308906762,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.012654155495978552,
      "loss": 0.784,
      "num_input_tokens_seen": 3285344,
      "step": 5665
    },
    {
      "epoch": 0.8445040214477212,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.01266532618409294,
      "loss": 0.82,
      "num_input_tokens_seen": 3288384,
      "step": 5670
    },
    {
      "epoch": 0.8452487339886804,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.012676496872207328,
      "loss": 0.7832,
      "num_input_tokens_seen": 3291360,
      "step": 5675
    },
    {
      "epoch": 0.8459934465296396,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.012687667560321716,
      "loss": 0.8049,
      "num_input_tokens_seen": 3294496,
      "step": 5680
    },
    {
      "epoch": 0.8467381590705988,
      "grad_norm": 0.037109375,
      "learning_rate": 0.012698838248436104,
      "loss": 0.8111,
      "num_input_tokens_seen": 3297216,
      "step": 5685
    },
    {
      "epoch": 0.8474828716115579,
      "grad_norm": 0.046875,
      "learning_rate": 0.012710008936550492,
      "loss": 0.8078,
      "num_input_tokens_seen": 3300000,
      "step": 5690
    },
    {
      "epoch": 0.8482275841525171,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.01272117962466488,
      "loss": 0.7891,
      "num_input_tokens_seen": 3302560,
      "step": 5695
    },
    {
      "epoch": 0.8489722966934763,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.012732350312779266,
      "loss": 0.7877,
      "num_input_tokens_seen": 3305312,
      "step": 5700
    },
    {
      "epoch": 0.8497170092344355,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.012743521000893654,
      "loss": 0.8203,
      "num_input_tokens_seen": 3308192,
      "step": 5705
    },
    {
      "epoch": 0.8504617217753947,
      "grad_norm": 0.05078125,
      "learning_rate": 0.012754691689008042,
      "loss": 0.8158,
      "num_input_tokens_seen": 3311296,
      "step": 5710
    },
    {
      "epoch": 0.8512064343163539,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.01276586237712243,
      "loss": 0.8054,
      "num_input_tokens_seen": 3313984,
      "step": 5715
    },
    {
      "epoch": 0.8519511468573131,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.012777033065236818,
      "loss": 0.8266,
      "num_input_tokens_seen": 3316864,
      "step": 5720
    },
    {
      "epoch": 0.8526958593982723,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.012788203753351206,
      "loss": 0.7955,
      "num_input_tokens_seen": 3319840,
      "step": 5725
    },
    {
      "epoch": 0.8534405719392315,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.012799374441465594,
      "loss": 0.8058,
      "num_input_tokens_seen": 3322720,
      "step": 5730
    },
    {
      "epoch": 0.8541852844801906,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.012810545129579982,
      "loss": 0.7936,
      "num_input_tokens_seen": 3325536,
      "step": 5735
    },
    {
      "epoch": 0.8549299970211498,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.01282171581769437,
      "loss": 0.7893,
      "num_input_tokens_seen": 3328544,
      "step": 5740
    },
    {
      "epoch": 0.855674709562109,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.012832886505808756,
      "loss": 0.8038,
      "num_input_tokens_seen": 3331520,
      "step": 5745
    },
    {
      "epoch": 0.8564194221030682,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.012844057193923144,
      "loss": 0.803,
      "num_input_tokens_seen": 3334464,
      "step": 5750
    },
    {
      "epoch": 0.8571641346440274,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.012855227882037532,
      "loss": 0.8052,
      "num_input_tokens_seen": 3337248,
      "step": 5755
    },
    {
      "epoch": 0.8579088471849866,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.012866398570151922,
      "loss": 0.7954,
      "num_input_tokens_seen": 3340128,
      "step": 5760
    },
    {
      "epoch": 0.8586535597259458,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.01287756925826631,
      "loss": 0.818,
      "num_input_tokens_seen": 3342784,
      "step": 5765
    },
    {
      "epoch": 0.859398272266905,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.012888739946380698,
      "loss": 0.8068,
      "num_input_tokens_seen": 3345504,
      "step": 5770
    },
    {
      "epoch": 0.8601429848078642,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.012899910634495085,
      "loss": 0.8134,
      "num_input_tokens_seen": 3348416,
      "step": 5775
    },
    {
      "epoch": 0.8608876973488233,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.012911081322609473,
      "loss": 0.7868,
      "num_input_tokens_seen": 3351296,
      "step": 5780
    },
    {
      "epoch": 0.8616324098897825,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.01292225201072386,
      "loss": 0.8039,
      "num_input_tokens_seen": 3354368,
      "step": 5785
    },
    {
      "epoch": 0.8623771224307417,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.012933422698838248,
      "loss": 0.8047,
      "num_input_tokens_seen": 3357120,
      "step": 5790
    },
    {
      "epoch": 0.8631218349717009,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.012944593386952636,
      "loss": 0.8105,
      "num_input_tokens_seen": 3360192,
      "step": 5795
    },
    {
      "epoch": 0.8638665475126601,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.012955764075067024,
      "loss": 0.8054,
      "num_input_tokens_seen": 3362944,
      "step": 5800
    },
    {
      "epoch": 0.8646112600536193,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.012966934763181412,
      "loss": 0.7974,
      "num_input_tokens_seen": 3365920,
      "step": 5805
    },
    {
      "epoch": 0.8653559725945785,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0129781054512958,
      "loss": 0.8119,
      "num_input_tokens_seen": 3368832,
      "step": 5810
    },
    {
      "epoch": 0.8661006851355377,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.012989276139410188,
      "loss": 0.8121,
      "num_input_tokens_seen": 3371808,
      "step": 5815
    },
    {
      "epoch": 0.8668453976764968,
      "grad_norm": 0.0390625,
      "learning_rate": 0.013000446827524575,
      "loss": 0.8148,
      "num_input_tokens_seen": 3374976,
      "step": 5820
    },
    {
      "epoch": 0.867590110217456,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.013011617515638963,
      "loss": 0.7944,
      "num_input_tokens_seen": 3378112,
      "step": 5825
    },
    {
      "epoch": 0.8683348227584152,
      "grad_norm": 0.03125,
      "learning_rate": 0.01302278820375335,
      "loss": 0.8143,
      "num_input_tokens_seen": 3381024,
      "step": 5830
    },
    {
      "epoch": 0.8690795352993744,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.013033958891867738,
      "loss": 0.8114,
      "num_input_tokens_seen": 3383808,
      "step": 5835
    },
    {
      "epoch": 0.8698242478403336,
      "grad_norm": 0.046875,
      "learning_rate": 0.013045129579982126,
      "loss": 0.8025,
      "num_input_tokens_seen": 3386848,
      "step": 5840
    },
    {
      "epoch": 0.8705689603812928,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.013056300268096514,
      "loss": 0.8064,
      "num_input_tokens_seen": 3389760,
      "step": 5845
    },
    {
      "epoch": 0.871313672922252,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.013067470956210902,
      "loss": 0.8154,
      "num_input_tokens_seen": 3392704,
      "step": 5850
    },
    {
      "epoch": 0.8720583854632112,
      "grad_norm": 0.07421875,
      "learning_rate": 0.01307864164432529,
      "loss": 0.8065,
      "num_input_tokens_seen": 3395744,
      "step": 5855
    },
    {
      "epoch": 0.8728030980041704,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.013089812332439677,
      "loss": 0.8019,
      "num_input_tokens_seen": 3398592,
      "step": 5860
    },
    {
      "epoch": 0.8735478105451295,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.013100983020554065,
      "loss": 0.805,
      "num_input_tokens_seen": 3401216,
      "step": 5865
    },
    {
      "epoch": 0.8742925230860887,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.013112153708668453,
      "loss": 0.8075,
      "num_input_tokens_seen": 3404256,
      "step": 5870
    },
    {
      "epoch": 0.8750372356270479,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.013123324396782841,
      "loss": 0.8127,
      "num_input_tokens_seen": 3407232,
      "step": 5875
    },
    {
      "epoch": 0.8757819481680071,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.01313449508489723,
      "loss": 0.8013,
      "num_input_tokens_seen": 3409984,
      "step": 5880
    },
    {
      "epoch": 0.8765266607089663,
      "grad_norm": 0.052734375,
      "learning_rate": 0.013145665773011617,
      "loss": 0.8063,
      "num_input_tokens_seen": 3412768,
      "step": 5885
    },
    {
      "epoch": 0.8772713732499255,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.013156836461126005,
      "loss": 0.8171,
      "num_input_tokens_seen": 3415456,
      "step": 5890
    },
    {
      "epoch": 0.8780160857908847,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.013168007149240393,
      "loss": 0.8067,
      "num_input_tokens_seen": 3418368,
      "step": 5895
    },
    {
      "epoch": 0.878760798331844,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.013179177837354781,
      "loss": 0.816,
      "num_input_tokens_seen": 3421248,
      "step": 5900
    },
    {
      "epoch": 0.8795055108728032,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.01319034852546917,
      "loss": 0.8065,
      "num_input_tokens_seen": 3424544,
      "step": 5905
    },
    {
      "epoch": 0.8802502234137622,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.013201519213583557,
      "loss": 0.8183,
      "num_input_tokens_seen": 3427584,
      "step": 5910
    },
    {
      "epoch": 0.8809949359547214,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.013212689901697943,
      "loss": 0.8124,
      "num_input_tokens_seen": 3430400,
      "step": 5915
    },
    {
      "epoch": 0.8817396484956807,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.013223860589812331,
      "loss": 0.8006,
      "num_input_tokens_seen": 3433248,
      "step": 5920
    },
    {
      "epoch": 0.8824843610366399,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.01323503127792672,
      "loss": 0.7992,
      "num_input_tokens_seen": 3436032,
      "step": 5925
    },
    {
      "epoch": 0.883229073577599,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.013246201966041107,
      "loss": 0.8244,
      "num_input_tokens_seen": 3438912,
      "step": 5930
    },
    {
      "epoch": 0.8839737861185583,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.013257372654155495,
      "loss": 0.7996,
      "num_input_tokens_seen": 3441760,
      "step": 5935
    },
    {
      "epoch": 0.8847184986595175,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.013268543342269883,
      "loss": 0.8019,
      "num_input_tokens_seen": 3444448,
      "step": 5940
    },
    {
      "epoch": 0.8854632112004767,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.013279714030384271,
      "loss": 0.8066,
      "num_input_tokens_seen": 3447104,
      "step": 5945
    },
    {
      "epoch": 0.8862079237414358,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.01329088471849866,
      "loss": 0.8091,
      "num_input_tokens_seen": 3449856,
      "step": 5950
    },
    {
      "epoch": 0.886952636282395,
      "grad_norm": 0.02734375,
      "learning_rate": 0.013302055406613047,
      "loss": 0.7831,
      "num_input_tokens_seen": 3452544,
      "step": 5955
    },
    {
      "epoch": 0.8876973488233542,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.013313226094727433,
      "loss": 0.8009,
      "num_input_tokens_seen": 3455136,
      "step": 5960
    },
    {
      "epoch": 0.8884420613643134,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.013324396782841821,
      "loss": 0.8075,
      "num_input_tokens_seen": 3458272,
      "step": 5965
    },
    {
      "epoch": 0.8891867739052726,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.01333556747095621,
      "loss": 0.8151,
      "num_input_tokens_seen": 3461504,
      "step": 5970
    },
    {
      "epoch": 0.8899314864462318,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.013346738159070597,
      "loss": 0.7904,
      "num_input_tokens_seen": 3464448,
      "step": 5975
    },
    {
      "epoch": 0.890676198987191,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.013357908847184985,
      "loss": 0.8159,
      "num_input_tokens_seen": 3467264,
      "step": 5980
    },
    {
      "epoch": 0.8914209115281502,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.013369079535299375,
      "loss": 0.7972,
      "num_input_tokens_seen": 3470080,
      "step": 5985
    },
    {
      "epoch": 0.8921656240691094,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.013380250223413763,
      "loss": 0.8169,
      "num_input_tokens_seen": 3473088,
      "step": 5990
    },
    {
      "epoch": 0.8929103366100685,
      "grad_norm": 0.05078125,
      "learning_rate": 0.013391420911528151,
      "loss": 0.8193,
      "num_input_tokens_seen": 3476032,
      "step": 5995
    },
    {
      "epoch": 0.8936550491510277,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.013402591599642539,
      "loss": 0.8121,
      "num_input_tokens_seen": 3478560,
      "step": 6000
    },
    {
      "epoch": 0.8943997616919869,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.013413762287756925,
      "loss": 0.8072,
      "num_input_tokens_seen": 3482048,
      "step": 6005
    },
    {
      "epoch": 0.8951444742329461,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.013424932975871313,
      "loss": 0.8076,
      "num_input_tokens_seen": 3485184,
      "step": 6010
    },
    {
      "epoch": 0.8958891867739053,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.013436103663985701,
      "loss": 0.7952,
      "num_input_tokens_seen": 3488000,
      "step": 6015
    },
    {
      "epoch": 0.8966338993148645,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.013447274352100089,
      "loss": 0.7965,
      "num_input_tokens_seen": 3490816,
      "step": 6020
    },
    {
      "epoch": 0.8973786118558237,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.013458445040214477,
      "loss": 0.7955,
      "num_input_tokens_seen": 3493728,
      "step": 6025
    },
    {
      "epoch": 0.8981233243967829,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.013469615728328865,
      "loss": 0.8088,
      "num_input_tokens_seen": 3496384,
      "step": 6030
    },
    {
      "epoch": 0.898868036937742,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.013480786416443253,
      "loss": 0.8103,
      "num_input_tokens_seen": 3499104,
      "step": 6035
    },
    {
      "epoch": 0.8996127494787012,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.013491957104557641,
      "loss": 0.8183,
      "num_input_tokens_seen": 3501888,
      "step": 6040
    },
    {
      "epoch": 0.9003574620196604,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.013503127792672027,
      "loss": 0.8303,
      "num_input_tokens_seen": 3504608,
      "step": 6045
    },
    {
      "epoch": 0.9011021745606196,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.013514298480786415,
      "loss": 0.7928,
      "num_input_tokens_seen": 3507232,
      "step": 6050
    },
    {
      "epoch": 0.9018468871015788,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.013525469168900803,
      "loss": 0.8083,
      "num_input_tokens_seen": 3510080,
      "step": 6055
    },
    {
      "epoch": 0.902591599642538,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.013536639857015191,
      "loss": 0.8051,
      "num_input_tokens_seen": 3513280,
      "step": 6060
    },
    {
      "epoch": 0.9033363121834972,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.013547810545129579,
      "loss": 0.7937,
      "num_input_tokens_seen": 3516128,
      "step": 6065
    },
    {
      "epoch": 0.9040810247244564,
      "grad_norm": 0.0546875,
      "learning_rate": 0.013558981233243967,
      "loss": 0.8062,
      "num_input_tokens_seen": 3518944,
      "step": 6070
    },
    {
      "epoch": 0.9048257372654156,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.013570151921358355,
      "loss": 0.7959,
      "num_input_tokens_seen": 3521824,
      "step": 6075
    },
    {
      "epoch": 0.9055704498063747,
      "grad_norm": 0.037109375,
      "learning_rate": 0.013581322609472743,
      "loss": 0.7991,
      "num_input_tokens_seen": 3524640,
      "step": 6080
    },
    {
      "epoch": 0.9063151623473339,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.013592493297587131,
      "loss": 0.813,
      "num_input_tokens_seen": 3527648,
      "step": 6085
    },
    {
      "epoch": 0.9070598748882931,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.013603663985701517,
      "loss": 0.7946,
      "num_input_tokens_seen": 3530720,
      "step": 6090
    },
    {
      "epoch": 0.9078045874292523,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.013614834673815907,
      "loss": 0.7993,
      "num_input_tokens_seen": 3533600,
      "step": 6095
    },
    {
      "epoch": 0.9085492999702115,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.013626005361930295,
      "loss": 0.7929,
      "num_input_tokens_seen": 3536320,
      "step": 6100
    },
    {
      "epoch": 0.9092940125111707,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.013637176050044683,
      "loss": 0.7831,
      "num_input_tokens_seen": 3539104,
      "step": 6105
    },
    {
      "epoch": 0.9100387250521299,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.01364834673815907,
      "loss": 0.8207,
      "num_input_tokens_seen": 3542272,
      "step": 6110
    },
    {
      "epoch": 0.9107834375930891,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.013659517426273459,
      "loss": 0.8237,
      "num_input_tokens_seen": 3545120,
      "step": 6115
    },
    {
      "epoch": 0.9115281501340483,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.013670688114387847,
      "loss": 0.7968,
      "num_input_tokens_seen": 3548128,
      "step": 6120
    },
    {
      "epoch": 0.9122728626750074,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.013681858802502235,
      "loss": 0.8067,
      "num_input_tokens_seen": 3551712,
      "step": 6125
    },
    {
      "epoch": 0.9130175752159666,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.013693029490616623,
      "loss": 0.8164,
      "num_input_tokens_seen": 3554784,
      "step": 6130
    },
    {
      "epoch": 0.9137622877569258,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.013704200178731009,
      "loss": 0.7939,
      "num_input_tokens_seen": 3557888,
      "step": 6135
    },
    {
      "epoch": 0.914507000297885,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.013715370866845397,
      "loss": 0.7987,
      "num_input_tokens_seen": 3561248,
      "step": 6140
    },
    {
      "epoch": 0.9152517128388442,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.013726541554959785,
      "loss": 0.8002,
      "num_input_tokens_seen": 3564256,
      "step": 6145
    },
    {
      "epoch": 0.9159964253798034,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.013737712243074173,
      "loss": 0.8237,
      "num_input_tokens_seen": 3567424,
      "step": 6150
    },
    {
      "epoch": 0.9167411379207626,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.01374888293118856,
      "loss": 0.7894,
      "num_input_tokens_seen": 3570432,
      "step": 6155
    },
    {
      "epoch": 0.9174858504617218,
      "grad_norm": 0.05859375,
      "learning_rate": 0.013760053619302949,
      "loss": 0.8018,
      "num_input_tokens_seen": 3573216,
      "step": 6160
    },
    {
      "epoch": 0.9182305630026809,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.013771224307417337,
      "loss": 0.8075,
      "num_input_tokens_seen": 3575872,
      "step": 6165
    },
    {
      "epoch": 0.9189752755436401,
      "grad_norm": 0.033203125,
      "learning_rate": 0.013782394995531725,
      "loss": 0.8003,
      "num_input_tokens_seen": 3578848,
      "step": 6170
    },
    {
      "epoch": 0.9197199880845993,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.013793565683646111,
      "loss": 0.8148,
      "num_input_tokens_seen": 3581952,
      "step": 6175
    },
    {
      "epoch": 0.9204647006255585,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.013804736371760499,
      "loss": 0.8258,
      "num_input_tokens_seen": 3584960,
      "step": 6180
    },
    {
      "epoch": 0.9212094131665177,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.013815907059874887,
      "loss": 0.8054,
      "num_input_tokens_seen": 3588192,
      "step": 6185
    },
    {
      "epoch": 0.9219541257074769,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.013827077747989275,
      "loss": 0.813,
      "num_input_tokens_seen": 3591072,
      "step": 6190
    },
    {
      "epoch": 0.9226988382484361,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.013838248436103663,
      "loss": 0.8193,
      "num_input_tokens_seen": 3594048,
      "step": 6195
    },
    {
      "epoch": 0.9234435507893953,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01384941912421805,
      "loss": 0.7956,
      "num_input_tokens_seen": 3596960,
      "step": 6200
    },
    {
      "epoch": 0.9241882633303545,
      "grad_norm": 0.048828125,
      "learning_rate": 0.013860589812332439,
      "loss": 0.8127,
      "num_input_tokens_seen": 3600224,
      "step": 6205
    },
    {
      "epoch": 0.9249329758713136,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.013871760500446828,
      "loss": 0.7939,
      "num_input_tokens_seen": 3603360,
      "step": 6210
    },
    {
      "epoch": 0.9256776884122728,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.013882931188561216,
      "loss": 0.8018,
      "num_input_tokens_seen": 3606144,
      "step": 6215
    },
    {
      "epoch": 0.926422400953232,
      "grad_norm": 0.064453125,
      "learning_rate": 0.013894101876675603,
      "loss": 0.8331,
      "num_input_tokens_seen": 3609280,
      "step": 6220
    },
    {
      "epoch": 0.9271671134941912,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.01390527256478999,
      "loss": 0.8229,
      "num_input_tokens_seen": 3612320,
      "step": 6225
    },
    {
      "epoch": 0.9279118260351504,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.013916443252904379,
      "loss": 0.8139,
      "num_input_tokens_seen": 3615296,
      "step": 6230
    },
    {
      "epoch": 0.9286565385761096,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.013927613941018767,
      "loss": 0.8148,
      "num_input_tokens_seen": 3618336,
      "step": 6235
    },
    {
      "epoch": 0.9294012511170688,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.013938784629133154,
      "loss": 0.8099,
      "num_input_tokens_seen": 3621280,
      "step": 6240
    },
    {
      "epoch": 0.930145963658028,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.013949955317247542,
      "loss": 0.7992,
      "num_input_tokens_seen": 3624352,
      "step": 6245
    },
    {
      "epoch": 0.9308906761989872,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.01396112600536193,
      "loss": 0.8125,
      "num_input_tokens_seen": 3627488,
      "step": 6250
    },
    {
      "epoch": 0.9316353887399463,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.013972296693476318,
      "loss": 0.8096,
      "num_input_tokens_seen": 3630400,
      "step": 6255
    },
    {
      "epoch": 0.9323801012809055,
      "grad_norm": 0.0234375,
      "learning_rate": 0.013983467381590706,
      "loss": 0.8004,
      "num_input_tokens_seen": 3632928,
      "step": 6260
    },
    {
      "epoch": 0.9331248138218647,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.013994638069705093,
      "loss": 0.8005,
      "num_input_tokens_seen": 3635840,
      "step": 6265
    },
    {
      "epoch": 0.9338695263628239,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.01400580875781948,
      "loss": 0.8054,
      "num_input_tokens_seen": 3638656,
      "step": 6270
    },
    {
      "epoch": 0.9346142389037831,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.014016979445933869,
      "loss": 0.8107,
      "num_input_tokens_seen": 3641472,
      "step": 6275
    },
    {
      "epoch": 0.9353589514447423,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.014028150134048257,
      "loss": 0.8041,
      "num_input_tokens_seen": 3644192,
      "step": 6280
    },
    {
      "epoch": 0.9361036639857016,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.014039320822162644,
      "loss": 0.818,
      "num_input_tokens_seen": 3646976,
      "step": 6285
    },
    {
      "epoch": 0.9368483765266608,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.014050491510277032,
      "loss": 0.8055,
      "num_input_tokens_seen": 3649856,
      "step": 6290
    },
    {
      "epoch": 0.9375930890676198,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01406166219839142,
      "loss": 0.7981,
      "num_input_tokens_seen": 3652736,
      "step": 6295
    },
    {
      "epoch": 0.938337801608579,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.014072832886505808,
      "loss": 0.7895,
      "num_input_tokens_seen": 3655648,
      "step": 6300
    },
    {
      "epoch": 0.9390825141495382,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.014084003574620195,
      "loss": 0.8139,
      "num_input_tokens_seen": 3658432,
      "step": 6305
    },
    {
      "epoch": 0.9398272266904975,
      "grad_norm": 0.060546875,
      "learning_rate": 0.014095174262734583,
      "loss": 0.8006,
      "num_input_tokens_seen": 3661280,
      "step": 6310
    },
    {
      "epoch": 0.9405719392314567,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.01410634495084897,
      "loss": 0.8163,
      "num_input_tokens_seen": 3664576,
      "step": 6315
    },
    {
      "epoch": 0.9413166517724159,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.014117515638963359,
      "loss": 0.8059,
      "num_input_tokens_seen": 3667264,
      "step": 6320
    },
    {
      "epoch": 0.9420613643133751,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.014128686327077748,
      "loss": 0.7986,
      "num_input_tokens_seen": 3670144,
      "step": 6325
    },
    {
      "epoch": 0.9428060768543343,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.014139857015192136,
      "loss": 0.8215,
      "num_input_tokens_seen": 3672992,
      "step": 6330
    },
    {
      "epoch": 0.9435507893952935,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.014151027703306524,
      "loss": 0.8138,
      "num_input_tokens_seen": 3675840,
      "step": 6335
    },
    {
      "epoch": 0.9442955019362526,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.014162198391420912,
      "loss": 0.7865,
      "num_input_tokens_seen": 3678944,
      "step": 6340
    },
    {
      "epoch": 0.9450402144772118,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.0141733690795353,
      "loss": 0.8031,
      "num_input_tokens_seen": 3681728,
      "step": 6345
    },
    {
      "epoch": 0.945784927018171,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.014184539767649686,
      "loss": 0.8163,
      "num_input_tokens_seen": 3684672,
      "step": 6350
    },
    {
      "epoch": 0.9465296395591302,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.014195710455764074,
      "loss": 0.8092,
      "num_input_tokens_seen": 3687296,
      "step": 6355
    },
    {
      "epoch": 0.9472743521000894,
      "grad_norm": 0.046875,
      "learning_rate": 0.014206881143878462,
      "loss": 0.7923,
      "num_input_tokens_seen": 3689856,
      "step": 6360
    },
    {
      "epoch": 0.9480190646410486,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.01421805183199285,
      "loss": 0.7985,
      "num_input_tokens_seen": 3692704,
      "step": 6365
    },
    {
      "epoch": 0.9487637771820078,
      "grad_norm": 0.0234375,
      "learning_rate": 0.014229222520107238,
      "loss": 0.7914,
      "num_input_tokens_seen": 3695424,
      "step": 6370
    },
    {
      "epoch": 0.949508489722967,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.014240393208221626,
      "loss": 0.8087,
      "num_input_tokens_seen": 3698592,
      "step": 6375
    },
    {
      "epoch": 0.9502532022639262,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.014251563896336014,
      "loss": 0.8106,
      "num_input_tokens_seen": 3701568,
      "step": 6380
    },
    {
      "epoch": 0.9509979148048853,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.014262734584450402,
      "loss": 0.81,
      "num_input_tokens_seen": 3704672,
      "step": 6385
    },
    {
      "epoch": 0.9517426273458445,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.01427390527256479,
      "loss": 0.7914,
      "num_input_tokens_seen": 3707488,
      "step": 6390
    },
    {
      "epoch": 0.9524873398868037,
      "grad_norm": 0.0546875,
      "learning_rate": 0.014285075960679176,
      "loss": 0.7909,
      "num_input_tokens_seen": 3710528,
      "step": 6395
    },
    {
      "epoch": 0.9532320524277629,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.014296246648793564,
      "loss": 0.8176,
      "num_input_tokens_seen": 3713664,
      "step": 6400
    },
    {
      "epoch": 0.9539767649687221,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.014307417336907952,
      "loss": 0.7824,
      "num_input_tokens_seen": 3716704,
      "step": 6405
    },
    {
      "epoch": 0.9547214775096813,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01431858802502234,
      "loss": 0.8177,
      "num_input_tokens_seen": 3719456,
      "step": 6410
    },
    {
      "epoch": 0.9554661900506405,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.014329758713136728,
      "loss": 0.7949,
      "num_input_tokens_seen": 3722400,
      "step": 6415
    },
    {
      "epoch": 0.9562109025915997,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.014340929401251116,
      "loss": 0.8079,
      "num_input_tokens_seen": 3725024,
      "step": 6420
    },
    {
      "epoch": 0.9569556151325588,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.014352100089365504,
      "loss": 0.823,
      "num_input_tokens_seen": 3727808,
      "step": 6425
    },
    {
      "epoch": 0.957700327673518,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.014363270777479892,
      "loss": 0.8397,
      "num_input_tokens_seen": 3730432,
      "step": 6430
    },
    {
      "epoch": 0.9584450402144772,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.01437444146559428,
      "loss": 0.8023,
      "num_input_tokens_seen": 3733088,
      "step": 6435
    },
    {
      "epoch": 0.9591897527554364,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.014385612153708668,
      "loss": 0.7973,
      "num_input_tokens_seen": 3735840,
      "step": 6440
    },
    {
      "epoch": 0.9599344652963956,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.014396782841823056,
      "loss": 0.7992,
      "num_input_tokens_seen": 3738464,
      "step": 6445
    },
    {
      "epoch": 0.9606791778373548,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.014407953529937444,
      "loss": 0.7993,
      "num_input_tokens_seen": 3741280,
      "step": 6450
    },
    {
      "epoch": 0.961423890378314,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.014419124218051832,
      "loss": 0.7792,
      "num_input_tokens_seen": 3744256,
      "step": 6455
    },
    {
      "epoch": 0.9621686029192732,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.01443029490616622,
      "loss": 0.8167,
      "num_input_tokens_seen": 3747456,
      "step": 6460
    },
    {
      "epoch": 0.9629133154602324,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.014441465594280608,
      "loss": 0.787,
      "num_input_tokens_seen": 3750240,
      "step": 6465
    },
    {
      "epoch": 0.9636580280011915,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.014452636282394996,
      "loss": 0.8191,
      "num_input_tokens_seen": 3752960,
      "step": 6470
    },
    {
      "epoch": 0.9644027405421507,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.014463806970509384,
      "loss": 0.7898,
      "num_input_tokens_seen": 3755840,
      "step": 6475
    },
    {
      "epoch": 0.9651474530831099,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.01447497765862377,
      "loss": 0.8084,
      "num_input_tokens_seen": 3758560,
      "step": 6480
    },
    {
      "epoch": 0.9658921656240691,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.014486148346738158,
      "loss": 0.8083,
      "num_input_tokens_seen": 3761376,
      "step": 6485
    },
    {
      "epoch": 0.9666368781650283,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.014497319034852546,
      "loss": 0.8182,
      "num_input_tokens_seen": 3764128,
      "step": 6490
    },
    {
      "epoch": 0.9673815907059875,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.014508489722966934,
      "loss": 0.8054,
      "num_input_tokens_seen": 3768128,
      "step": 6495
    },
    {
      "epoch": 0.9681263032469467,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.014519660411081322,
      "loss": 0.8145,
      "num_input_tokens_seen": 3770976,
      "step": 6500
    },
    {
      "epoch": 0.9688710157879059,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.01453083109919571,
      "loss": 0.7892,
      "num_input_tokens_seen": 3773888,
      "step": 6505
    },
    {
      "epoch": 0.9696157283288651,
      "grad_norm": 0.03515625,
      "learning_rate": 0.014542001787310098,
      "loss": 0.7948,
      "num_input_tokens_seen": 3777024,
      "step": 6510
    },
    {
      "epoch": 0.9703604408698242,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.014553172475424486,
      "loss": 0.8057,
      "num_input_tokens_seen": 3779872,
      "step": 6515
    },
    {
      "epoch": 0.9711051534107834,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.014564343163538874,
      "loss": 0.804,
      "num_input_tokens_seen": 3782656,
      "step": 6520
    },
    {
      "epoch": 0.9718498659517426,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01457551385165326,
      "loss": 0.8154,
      "num_input_tokens_seen": 3785504,
      "step": 6525
    },
    {
      "epoch": 0.9725945784927018,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.014586684539767648,
      "loss": 0.7985,
      "num_input_tokens_seen": 3788160,
      "step": 6530
    },
    {
      "epoch": 0.973339291033661,
      "grad_norm": 0.056640625,
      "learning_rate": 0.014597855227882036,
      "loss": 0.812,
      "num_input_tokens_seen": 3791424,
      "step": 6535
    },
    {
      "epoch": 0.9740840035746202,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.014609025915996424,
      "loss": 0.8049,
      "num_input_tokens_seen": 3794400,
      "step": 6540
    },
    {
      "epoch": 0.9748287161155794,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.014620196604110812,
      "loss": 0.8068,
      "num_input_tokens_seen": 3797408,
      "step": 6545
    },
    {
      "epoch": 0.9755734286565386,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.014631367292225202,
      "loss": 0.8016,
      "num_input_tokens_seen": 3800352,
      "step": 6550
    },
    {
      "epoch": 0.9763181411974977,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.01464253798033959,
      "loss": 0.7978,
      "num_input_tokens_seen": 3803328,
      "step": 6555
    },
    {
      "epoch": 0.9770628537384569,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.014653708668453978,
      "loss": 0.8172,
      "num_input_tokens_seen": 3806176,
      "step": 6560
    },
    {
      "epoch": 0.9778075662794161,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.014664879356568364,
      "loss": 0.8204,
      "num_input_tokens_seen": 3809088,
      "step": 6565
    },
    {
      "epoch": 0.9785522788203753,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.014676050044682752,
      "loss": 0.799,
      "num_input_tokens_seen": 3811680,
      "step": 6570
    },
    {
      "epoch": 0.9792969913613345,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.01468722073279714,
      "loss": 0.8039,
      "num_input_tokens_seen": 3814592,
      "step": 6575
    },
    {
      "epoch": 0.9800417039022937,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.014698391420911528,
      "loss": 0.7918,
      "num_input_tokens_seen": 3817728,
      "step": 6580
    },
    {
      "epoch": 0.9807864164432529,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.014709562109025916,
      "loss": 0.8036,
      "num_input_tokens_seen": 3820608,
      "step": 6585
    },
    {
      "epoch": 0.9815311289842121,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.014720732797140304,
      "loss": 0.7988,
      "num_input_tokens_seen": 3823648,
      "step": 6590
    },
    {
      "epoch": 0.9822758415251713,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.014731903485254692,
      "loss": 0.7981,
      "num_input_tokens_seen": 3826624,
      "step": 6595
    },
    {
      "epoch": 0.9830205540661304,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.01474307417336908,
      "loss": 0.7964,
      "num_input_tokens_seen": 3829632,
      "step": 6600
    },
    {
      "epoch": 0.9837652666070896,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.014754244861483468,
      "loss": 0.8394,
      "num_input_tokens_seen": 3832384,
      "step": 6605
    },
    {
      "epoch": 0.9845099791480488,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.014765415549597854,
      "loss": 0.822,
      "num_input_tokens_seen": 3835488,
      "step": 6610
    },
    {
      "epoch": 0.985254691689008,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.014776586237712242,
      "loss": 0.7932,
      "num_input_tokens_seen": 3838560,
      "step": 6615
    },
    {
      "epoch": 0.9859994042299672,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01478775692582663,
      "loss": 0.7911,
      "num_input_tokens_seen": 3841376,
      "step": 6620
    },
    {
      "epoch": 0.9867441167709264,
      "grad_norm": 0.083984375,
      "learning_rate": 0.014798927613941018,
      "loss": 0.8141,
      "num_input_tokens_seen": 3844192,
      "step": 6625
    },
    {
      "epoch": 0.9874888293118856,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.014810098302055406,
      "loss": 0.8215,
      "num_input_tokens_seen": 3847424,
      "step": 6630
    },
    {
      "epoch": 0.9882335418528448,
      "grad_norm": 0.0625,
      "learning_rate": 0.014821268990169794,
      "loss": 0.8119,
      "num_input_tokens_seen": 3850048,
      "step": 6635
    },
    {
      "epoch": 0.988978254393804,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.014832439678284182,
      "loss": 0.8461,
      "num_input_tokens_seen": 3852736,
      "step": 6640
    },
    {
      "epoch": 0.9897229669347631,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.01484361036639857,
      "loss": 0.8075,
      "num_input_tokens_seen": 3855456,
      "step": 6645
    },
    {
      "epoch": 0.9904676794757223,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.014854781054512958,
      "loss": 0.8061,
      "num_input_tokens_seen": 3858080,
      "step": 6650
    },
    {
      "epoch": 0.9912123920166815,
      "grad_norm": 0.06640625,
      "learning_rate": 0.014865951742627344,
      "loss": 0.7976,
      "num_input_tokens_seen": 3860864,
      "step": 6655
    },
    {
      "epoch": 0.9919571045576407,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.014877122430741734,
      "loss": 0.8055,
      "num_input_tokens_seen": 3863840,
      "step": 6660
    },
    {
      "epoch": 0.9927018170986,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.014888293118856121,
      "loss": 0.7923,
      "num_input_tokens_seen": 3866720,
      "step": 6665
    },
    {
      "epoch": 0.9934465296395592,
      "grad_norm": 0.06640625,
      "learning_rate": 0.01489946380697051,
      "loss": 0.8172,
      "num_input_tokens_seen": 3869504,
      "step": 6670
    },
    {
      "epoch": 0.9941912421805184,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.014910634495084897,
      "loss": 0.8171,
      "num_input_tokens_seen": 3872608,
      "step": 6675
    },
    {
      "epoch": 0.9949359547214776,
      "grad_norm": 0.0703125,
      "learning_rate": 0.014921805183199285,
      "loss": 0.8407,
      "num_input_tokens_seen": 3875584,
      "step": 6680
    },
    {
      "epoch": 0.9956806672624366,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.014932975871313673,
      "loss": 0.8208,
      "num_input_tokens_seen": 3878336,
      "step": 6685
    },
    {
      "epoch": 0.9964253798033958,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.014944146559428061,
      "loss": 0.806,
      "num_input_tokens_seen": 3881600,
      "step": 6690
    },
    {
      "epoch": 0.997170092344355,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.014955317247542448,
      "loss": 0.822,
      "num_input_tokens_seen": 3884160,
      "step": 6695
    },
    {
      "epoch": 0.9979148048853143,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.014966487935656836,
      "loss": 0.8004,
      "num_input_tokens_seen": 3886944,
      "step": 6700
    },
    {
      "epoch": 0.9986595174262735,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.014977658623771224,
      "loss": 0.8172,
      "num_input_tokens_seen": 3889856,
      "step": 6705
    },
    {
      "epoch": 0.9994042299672327,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.014988829311885611,
      "loss": 0.8045,
      "num_input_tokens_seen": 3892704,
      "step": 6710
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.8024358153343201,
      "eval_runtime": 70.6851,
      "eval_samples_per_second": 42.215,
      "eval_steps_per_second": 10.554,
      "num_input_tokens_seen": 3894688,
      "step": 6714
    },
    {
      "epoch": 1.0001489425081918,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.015,
      "loss": 0.7923,
      "num_input_tokens_seen": 3895200,
      "step": 6715
    },
    {
      "epoch": 1.000893655049151,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.015011170688114387,
      "loss": 0.8033,
      "num_input_tokens_seen": 3897984,
      "step": 6720
    },
    {
      "epoch": 1.0016383675901102,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.015022341376228775,
      "loss": 0.8,
      "num_input_tokens_seen": 3901024,
      "step": 6725
    },
    {
      "epoch": 1.0023830801310694,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.015033512064343163,
      "loss": 0.8144,
      "num_input_tokens_seen": 3904032,
      "step": 6730
    },
    {
      "epoch": 1.0031277926720286,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.015044682752457551,
      "loss": 0.7985,
      "num_input_tokens_seen": 3906912,
      "step": 6735
    },
    {
      "epoch": 1.0038725052129878,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.01505585344057194,
      "loss": 0.8279,
      "num_input_tokens_seen": 3909952,
      "step": 6740
    },
    {
      "epoch": 1.004617217753947,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.015067024128686327,
      "loss": 0.795,
      "num_input_tokens_seen": 3912800,
      "step": 6745
    },
    {
      "epoch": 1.0053619302949062,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.015078194816800715,
      "loss": 0.8183,
      "num_input_tokens_seen": 3915584,
      "step": 6750
    },
    {
      "epoch": 1.0061066428358654,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.015089365504915103,
      "loss": 0.8189,
      "num_input_tokens_seen": 3918432,
      "step": 6755
    },
    {
      "epoch": 1.0068513553768246,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.015100536193029488,
      "loss": 0.8139,
      "num_input_tokens_seen": 3921216,
      "step": 6760
    },
    {
      "epoch": 1.0075960679177838,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.015111706881143876,
      "loss": 0.8178,
      "num_input_tokens_seen": 3924256,
      "step": 6765
    },
    {
      "epoch": 1.008340780458743,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.015122877569258265,
      "loss": 0.8088,
      "num_input_tokens_seen": 3926976,
      "step": 6770
    },
    {
      "epoch": 1.0090854929997022,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.015134048257372653,
      "loss": 0.7981,
      "num_input_tokens_seen": 3929792,
      "step": 6775
    },
    {
      "epoch": 1.0098302055406614,
      "grad_norm": 0.060546875,
      "learning_rate": 0.015145218945487041,
      "loss": 0.7922,
      "num_input_tokens_seen": 3932768,
      "step": 6780
    },
    {
      "epoch": 1.0105749180816206,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.01515638963360143,
      "loss": 0.8037,
      "num_input_tokens_seen": 3935456,
      "step": 6785
    },
    {
      "epoch": 1.0113196306225798,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.015167560321715817,
      "loss": 0.8015,
      "num_input_tokens_seen": 3938208,
      "step": 6790
    },
    {
      "epoch": 1.0120643431635388,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.015178731009830205,
      "loss": 0.8209,
      "num_input_tokens_seen": 3941088,
      "step": 6795
    },
    {
      "epoch": 1.012809055704498,
      "grad_norm": 0.041015625,
      "learning_rate": 0.015189901697944593,
      "loss": 0.8025,
      "num_input_tokens_seen": 3943904,
      "step": 6800
    },
    {
      "epoch": 1.0135537682454572,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.015201072386058981,
      "loss": 0.7822,
      "num_input_tokens_seen": 3946752,
      "step": 6805
    },
    {
      "epoch": 1.0142984807864164,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.015212243074173369,
      "loss": 0.7825,
      "num_input_tokens_seen": 3949632,
      "step": 6810
    },
    {
      "epoch": 1.0150431933273756,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.015223413762287757,
      "loss": 0.8204,
      "num_input_tokens_seen": 3952864,
      "step": 6815
    },
    {
      "epoch": 1.0157879058683348,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.015234584450402145,
      "loss": 0.7648,
      "num_input_tokens_seen": 3956032,
      "step": 6820
    },
    {
      "epoch": 1.016532618409294,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.015245755138516533,
      "loss": 0.7856,
      "num_input_tokens_seen": 3958944,
      "step": 6825
    },
    {
      "epoch": 1.0172773309502532,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.015256925826630921,
      "loss": 0.8415,
      "num_input_tokens_seen": 3961984,
      "step": 6830
    },
    {
      "epoch": 1.0180220434912124,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.015268096514745309,
      "loss": 0.8105,
      "num_input_tokens_seen": 3964768,
      "step": 6835
    },
    {
      "epoch": 1.0187667560321716,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.015279267202859697,
      "loss": 0.809,
      "num_input_tokens_seen": 3967584,
      "step": 6840
    },
    {
      "epoch": 1.0195114685731308,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.015290437890974081,
      "loss": 0.8168,
      "num_input_tokens_seen": 3970336,
      "step": 6845
    },
    {
      "epoch": 1.02025618111409,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.01530160857908847,
      "loss": 0.7899,
      "num_input_tokens_seen": 3972896,
      "step": 6850
    },
    {
      "epoch": 1.0210008936550492,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.015312779267202857,
      "loss": 0.818,
      "num_input_tokens_seen": 3975808,
      "step": 6855
    },
    {
      "epoch": 1.0217456061960084,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.015323949955317245,
      "loss": 0.8074,
      "num_input_tokens_seen": 3978688,
      "step": 6860
    },
    {
      "epoch": 1.0224903187369676,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.015335120643431633,
      "loss": 0.8066,
      "num_input_tokens_seen": 3981408,
      "step": 6865
    },
    {
      "epoch": 1.0232350312779268,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.015346291331546021,
      "loss": 0.823,
      "num_input_tokens_seen": 3984384,
      "step": 6870
    },
    {
      "epoch": 1.023979743818886,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01535746201966041,
      "loss": 0.813,
      "num_input_tokens_seen": 3987040,
      "step": 6875
    },
    {
      "epoch": 1.024724456359845,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.015368632707774797,
      "loss": 0.7965,
      "num_input_tokens_seen": 3990016,
      "step": 6880
    },
    {
      "epoch": 1.0254691689008042,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.015379803395889187,
      "loss": 0.7969,
      "num_input_tokens_seen": 3993248,
      "step": 6885
    },
    {
      "epoch": 1.0262138814417634,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.015390974084003575,
      "loss": 0.8033,
      "num_input_tokens_seen": 3996256,
      "step": 6890
    },
    {
      "epoch": 1.0269585939827226,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.015402144772117963,
      "loss": 0.8123,
      "num_input_tokens_seen": 3999360,
      "step": 6895
    },
    {
      "epoch": 1.0277033065236818,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.01541331546023235,
      "loss": 0.8015,
      "num_input_tokens_seen": 4002400,
      "step": 6900
    },
    {
      "epoch": 1.028448019064641,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.015424486148346739,
      "loss": 0.8289,
      "num_input_tokens_seen": 4005664,
      "step": 6905
    },
    {
      "epoch": 1.0291927316056002,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.015435656836461127,
      "loss": 0.8076,
      "num_input_tokens_seen": 4008736,
      "step": 6910
    },
    {
      "epoch": 1.0299374441465594,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.015446827524575515,
      "loss": 0.8024,
      "num_input_tokens_seen": 4011616,
      "step": 6915
    },
    {
      "epoch": 1.0306821566875186,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.015457998212689903,
      "loss": 0.8047,
      "num_input_tokens_seen": 4014272,
      "step": 6920
    },
    {
      "epoch": 1.0314268692284778,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.01546916890080429,
      "loss": 0.7993,
      "num_input_tokens_seen": 4017248,
      "step": 6925
    },
    {
      "epoch": 1.032171581769437,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.015480339588918679,
      "loss": 0.8188,
      "num_input_tokens_seen": 4019680,
      "step": 6930
    },
    {
      "epoch": 1.0329162943103962,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.015491510277033063,
      "loss": 0.8175,
      "num_input_tokens_seen": 4022336,
      "step": 6935
    },
    {
      "epoch": 1.0336610068513554,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.015502680965147451,
      "loss": 0.8063,
      "num_input_tokens_seen": 4025216,
      "step": 6940
    },
    {
      "epoch": 1.0344057193923146,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.015513851653261839,
      "loss": 0.7984,
      "num_input_tokens_seen": 4028064,
      "step": 6945
    },
    {
      "epoch": 1.0351504319332738,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.015525022341376227,
      "loss": 0.7955,
      "num_input_tokens_seen": 4030912,
      "step": 6950
    },
    {
      "epoch": 1.035895144474233,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.015536193029490615,
      "loss": 0.7933,
      "num_input_tokens_seen": 4033952,
      "step": 6955
    },
    {
      "epoch": 1.0366398570151922,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.015547363717605003,
      "loss": 0.8143,
      "num_input_tokens_seen": 4036544,
      "step": 6960
    },
    {
      "epoch": 1.0373845695561514,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.015558534405719391,
      "loss": 0.7787,
      "num_input_tokens_seen": 4039296,
      "step": 6965
    },
    {
      "epoch": 1.0381292820971104,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.015569705093833779,
      "loss": 0.8042,
      "num_input_tokens_seen": 4042400,
      "step": 6970
    },
    {
      "epoch": 1.0388739946380696,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.015580875781948167,
      "loss": 0.8285,
      "num_input_tokens_seen": 4045152,
      "step": 6975
    },
    {
      "epoch": 1.0396187071790288,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.015592046470062555,
      "loss": 0.7893,
      "num_input_tokens_seen": 4048032,
      "step": 6980
    },
    {
      "epoch": 1.040363419719988,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.015603217158176943,
      "loss": 0.8091,
      "num_input_tokens_seen": 4050752,
      "step": 6985
    },
    {
      "epoch": 1.0411081322609472,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.01561438784629133,
      "loss": 0.8025,
      "num_input_tokens_seen": 4053600,
      "step": 6990
    },
    {
      "epoch": 1.0418528448019064,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.01562555853440572,
      "loss": 0.798,
      "num_input_tokens_seen": 4056608,
      "step": 6995
    },
    {
      "epoch": 1.0425975573428656,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.01563672922252011,
      "loss": 0.8131,
      "num_input_tokens_seen": 4059616,
      "step": 7000
    },
    {
      "epoch": 1.0433422698838248,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.015647899910634495,
      "loss": 0.8044,
      "num_input_tokens_seen": 4062688,
      "step": 7005
    },
    {
      "epoch": 1.044086982424784,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.015659070598748884,
      "loss": 0.8073,
      "num_input_tokens_seen": 4065408,
      "step": 7010
    },
    {
      "epoch": 1.0448316949657432,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01567024128686327,
      "loss": 0.7934,
      "num_input_tokens_seen": 4068288,
      "step": 7015
    },
    {
      "epoch": 1.0455764075067024,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.015681411974977657,
      "loss": 0.8063,
      "num_input_tokens_seen": 4070944,
      "step": 7020
    },
    {
      "epoch": 1.0463211200476616,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.015692582663092043,
      "loss": 0.7997,
      "num_input_tokens_seen": 4073696,
      "step": 7025
    },
    {
      "epoch": 1.0470658325886208,
      "grad_norm": 0.033203125,
      "learning_rate": 0.015703753351206433,
      "loss": 0.795,
      "num_input_tokens_seen": 4076576,
      "step": 7030
    },
    {
      "epoch": 1.04781054512958,
      "grad_norm": 0.068359375,
      "learning_rate": 0.01571492403932082,
      "loss": 0.786,
      "num_input_tokens_seen": 4079136,
      "step": 7035
    },
    {
      "epoch": 1.0485552576705393,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.01572609472743521,
      "loss": 0.818,
      "num_input_tokens_seen": 4082208,
      "step": 7040
    },
    {
      "epoch": 1.0492999702114985,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.015737265415549595,
      "loss": 0.8051,
      "num_input_tokens_seen": 4085120,
      "step": 7045
    },
    {
      "epoch": 1.0500446827524577,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.015748436103663985,
      "loss": 0.8095,
      "num_input_tokens_seen": 4087712,
      "step": 7050
    },
    {
      "epoch": 1.0507893952934166,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.015759606791778374,
      "loss": 0.8034,
      "num_input_tokens_seen": 4090464,
      "step": 7055
    },
    {
      "epoch": 1.0515341078343758,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.01577077747989276,
      "loss": 0.8131,
      "num_input_tokens_seen": 4093184,
      "step": 7060
    },
    {
      "epoch": 1.052278820375335,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.01578194816800715,
      "loss": 0.8109,
      "num_input_tokens_seen": 4095968,
      "step": 7065
    },
    {
      "epoch": 1.0530235329162942,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.015793118856121537,
      "loss": 0.787,
      "num_input_tokens_seen": 4098912,
      "step": 7070
    },
    {
      "epoch": 1.0537682454572534,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.015804289544235926,
      "loss": 0.7734,
      "num_input_tokens_seen": 4101696,
      "step": 7075
    },
    {
      "epoch": 1.0545129579982127,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.015815460232350313,
      "loss": 0.8148,
      "num_input_tokens_seen": 4104736,
      "step": 7080
    },
    {
      "epoch": 1.0552576705391719,
      "grad_norm": 0.05078125,
      "learning_rate": 0.015826630920464702,
      "loss": 0.8221,
      "num_input_tokens_seen": 4107776,
      "step": 7085
    },
    {
      "epoch": 1.056002383080131,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01583780160857909,
      "loss": 0.8062,
      "num_input_tokens_seen": 4110592,
      "step": 7090
    },
    {
      "epoch": 1.0567470956210903,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.015848972296693478,
      "loss": 0.8328,
      "num_input_tokens_seen": 4113568,
      "step": 7095
    },
    {
      "epoch": 1.0574918081620495,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.015860142984807864,
      "loss": 0.7884,
      "num_input_tokens_seen": 4116608,
      "step": 7100
    },
    {
      "epoch": 1.0582365207030087,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01587131367292225,
      "loss": 0.8045,
      "num_input_tokens_seen": 4119712,
      "step": 7105
    },
    {
      "epoch": 1.0589812332439679,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.015882484361036637,
      "loss": 0.7895,
      "num_input_tokens_seen": 4122336,
      "step": 7110
    },
    {
      "epoch": 1.059725945784927,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.015893655049151027,
      "loss": 0.8097,
      "num_input_tokens_seen": 4125088,
      "step": 7115
    },
    {
      "epoch": 1.0604706583258863,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.015904825737265413,
      "loss": 0.7902,
      "num_input_tokens_seen": 4128448,
      "step": 7120
    },
    {
      "epoch": 1.0612153708668455,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.015915996425379803,
      "loss": 0.8035,
      "num_input_tokens_seen": 4131520,
      "step": 7125
    },
    {
      "epoch": 1.0619600834078047,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.01592716711349419,
      "loss": 0.8241,
      "num_input_tokens_seen": 4134944,
      "step": 7130
    },
    {
      "epoch": 1.0627047959487639,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.01593833780160858,
      "loss": 0.8201,
      "num_input_tokens_seen": 4138016,
      "step": 7135
    },
    {
      "epoch": 1.063449508489723,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.015949508489722965,
      "loss": 0.8077,
      "num_input_tokens_seen": 4140896,
      "step": 7140
    },
    {
      "epoch": 1.064194221030682,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.015960679177837354,
      "loss": 0.7862,
      "num_input_tokens_seen": 4143584,
      "step": 7145
    },
    {
      "epoch": 1.0649389335716413,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.01597184986595174,
      "loss": 0.7973,
      "num_input_tokens_seen": 4146528,
      "step": 7150
    },
    {
      "epoch": 1.0656836461126005,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.01598302055406613,
      "loss": 0.8095,
      "num_input_tokens_seen": 4149536,
      "step": 7155
    },
    {
      "epoch": 1.0664283586535597,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.015994191242180517,
      "loss": 0.8007,
      "num_input_tokens_seen": 4152384,
      "step": 7160
    },
    {
      "epoch": 1.0671730711945189,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.016005361930294906,
      "loss": 0.7997,
      "num_input_tokens_seen": 4155200,
      "step": 7165
    },
    {
      "epoch": 1.067917783735478,
      "grad_norm": 0.033203125,
      "learning_rate": 0.016016532618409296,
      "loss": 0.7964,
      "num_input_tokens_seen": 4158112,
      "step": 7170
    },
    {
      "epoch": 1.0686624962764373,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.016027703306523682,
      "loss": 0.7972,
      "num_input_tokens_seen": 4160768,
      "step": 7175
    },
    {
      "epoch": 1.0694072088173965,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.016038873994638072,
      "loss": 0.8107,
      "num_input_tokens_seen": 4163872,
      "step": 7180
    },
    {
      "epoch": 1.0701519213583557,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.016050044682752458,
      "loss": 0.8276,
      "num_input_tokens_seen": 4166816,
      "step": 7185
    },
    {
      "epoch": 1.0708966338993149,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.016061215370866848,
      "loss": 0.7928,
      "num_input_tokens_seen": 4169568,
      "step": 7190
    },
    {
      "epoch": 1.071641346440274,
      "grad_norm": 0.0390625,
      "learning_rate": 0.01607238605898123,
      "loss": 0.8006,
      "num_input_tokens_seen": 4172480,
      "step": 7195
    },
    {
      "epoch": 1.0723860589812333,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01608355674709562,
      "loss": 0.8227,
      "num_input_tokens_seen": 4175744,
      "step": 7200
    },
    {
      "epoch": 1.0731307715221925,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.016094727435210007,
      "loss": 0.7819,
      "num_input_tokens_seen": 4178528,
      "step": 7205
    },
    {
      "epoch": 1.0738754840631517,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.016105898123324396,
      "loss": 0.8053,
      "num_input_tokens_seen": 4181728,
      "step": 7210
    },
    {
      "epoch": 1.074620196604111,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.016117068811438783,
      "loss": 0.7897,
      "num_input_tokens_seen": 4184320,
      "step": 7215
    },
    {
      "epoch": 1.07536490914507,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.016128239499553172,
      "loss": 0.7937,
      "num_input_tokens_seen": 4187040,
      "step": 7220
    },
    {
      "epoch": 1.076109621686029,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.01613941018766756,
      "loss": 0.8075,
      "num_input_tokens_seen": 4190144,
      "step": 7225
    },
    {
      "epoch": 1.0768543342269883,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.016150580875781948,
      "loss": 0.8057,
      "num_input_tokens_seen": 4193152,
      "step": 7230
    },
    {
      "epoch": 1.0775990467679475,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.016161751563896334,
      "loss": 0.7825,
      "num_input_tokens_seen": 4196000,
      "step": 7235
    },
    {
      "epoch": 1.0783437593089067,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.016172922252010724,
      "loss": 0.8082,
      "num_input_tokens_seen": 4198880,
      "step": 7240
    },
    {
      "epoch": 1.079088471849866,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.01618409294012511,
      "loss": 0.7915,
      "num_input_tokens_seen": 4201664,
      "step": 7245
    },
    {
      "epoch": 1.079833184390825,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0161952636282395,
      "loss": 0.8199,
      "num_input_tokens_seen": 4204672,
      "step": 7250
    },
    {
      "epoch": 1.0805778969317843,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.016206434316353886,
      "loss": 0.8064,
      "num_input_tokens_seen": 4207552,
      "step": 7255
    },
    {
      "epoch": 1.0813226094727435,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.016217605004468276,
      "loss": 0.8127,
      "num_input_tokens_seen": 4210336,
      "step": 7260
    },
    {
      "epoch": 1.0820673220137027,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.016228775692582662,
      "loss": 0.7895,
      "num_input_tokens_seen": 4212992,
      "step": 7265
    },
    {
      "epoch": 1.082812034554662,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.016239946380697052,
      "loss": 0.8101,
      "num_input_tokens_seen": 4215712,
      "step": 7270
    },
    {
      "epoch": 1.083556747095621,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.016251117068811438,
      "loss": 0.7959,
      "num_input_tokens_seen": 4219072,
      "step": 7275
    },
    {
      "epoch": 1.0843014596365803,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.016262287756925824,
      "loss": 0.8017,
      "num_input_tokens_seen": 4221856,
      "step": 7280
    },
    {
      "epoch": 1.0850461721775395,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.016273458445040214,
      "loss": 0.8312,
      "num_input_tokens_seen": 4224896,
      "step": 7285
    },
    {
      "epoch": 1.0857908847184987,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0162846291331546,
      "loss": 0.8059,
      "num_input_tokens_seen": 4227680,
      "step": 7290
    },
    {
      "epoch": 1.086535597259458,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.01629579982126899,
      "loss": 0.7765,
      "num_input_tokens_seen": 4230432,
      "step": 7295
    },
    {
      "epoch": 1.0872803098004171,
      "grad_norm": 1.3046875,
      "learning_rate": 0.016306970509383376,
      "loss": 0.9055,
      "num_input_tokens_seen": 4233632,
      "step": 7300
    },
    {
      "epoch": 1.0880250223413763,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.016318141197497766,
      "loss": 0.8277,
      "num_input_tokens_seen": 4237024,
      "step": 7305
    },
    {
      "epoch": 1.0887697348823355,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.016329311885612152,
      "loss": 0.808,
      "num_input_tokens_seen": 4240320,
      "step": 7310
    },
    {
      "epoch": 1.0895144474232945,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.016340482573726542,
      "loss": 0.7882,
      "num_input_tokens_seen": 4242976,
      "step": 7315
    },
    {
      "epoch": 1.0902591599642537,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.016351653261840928,
      "loss": 0.7969,
      "num_input_tokens_seen": 4245792,
      "step": 7320
    },
    {
      "epoch": 1.091003872505213,
      "grad_norm": 0.056640625,
      "learning_rate": 0.016362823949955318,
      "loss": 0.7941,
      "num_input_tokens_seen": 4248608,
      "step": 7325
    },
    {
      "epoch": 1.0917485850461721,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.016373994638069704,
      "loss": 0.7962,
      "num_input_tokens_seen": 4251616,
      "step": 7330
    },
    {
      "epoch": 1.0924932975871313,
      "grad_norm": 0.150390625,
      "learning_rate": 0.016385165326184094,
      "loss": 0.8012,
      "num_input_tokens_seen": 4254720,
      "step": 7335
    },
    {
      "epoch": 1.0932380101280905,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.01639633601429848,
      "loss": 0.8109,
      "num_input_tokens_seen": 4257568,
      "step": 7340
    },
    {
      "epoch": 1.0939827226690497,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.01640750670241287,
      "loss": 0.822,
      "num_input_tokens_seen": 4260384,
      "step": 7345
    },
    {
      "epoch": 1.094727435210009,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.016418677390527256,
      "loss": 0.8082,
      "num_input_tokens_seen": 4263616,
      "step": 7350
    },
    {
      "epoch": 1.0954721477509681,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.016429848078641646,
      "loss": 0.8062,
      "num_input_tokens_seen": 4266368,
      "step": 7355
    },
    {
      "epoch": 1.0962168602919273,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.016441018766756032,
      "loss": 0.8016,
      "num_input_tokens_seen": 4269120,
      "step": 7360
    },
    {
      "epoch": 1.0969615728328865,
      "grad_norm": 0.07421875,
      "learning_rate": 0.016452189454870418,
      "loss": 0.8172,
      "num_input_tokens_seen": 4272064,
      "step": 7365
    },
    {
      "epoch": 1.0977062853738457,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.016463360142984804,
      "loss": 0.7935,
      "num_input_tokens_seen": 4275072,
      "step": 7370
    },
    {
      "epoch": 1.098450997914805,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.016474530831099194,
      "loss": 0.8096,
      "num_input_tokens_seen": 4278240,
      "step": 7375
    },
    {
      "epoch": 1.0991957104557641,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.01648570151921358,
      "loss": 0.7946,
      "num_input_tokens_seen": 4281184,
      "step": 7380
    },
    {
      "epoch": 1.0999404229967233,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.01649687220732797,
      "loss": 0.8011,
      "num_input_tokens_seen": 4284032,
      "step": 7385
    },
    {
      "epoch": 1.1006851355376825,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.01650804289544236,
      "loss": 0.7974,
      "num_input_tokens_seen": 4286976,
      "step": 7390
    },
    {
      "epoch": 1.1014298480786415,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.016519213583556746,
      "loss": 0.7994,
      "num_input_tokens_seen": 4289920,
      "step": 7395
    },
    {
      "epoch": 1.1021745606196007,
      "grad_norm": 0.189453125,
      "learning_rate": 0.016530384271671136,
      "loss": 0.8047,
      "num_input_tokens_seen": 4292768,
      "step": 7400
    },
    {
      "epoch": 1.10291927316056,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.016541554959785522,
      "loss": 0.753,
      "num_input_tokens_seen": 4295840,
      "step": 7405
    },
    {
      "epoch": 1.1036639857015191,
      "grad_norm": 0.1328125,
      "learning_rate": 0.01655272564789991,
      "loss": 0.7888,
      "num_input_tokens_seen": 4298976,
      "step": 7410
    },
    {
      "epoch": 1.1044086982424783,
      "grad_norm": 0.265625,
      "learning_rate": 0.016563896336014298,
      "loss": 0.7777,
      "num_input_tokens_seen": 4302016,
      "step": 7415
    },
    {
      "epoch": 1.1051534107834375,
      "grad_norm": 0.076171875,
      "learning_rate": 0.016575067024128688,
      "loss": 0.8144,
      "num_input_tokens_seen": 4305216,
      "step": 7420
    },
    {
      "epoch": 1.1058981233243967,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.016586237712243074,
      "loss": 0.8113,
      "num_input_tokens_seen": 4308032,
      "step": 7425
    },
    {
      "epoch": 1.106642835865356,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.016597408400357463,
      "loss": 0.8295,
      "num_input_tokens_seen": 4310752,
      "step": 7430
    },
    {
      "epoch": 1.1073875484063151,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.01660857908847185,
      "loss": 0.8187,
      "num_input_tokens_seen": 4313344,
      "step": 7435
    },
    {
      "epoch": 1.1081322609472744,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.01661974977658624,
      "loss": 0.8004,
      "num_input_tokens_seen": 4316000,
      "step": 7440
    },
    {
      "epoch": 1.1088769734882336,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.016630920464700626,
      "loss": 0.8225,
      "num_input_tokens_seen": 4318688,
      "step": 7445
    },
    {
      "epoch": 1.1096216860291928,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.016642091152815015,
      "loss": 0.7985,
      "num_input_tokens_seen": 4321440,
      "step": 7450
    },
    {
      "epoch": 1.110366398570152,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.016653261840929398,
      "loss": 0.8119,
      "num_input_tokens_seen": 4324224,
      "step": 7455
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.016664432529043788,
      "loss": 0.8032,
      "num_input_tokens_seen": 4327008,
      "step": 7460
    },
    {
      "epoch": 1.1118558236520704,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.016675603217158174,
      "loss": 0.795,
      "num_input_tokens_seen": 4330112,
      "step": 7465
    },
    {
      "epoch": 1.1126005361930296,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.016686773905272564,
      "loss": 0.781,
      "num_input_tokens_seen": 4332992,
      "step": 7470
    },
    {
      "epoch": 1.1133452487339888,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.01669794459338695,
      "loss": 0.8071,
      "num_input_tokens_seen": 4336128,
      "step": 7475
    },
    {
      "epoch": 1.114089961274948,
      "grad_norm": 0.068359375,
      "learning_rate": 0.01670911528150134,
      "loss": 0.8279,
      "num_input_tokens_seen": 4339040,
      "step": 7480
    },
    {
      "epoch": 1.1148346738159072,
      "grad_norm": 0.068359375,
      "learning_rate": 0.016720285969615726,
      "loss": 0.8166,
      "num_input_tokens_seen": 4341920,
      "step": 7485
    },
    {
      "epoch": 1.1155793863568662,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.016731456657730116,
      "loss": 0.8095,
      "num_input_tokens_seen": 4344960,
      "step": 7490
    },
    {
      "epoch": 1.1163240988978254,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.016742627345844502,
      "loss": 0.8099,
      "num_input_tokens_seen": 4347936,
      "step": 7495
    },
    {
      "epoch": 1.1170688114387846,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.01675379803395889,
      "loss": 0.8128,
      "num_input_tokens_seen": 4350688,
      "step": 7500
    },
    {
      "epoch": 1.1178135239797438,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.01676496872207328,
      "loss": 0.8191,
      "num_input_tokens_seen": 4353664,
      "step": 7505
    },
    {
      "epoch": 1.118558236520703,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.016776139410187667,
      "loss": 0.8198,
      "num_input_tokens_seen": 4356736,
      "step": 7510
    },
    {
      "epoch": 1.1193029490616622,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.016787310098302057,
      "loss": 0.8036,
      "num_input_tokens_seen": 4359808,
      "step": 7515
    },
    {
      "epoch": 1.1200476616026214,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.016798480786416443,
      "loss": 0.8121,
      "num_input_tokens_seen": 4362752,
      "step": 7520
    },
    {
      "epoch": 1.1207923741435806,
      "grad_norm": 0.05859375,
      "learning_rate": 0.016809651474530833,
      "loss": 0.8064,
      "num_input_tokens_seen": 4365632,
      "step": 7525
    },
    {
      "epoch": 1.1215370866845398,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01682082216264522,
      "loss": 0.82,
      "num_input_tokens_seen": 4368512,
      "step": 7530
    },
    {
      "epoch": 1.122281799225499,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.01683199285075961,
      "loss": 0.8132,
      "num_input_tokens_seen": 4371584,
      "step": 7535
    },
    {
      "epoch": 1.1230265117664582,
      "grad_norm": 0.046875,
      "learning_rate": 0.016843163538873992,
      "loss": 0.8021,
      "num_input_tokens_seen": 4374336,
      "step": 7540
    },
    {
      "epoch": 1.1237712243074174,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.01685433422698838,
      "loss": 0.8075,
      "num_input_tokens_seen": 4377312,
      "step": 7545
    },
    {
      "epoch": 1.1245159368483766,
      "grad_norm": 0.037109375,
      "learning_rate": 0.016865504915102768,
      "loss": 0.8063,
      "num_input_tokens_seen": 4380064,
      "step": 7550
    },
    {
      "epoch": 1.1252606493893358,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.016876675603217157,
      "loss": 0.7939,
      "num_input_tokens_seen": 4382816,
      "step": 7555
    },
    {
      "epoch": 1.126005361930295,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.016887846291331544,
      "loss": 0.7972,
      "num_input_tokens_seen": 4385568,
      "step": 7560
    },
    {
      "epoch": 1.1267500744712542,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.016899016979445933,
      "loss": 0.8083,
      "num_input_tokens_seen": 4388608,
      "step": 7565
    },
    {
      "epoch": 1.1274947870122132,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.01691018766756032,
      "loss": 0.8395,
      "num_input_tokens_seen": 4391744,
      "step": 7570
    },
    {
      "epoch": 1.1282394995531724,
      "grad_norm": 0.044921875,
      "learning_rate": 0.01692135835567471,
      "loss": 0.8066,
      "num_input_tokens_seen": 4394688,
      "step": 7575
    },
    {
      "epoch": 1.1289842120941316,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.016932529043789096,
      "loss": 0.807,
      "num_input_tokens_seen": 4397696,
      "step": 7580
    },
    {
      "epoch": 1.1297289246350908,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.016943699731903485,
      "loss": 0.8061,
      "num_input_tokens_seen": 4400608,
      "step": 7585
    },
    {
      "epoch": 1.13047363717605,
      "grad_norm": 0.0390625,
      "learning_rate": 0.01695487042001787,
      "loss": 0.8005,
      "num_input_tokens_seen": 4403552,
      "step": 7590
    },
    {
      "epoch": 1.1312183497170092,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01696604110813226,
      "loss": 0.8161,
      "num_input_tokens_seen": 4406496,
      "step": 7595
    },
    {
      "epoch": 1.1319630622579684,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.016977211796246647,
      "loss": 0.7926,
      "num_input_tokens_seen": 4409120,
      "step": 7600
    },
    {
      "epoch": 1.1327077747989276,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.016988382484361037,
      "loss": 0.7986,
      "num_input_tokens_seen": 4411744,
      "step": 7605
    },
    {
      "epoch": 1.1334524873398868,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.016999553172475423,
      "loss": 0.8287,
      "num_input_tokens_seen": 4414720,
      "step": 7610
    },
    {
      "epoch": 1.134197199880846,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.017010723860589813,
      "loss": 0.8044,
      "num_input_tokens_seen": 4417472,
      "step": 7615
    },
    {
      "epoch": 1.1349419124218052,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0170218945487042,
      "loss": 0.825,
      "num_input_tokens_seen": 4420640,
      "step": 7620
    },
    {
      "epoch": 1.1356866249627644,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.017033065236818586,
      "loss": 0.7904,
      "num_input_tokens_seen": 4423520,
      "step": 7625
    },
    {
      "epoch": 1.1364313375037236,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.017044235924932975,
      "loss": 0.8013,
      "num_input_tokens_seen": 4426496,
      "step": 7630
    },
    {
      "epoch": 1.1371760500446828,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.01705540661304736,
      "loss": 0.8184,
      "num_input_tokens_seen": 4429440,
      "step": 7635
    },
    {
      "epoch": 1.137920762585642,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01706657730116175,
      "loss": 0.8104,
      "num_input_tokens_seen": 4432320,
      "step": 7640
    },
    {
      "epoch": 1.1386654751266012,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.017077747989276137,
      "loss": 0.8065,
      "num_input_tokens_seen": 4435584,
      "step": 7645
    },
    {
      "epoch": 1.1394101876675604,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.017088918677390527,
      "loss": 0.8009,
      "num_input_tokens_seen": 4438464,
      "step": 7650
    },
    {
      "epoch": 1.1401549002085196,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.017100089365504913,
      "loss": 0.7967,
      "num_input_tokens_seen": 4441120,
      "step": 7655
    },
    {
      "epoch": 1.1408996127494788,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.017111260053619303,
      "loss": 0.7947,
      "num_input_tokens_seen": 4443680,
      "step": 7660
    },
    {
      "epoch": 1.1416443252904378,
      "grad_norm": 0.044921875,
      "learning_rate": 0.01712243074173369,
      "loss": 0.8023,
      "num_input_tokens_seen": 4446528,
      "step": 7665
    },
    {
      "epoch": 1.142389037831397,
      "grad_norm": 0.05078125,
      "learning_rate": 0.01713360142984808,
      "loss": 0.8091,
      "num_input_tokens_seen": 4449504,
      "step": 7670
    },
    {
      "epoch": 1.1431337503723562,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.017144772117962465,
      "loss": 0.8177,
      "num_input_tokens_seen": 4452352,
      "step": 7675
    },
    {
      "epoch": 1.1438784629133154,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.017155942806076855,
      "loss": 0.8242,
      "num_input_tokens_seen": 4455136,
      "step": 7680
    },
    {
      "epoch": 1.1446231754542746,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.01716711349419124,
      "loss": 0.8139,
      "num_input_tokens_seen": 4458144,
      "step": 7685
    },
    {
      "epoch": 1.1453678879952338,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.01717828418230563,
      "loss": 0.8093,
      "num_input_tokens_seen": 4460800,
      "step": 7690
    },
    {
      "epoch": 1.146112600536193,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.017189454870420017,
      "loss": 0.7968,
      "num_input_tokens_seen": 4463904,
      "step": 7695
    },
    {
      "epoch": 1.1468573130771522,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.017200625558534407,
      "loss": 0.8186,
      "num_input_tokens_seen": 4466816,
      "step": 7700
    },
    {
      "epoch": 1.1476020256181114,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.017211796246648793,
      "loss": 0.8187,
      "num_input_tokens_seen": 4469504,
      "step": 7705
    },
    {
      "epoch": 1.1483467381590706,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.01722296693476318,
      "loss": 0.7986,
      "num_input_tokens_seen": 4472736,
      "step": 7710
    },
    {
      "epoch": 1.1490914507000298,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.017234137622877566,
      "loss": 0.8093,
      "num_input_tokens_seen": 4475552,
      "step": 7715
    },
    {
      "epoch": 1.149836163240989,
      "grad_norm": 0.0390625,
      "learning_rate": 0.017245308310991955,
      "loss": 0.8104,
      "num_input_tokens_seen": 4478432,
      "step": 7720
    },
    {
      "epoch": 1.1505808757819482,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.01725647899910634,
      "loss": 0.7934,
      "num_input_tokens_seen": 4481312,
      "step": 7725
    },
    {
      "epoch": 1.1513255883229074,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.01726764968722073,
      "loss": 0.801,
      "num_input_tokens_seen": 4484608,
      "step": 7730
    },
    {
      "epoch": 1.1520703008638666,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01727882037533512,
      "loss": 0.7918,
      "num_input_tokens_seen": 4487808,
      "step": 7735
    },
    {
      "epoch": 1.1528150134048256,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.017289991063449507,
      "loss": 0.7871,
      "num_input_tokens_seen": 4490688,
      "step": 7740
    },
    {
      "epoch": 1.1535597259457848,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.017301161751563897,
      "loss": 0.8181,
      "num_input_tokens_seen": 4493920,
      "step": 7745
    },
    {
      "epoch": 1.154304438486744,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.017312332439678283,
      "loss": 0.8145,
      "num_input_tokens_seen": 4496768,
      "step": 7750
    },
    {
      "epoch": 1.1550491510277032,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.017323503127792673,
      "loss": 0.7917,
      "num_input_tokens_seen": 4499840,
      "step": 7755
    },
    {
      "epoch": 1.1557938635686624,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.01733467381590706,
      "loss": 0.8015,
      "num_input_tokens_seen": 4502912,
      "step": 7760
    },
    {
      "epoch": 1.1565385761096216,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01734584450402145,
      "loss": 0.7961,
      "num_input_tokens_seen": 4505856,
      "step": 7765
    },
    {
      "epoch": 1.1572832886505808,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.017357015192135835,
      "loss": 0.8117,
      "num_input_tokens_seen": 4508928,
      "step": 7770
    },
    {
      "epoch": 1.15802800119154,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.017368185880250225,
      "loss": 0.786,
      "num_input_tokens_seen": 4512064,
      "step": 7775
    },
    {
      "epoch": 1.1587727137324992,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01737935656836461,
      "loss": 0.7695,
      "num_input_tokens_seen": 4514880,
      "step": 7780
    },
    {
      "epoch": 1.1595174262734584,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.017390527256479,
      "loss": 0.8151,
      "num_input_tokens_seen": 4517632,
      "step": 7785
    },
    {
      "epoch": 1.1602621388144176,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.017401697944593387,
      "loss": 0.8238,
      "num_input_tokens_seen": 4520416,
      "step": 7790
    },
    {
      "epoch": 1.1610068513553768,
      "grad_norm": 0.03125,
      "learning_rate": 0.017412868632707777,
      "loss": 0.8142,
      "num_input_tokens_seen": 4523296,
      "step": 7795
    },
    {
      "epoch": 1.161751563896336,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.01742403932082216,
      "loss": 0.772,
      "num_input_tokens_seen": 4526880,
      "step": 7800
    },
    {
      "epoch": 1.1624962764372953,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.01743521000893655,
      "loss": 0.7983,
      "num_input_tokens_seen": 4529504,
      "step": 7805
    },
    {
      "epoch": 1.1632409889782545,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.017446380697050935,
      "loss": 0.7803,
      "num_input_tokens_seen": 4532320,
      "step": 7810
    },
    {
      "epoch": 1.1639857015192137,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.017457551385165325,
      "loss": 0.798,
      "num_input_tokens_seen": 4535104,
      "step": 7815
    },
    {
      "epoch": 1.1647304140601729,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01746872207327971,
      "loss": 0.813,
      "num_input_tokens_seen": 4537600,
      "step": 7820
    },
    {
      "epoch": 1.165475126601132,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0174798927613941,
      "loss": 0.8204,
      "num_input_tokens_seen": 4540512,
      "step": 7825
    },
    {
      "epoch": 1.1662198391420913,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.017491063449508487,
      "loss": 0.7994,
      "num_input_tokens_seen": 4543328,
      "step": 7830
    },
    {
      "epoch": 1.1669645516830505,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.017502234137622877,
      "loss": 0.797,
      "num_input_tokens_seen": 4546176,
      "step": 7835
    },
    {
      "epoch": 1.1677092642240094,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.017513404825737263,
      "loss": 0.8166,
      "num_input_tokens_seen": 4549344,
      "step": 7840
    },
    {
      "epoch": 1.1684539767649686,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.017524575513851653,
      "loss": 0.8188,
      "num_input_tokens_seen": 4552160,
      "step": 7845
    },
    {
      "epoch": 1.1691986893059279,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.017535746201966042,
      "loss": 0.8256,
      "num_input_tokens_seen": 4554880,
      "step": 7850
    },
    {
      "epoch": 1.169943401846887,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.01754691689008043,
      "loss": 0.7967,
      "num_input_tokens_seen": 4557696,
      "step": 7855
    },
    {
      "epoch": 1.1706881143878463,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.01755808757819482,
      "loss": 0.7922,
      "num_input_tokens_seen": 4560224,
      "step": 7860
    },
    {
      "epoch": 1.1714328269288055,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.017569258266309205,
      "loss": 0.7961,
      "num_input_tokens_seen": 4563168,
      "step": 7865
    },
    {
      "epoch": 1.1721775394697647,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.017580428954423594,
      "loss": 0.7987,
      "num_input_tokens_seen": 4565856,
      "step": 7870
    },
    {
      "epoch": 1.1729222520107239,
      "grad_norm": 0.046875,
      "learning_rate": 0.01759159964253798,
      "loss": 0.8026,
      "num_input_tokens_seen": 4569216,
      "step": 7875
    },
    {
      "epoch": 1.173666964551683,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01760277033065237,
      "loss": 0.8038,
      "num_input_tokens_seen": 4572064,
      "step": 7880
    },
    {
      "epoch": 1.1744116770926423,
      "grad_norm": 0.041015625,
      "learning_rate": 0.017613941018766753,
      "loss": 0.8147,
      "num_input_tokens_seen": 4575136,
      "step": 7885
    },
    {
      "epoch": 1.1751563896336015,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.017625111706881143,
      "loss": 0.8055,
      "num_input_tokens_seen": 4577856,
      "step": 7890
    },
    {
      "epoch": 1.1759011021745607,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01763628239499553,
      "loss": 0.8287,
      "num_input_tokens_seen": 4580608,
      "step": 7895
    },
    {
      "epoch": 1.1766458147155199,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.01764745308310992,
      "loss": 0.8042,
      "num_input_tokens_seen": 4583200,
      "step": 7900
    },
    {
      "epoch": 1.177390527256479,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.017658623771224305,
      "loss": 0.7864,
      "num_input_tokens_seen": 4585952,
      "step": 7905
    },
    {
      "epoch": 1.1781352397974383,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.017669794459338695,
      "loss": 0.8252,
      "num_input_tokens_seen": 4589056,
      "step": 7910
    },
    {
      "epoch": 1.1788799523383973,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.01768096514745308,
      "loss": 0.807,
      "num_input_tokens_seen": 4592096,
      "step": 7915
    },
    {
      "epoch": 1.1796246648793565,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.01769213583556747,
      "loss": 0.8059,
      "num_input_tokens_seen": 4594784,
      "step": 7920
    },
    {
      "epoch": 1.1803693774203157,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.017703306523681857,
      "loss": 0.8071,
      "num_input_tokens_seen": 4597856,
      "step": 7925
    },
    {
      "epoch": 1.1811140899612749,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.017714477211796247,
      "loss": 0.7971,
      "num_input_tokens_seen": 4600512,
      "step": 7930
    },
    {
      "epoch": 1.181858802502234,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.017725647899910633,
      "loss": 0.8098,
      "num_input_tokens_seen": 4603328,
      "step": 7935
    },
    {
      "epoch": 1.1826035150431933,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.017736818588025022,
      "loss": 0.7903,
      "num_input_tokens_seen": 4606144,
      "step": 7940
    },
    {
      "epoch": 1.1833482275841525,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01774798927613941,
      "loss": 0.7985,
      "num_input_tokens_seen": 4609088,
      "step": 7945
    },
    {
      "epoch": 1.1840929401251117,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0177591599642538,
      "loss": 0.8039,
      "num_input_tokens_seen": 4612160,
      "step": 7950
    },
    {
      "epoch": 1.1848376526660709,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.017770330652368185,
      "loss": 0.8053,
      "num_input_tokens_seen": 4615072,
      "step": 7955
    },
    {
      "epoch": 1.18558236520703,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.017781501340482574,
      "loss": 0.8059,
      "num_input_tokens_seen": 4617600,
      "step": 7960
    },
    {
      "epoch": 1.1863270777479893,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.017792672028596964,
      "loss": 0.7983,
      "num_input_tokens_seen": 4620320,
      "step": 7965
    },
    {
      "epoch": 1.1870717902889485,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.017803842716711347,
      "loss": 0.8106,
      "num_input_tokens_seen": 4623520,
      "step": 7970
    },
    {
      "epoch": 1.1878165028299077,
      "grad_norm": 0.0234375,
      "learning_rate": 0.017815013404825737,
      "loss": 0.8165,
      "num_input_tokens_seen": 4626656,
      "step": 7975
    },
    {
      "epoch": 1.188561215370867,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.017826184092940123,
      "loss": 0.806,
      "num_input_tokens_seen": 4629408,
      "step": 7980
    },
    {
      "epoch": 1.189305927911826,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.017837354781054512,
      "loss": 0.7847,
      "num_input_tokens_seen": 4632320,
      "step": 7985
    },
    {
      "epoch": 1.1900506404527853,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0178485254691689,
      "loss": 0.8236,
      "num_input_tokens_seen": 4635040,
      "step": 7990
    },
    {
      "epoch": 1.1907953529937445,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01785969615728329,
      "loss": 0.834,
      "num_input_tokens_seen": 4638048,
      "step": 7995
    },
    {
      "epoch": 1.1915400655347037,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.017870866845397675,
      "loss": 0.8111,
      "num_input_tokens_seen": 4640864,
      "step": 8000
    },
    {
      "epoch": 1.192284778075663,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.017882037533512064,
      "loss": 0.7837,
      "num_input_tokens_seen": 4643488,
      "step": 8005
    },
    {
      "epoch": 1.193029490616622,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.01789320822162645,
      "loss": 0.8101,
      "num_input_tokens_seen": 4646432,
      "step": 8010
    },
    {
      "epoch": 1.193774203157581,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01790437890974084,
      "loss": 0.8026,
      "num_input_tokens_seen": 4649280,
      "step": 8015
    },
    {
      "epoch": 1.1945189156985403,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.017915549597855226,
      "loss": 0.8145,
      "num_input_tokens_seen": 4652192,
      "step": 8020
    },
    {
      "epoch": 1.1952636282394995,
      "grad_norm": 0.03125,
      "learning_rate": 0.017926720285969616,
      "loss": 0.7982,
      "num_input_tokens_seen": 4654848,
      "step": 8025
    },
    {
      "epoch": 1.1960083407804587,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.017937890974084002,
      "loss": 0.8189,
      "num_input_tokens_seen": 4657600,
      "step": 8030
    },
    {
      "epoch": 1.196753053321418,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.017949061662198392,
      "loss": 0.7872,
      "num_input_tokens_seen": 4660352,
      "step": 8035
    },
    {
      "epoch": 1.197497765862377,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.01796023235031278,
      "loss": 0.7923,
      "num_input_tokens_seen": 4663296,
      "step": 8040
    },
    {
      "epoch": 1.1982424784033363,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.017971403038427168,
      "loss": 0.7954,
      "num_input_tokens_seen": 4666240,
      "step": 8045
    },
    {
      "epoch": 1.1989871909442955,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.017982573726541554,
      "loss": 0.825,
      "num_input_tokens_seen": 4668992,
      "step": 8050
    },
    {
      "epoch": 1.1997319034852547,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.017993744414655944,
      "loss": 0.811,
      "num_input_tokens_seen": 4671584,
      "step": 8055
    },
    {
      "epoch": 1.200476616026214,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.018004915102770327,
      "loss": 0.796,
      "num_input_tokens_seen": 4674144,
      "step": 8060
    },
    {
      "epoch": 1.2012213285671731,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.018016085790884716,
      "loss": 0.8021,
      "num_input_tokens_seen": 4676864,
      "step": 8065
    },
    {
      "epoch": 1.2019660411081323,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.018027256478999106,
      "loss": 0.8315,
      "num_input_tokens_seen": 4679744,
      "step": 8070
    },
    {
      "epoch": 1.2027107536490915,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.018038427167113492,
      "loss": 0.8024,
      "num_input_tokens_seen": 4682592,
      "step": 8075
    },
    {
      "epoch": 1.2034554661900507,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.018049597855227882,
      "loss": 0.818,
      "num_input_tokens_seen": 4685344,
      "step": 8080
    },
    {
      "epoch": 1.2042001787310097,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01806076854334227,
      "loss": 0.8042,
      "num_input_tokens_seen": 4688000,
      "step": 8085
    },
    {
      "epoch": 1.204944891271969,
      "grad_norm": 0.052734375,
      "learning_rate": 0.018071939231456658,
      "loss": 0.7952,
      "num_input_tokens_seen": 4691456,
      "step": 8090
    },
    {
      "epoch": 1.2056896038129281,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.018083109919571044,
      "loss": 0.8064,
      "num_input_tokens_seen": 4694432,
      "step": 8095
    },
    {
      "epoch": 1.2064343163538873,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.018094280607685434,
      "loss": 0.8013,
      "num_input_tokens_seen": 4697376,
      "step": 8100
    },
    {
      "epoch": 1.2071790288948465,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01810545129579982,
      "loss": 0.8019,
      "num_input_tokens_seen": 4700320,
      "step": 8105
    },
    {
      "epoch": 1.2079237414358057,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.01811662198391421,
      "loss": 0.8116,
      "num_input_tokens_seen": 4703232,
      "step": 8110
    },
    {
      "epoch": 1.208668453976765,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.018127792672028596,
      "loss": 0.8193,
      "num_input_tokens_seen": 4706208,
      "step": 8115
    },
    {
      "epoch": 1.2094131665177241,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.018138963360142986,
      "loss": 0.7847,
      "num_input_tokens_seen": 4709056,
      "step": 8120
    },
    {
      "epoch": 1.2101578790586833,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.018150134048257372,
      "loss": 0.8142,
      "num_input_tokens_seen": 4711904,
      "step": 8125
    },
    {
      "epoch": 1.2109025915996425,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.018161304736371762,
      "loss": 0.8016,
      "num_input_tokens_seen": 4714560,
      "step": 8130
    },
    {
      "epoch": 1.2116473041406017,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.018172475424486148,
      "loss": 0.7886,
      "num_input_tokens_seen": 4717472,
      "step": 8135
    },
    {
      "epoch": 1.212392016681561,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.018183646112600538,
      "loss": 0.8101,
      "num_input_tokens_seen": 4720288,
      "step": 8140
    },
    {
      "epoch": 1.2131367292225201,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.01819481680071492,
      "loss": 0.7955,
      "num_input_tokens_seen": 4723360,
      "step": 8145
    },
    {
      "epoch": 1.2138814417634793,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01820598748882931,
      "loss": 0.8059,
      "num_input_tokens_seen": 4726208,
      "step": 8150
    },
    {
      "epoch": 1.2146261543044385,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.018217158176943696,
      "loss": 0.8093,
      "num_input_tokens_seen": 4729024,
      "step": 8155
    },
    {
      "epoch": 1.2153708668453977,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.018228328865058086,
      "loss": 0.7906,
      "num_input_tokens_seen": 4731904,
      "step": 8160
    },
    {
      "epoch": 1.216115579386357,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.018239499553172472,
      "loss": 0.7944,
      "num_input_tokens_seen": 4734432,
      "step": 8165
    },
    {
      "epoch": 1.2168602919273162,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.018250670241286862,
      "loss": 0.7999,
      "num_input_tokens_seen": 4737216,
      "step": 8170
    },
    {
      "epoch": 1.2176050044682754,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.01826184092940125,
      "loss": 0.8191,
      "num_input_tokens_seen": 4739936,
      "step": 8175
    },
    {
      "epoch": 1.2183497170092346,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.018273011617515638,
      "loss": 0.8053,
      "num_input_tokens_seen": 4743072,
      "step": 8180
    },
    {
      "epoch": 1.2190944295501935,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.018284182305630028,
      "loss": 0.8188,
      "num_input_tokens_seen": 4745984,
      "step": 8185
    },
    {
      "epoch": 1.2198391420911527,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.018295352993744414,
      "loss": 0.7919,
      "num_input_tokens_seen": 4748896,
      "step": 8190
    },
    {
      "epoch": 1.220583854632112,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.018306523681858804,
      "loss": 0.8043,
      "num_input_tokens_seen": 4751808,
      "step": 8195
    },
    {
      "epoch": 1.2213285671730711,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01831769436997319,
      "loss": 0.8138,
      "num_input_tokens_seen": 4754496,
      "step": 8200
    },
    {
      "epoch": 1.2220732797140303,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.01832886505808758,
      "loss": 0.7999,
      "num_input_tokens_seen": 4757344,
      "step": 8205
    },
    {
      "epoch": 1.2228179922549895,
      "grad_norm": 0.07421875,
      "learning_rate": 0.018340035746201966,
      "loss": 0.7868,
      "num_input_tokens_seen": 4760480,
      "step": 8210
    },
    {
      "epoch": 1.2235627047959488,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.018351206434316356,
      "loss": 0.8184,
      "num_input_tokens_seen": 4763200,
      "step": 8215
    },
    {
      "epoch": 1.224307417336908,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.018362377122430742,
      "loss": 0.8288,
      "num_input_tokens_seen": 4766144,
      "step": 8220
    },
    {
      "epoch": 1.2250521298778672,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01837354781054513,
      "loss": 0.8345,
      "num_input_tokens_seen": 4769440,
      "step": 8225
    },
    {
      "epoch": 1.2257968424188264,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.018384718498659514,
      "loss": 0.7958,
      "num_input_tokens_seen": 4772320,
      "step": 8230
    },
    {
      "epoch": 1.2265415549597856,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.018395889186773904,
      "loss": 0.8015,
      "num_input_tokens_seen": 4775232,
      "step": 8235
    },
    {
      "epoch": 1.2272862675007448,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.01840705987488829,
      "loss": 0.7855,
      "num_input_tokens_seen": 4778368,
      "step": 8240
    },
    {
      "epoch": 1.228030980041704,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.01841823056300268,
      "loss": 0.8219,
      "num_input_tokens_seen": 4781312,
      "step": 8245
    },
    {
      "epoch": 1.2287756925826632,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.018429401251117066,
      "loss": 0.7849,
      "num_input_tokens_seen": 4784128,
      "step": 8250
    },
    {
      "epoch": 1.2295204051236224,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.018440571939231456,
      "loss": 0.7923,
      "num_input_tokens_seen": 4786976,
      "step": 8255
    },
    {
      "epoch": 1.2302651176645814,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.018451742627345842,
      "loss": 0.7982,
      "num_input_tokens_seen": 4789856,
      "step": 8260
    },
    {
      "epoch": 1.2310098302055406,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.018462913315460232,
      "loss": 0.8214,
      "num_input_tokens_seen": 4792704,
      "step": 8265
    },
    {
      "epoch": 1.2317545427464998,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.018474084003574618,
      "loss": 0.7971,
      "num_input_tokens_seen": 4795712,
      "step": 8270
    },
    {
      "epoch": 1.232499255287459,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.018485254691689008,
      "loss": 0.8228,
      "num_input_tokens_seen": 4798560,
      "step": 8275
    },
    {
      "epoch": 1.2332439678284182,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.018496425379803394,
      "loss": 0.7764,
      "num_input_tokens_seen": 4801312,
      "step": 8280
    },
    {
      "epoch": 1.2339886803693774,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.018507596067917784,
      "loss": 0.8115,
      "num_input_tokens_seen": 4804192,
      "step": 8285
    },
    {
      "epoch": 1.2347333929103366,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.01851876675603217,
      "loss": 0.7947,
      "num_input_tokens_seen": 4807072,
      "step": 8290
    },
    {
      "epoch": 1.2354781054512958,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.01852993744414656,
      "loss": 0.8086,
      "num_input_tokens_seen": 4810048,
      "step": 8295
    },
    {
      "epoch": 1.236222817992255,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.01854110813226095,
      "loss": 0.8308,
      "num_input_tokens_seen": 4813120,
      "step": 8300
    },
    {
      "epoch": 1.2369675305332142,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.018552278820375336,
      "loss": 0.7936,
      "num_input_tokens_seen": 4816128,
      "step": 8305
    },
    {
      "epoch": 1.2377122430741734,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.018563449508489725,
      "loss": 0.8048,
      "num_input_tokens_seen": 4818976,
      "step": 8310
    },
    {
      "epoch": 1.2384569556151326,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.01857462019660411,
      "loss": 0.8372,
      "num_input_tokens_seen": 4821760,
      "step": 8315
    },
    {
      "epoch": 1.2392016681560918,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.018585790884718498,
      "loss": 0.8023,
      "num_input_tokens_seen": 4824544,
      "step": 8320
    },
    {
      "epoch": 1.239946380697051,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.018596961572832884,
      "loss": 0.8073,
      "num_input_tokens_seen": 4827648,
      "step": 8325
    },
    {
      "epoch": 1.2406910932380102,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.018608132260947274,
      "loss": 0.7913,
      "num_input_tokens_seen": 4830336,
      "step": 8330
    },
    {
      "epoch": 1.2414358057789694,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01861930294906166,
      "loss": 0.7997,
      "num_input_tokens_seen": 4833056,
      "step": 8335
    },
    {
      "epoch": 1.2421805183199286,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.01863047363717605,
      "loss": 0.792,
      "num_input_tokens_seen": 4836096,
      "step": 8340
    },
    {
      "epoch": 1.2429252308608878,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.018641644325290436,
      "loss": 0.8226,
      "num_input_tokens_seen": 4838912,
      "step": 8345
    },
    {
      "epoch": 1.243669943401847,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.018652815013404826,
      "loss": 0.8057,
      "num_input_tokens_seen": 4841728,
      "step": 8350
    },
    {
      "epoch": 1.244414655942806,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.018663985701519212,
      "loss": 0.8072,
      "num_input_tokens_seen": 4844928,
      "step": 8355
    },
    {
      "epoch": 1.2451593684837652,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0186751563896336,
      "loss": 0.8047,
      "num_input_tokens_seen": 4848096,
      "step": 8360
    },
    {
      "epoch": 1.2459040810247244,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.018686327077747988,
      "loss": 0.7866,
      "num_input_tokens_seen": 4851296,
      "step": 8365
    },
    {
      "epoch": 1.2466487935656836,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.018697497765862377,
      "loss": 0.8094,
      "num_input_tokens_seen": 4854208,
      "step": 8370
    },
    {
      "epoch": 1.2473935061066428,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.018708668453976764,
      "loss": 0.8027,
      "num_input_tokens_seen": 4856896,
      "step": 8375
    },
    {
      "epoch": 1.248138218647602,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.018719839142091153,
      "loss": 0.7879,
      "num_input_tokens_seen": 4859744,
      "step": 8380
    },
    {
      "epoch": 1.2488829311885612,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.01873100983020554,
      "loss": 0.8164,
      "num_input_tokens_seen": 4862560,
      "step": 8385
    },
    {
      "epoch": 1.2496276437295204,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.01874218051831993,
      "loss": 0.7958,
      "num_input_tokens_seen": 4865568,
      "step": 8390
    },
    {
      "epoch": 1.2503723562704796,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.018753351206434316,
      "loss": 0.8184,
      "num_input_tokens_seen": 4868384,
      "step": 8395
    },
    {
      "epoch": 1.2511170688114388,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.018764521894548705,
      "loss": 0.8075,
      "num_input_tokens_seen": 4870912,
      "step": 8400
    },
    {
      "epoch": 1.251861781352398,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01877569258266309,
      "loss": 0.8224,
      "num_input_tokens_seen": 4873792,
      "step": 8405
    },
    {
      "epoch": 1.2526064938933572,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.018786863270777478,
      "loss": 0.8077,
      "num_input_tokens_seen": 4876864,
      "step": 8410
    },
    {
      "epoch": 1.2533512064343164,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.018798033958891867,
      "loss": 0.7884,
      "num_input_tokens_seen": 4879872,
      "step": 8415
    },
    {
      "epoch": 1.2540959189752756,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.018809204647006254,
      "loss": 0.8193,
      "num_input_tokens_seen": 4882944,
      "step": 8420
    },
    {
      "epoch": 1.2548406315162346,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.018820375335120643,
      "loss": 0.8278,
      "num_input_tokens_seen": 4885952,
      "step": 8425
    },
    {
      "epoch": 1.2555853440571938,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.01883154602323503,
      "loss": 0.8239,
      "num_input_tokens_seen": 4888768,
      "step": 8430
    },
    {
      "epoch": 1.256330056598153,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.01884271671134942,
      "loss": 0.8052,
      "num_input_tokens_seen": 4891584,
      "step": 8435
    },
    {
      "epoch": 1.2570747691391122,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.018853887399463806,
      "loss": 0.8131,
      "num_input_tokens_seen": 4894432,
      "step": 8440
    },
    {
      "epoch": 1.2578194816800714,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.018865058087578195,
      "loss": 0.8125,
      "num_input_tokens_seen": 4897152,
      "step": 8445
    },
    {
      "epoch": 1.2585641942210306,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01887622877569258,
      "loss": 0.8003,
      "num_input_tokens_seen": 4900224,
      "step": 8450
    },
    {
      "epoch": 1.2593089067619898,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01888739946380697,
      "loss": 0.7999,
      "num_input_tokens_seen": 4903072,
      "step": 8455
    },
    {
      "epoch": 1.260053619302949,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.018898570151921357,
      "loss": 0.7907,
      "num_input_tokens_seen": 4905792,
      "step": 8460
    },
    {
      "epoch": 1.2607983318439082,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.018909740840035747,
      "loss": 0.7997,
      "num_input_tokens_seen": 4908608,
      "step": 8465
    },
    {
      "epoch": 1.2615430443848674,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.018920911528150133,
      "loss": 0.8166,
      "num_input_tokens_seen": 4911648,
      "step": 8470
    },
    {
      "epoch": 1.2622877569258266,
      "grad_norm": 0.048828125,
      "learning_rate": 0.018932082216264523,
      "loss": 0.8038,
      "num_input_tokens_seen": 4914816,
      "step": 8475
    },
    {
      "epoch": 1.2630324694667858,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01894325290437891,
      "loss": 0.7954,
      "num_input_tokens_seen": 4917728,
      "step": 8480
    },
    {
      "epoch": 1.263777182007745,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0189544235924933,
      "loss": 0.8171,
      "num_input_tokens_seen": 4920704,
      "step": 8485
    },
    {
      "epoch": 1.2645218945487042,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.018965594280607682,
      "loss": 0.7981,
      "num_input_tokens_seen": 4923392,
      "step": 8490
    },
    {
      "epoch": 1.2652666070896634,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01897676496872207,
      "loss": 0.7808,
      "num_input_tokens_seen": 4926432,
      "step": 8495
    },
    {
      "epoch": 1.2660113196306226,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.018987935656836458,
      "loss": 0.79,
      "num_input_tokens_seen": 4929600,
      "step": 8500
    },
    {
      "epoch": 1.2667560321715818,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.018999106344950847,
      "loss": 0.8064,
      "num_input_tokens_seen": 4932288,
      "step": 8505
    },
    {
      "epoch": 1.267500744712541,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.019010277033065234,
      "loss": 0.8144,
      "num_input_tokens_seen": 4935168,
      "step": 8510
    },
    {
      "epoch": 1.2682454572535002,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.019021447721179623,
      "loss": 0.7943,
      "num_input_tokens_seen": 4937920,
      "step": 8515
    },
    {
      "epoch": 1.2689901697944594,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.019032618409294013,
      "loss": 0.8092,
      "num_input_tokens_seen": 4940992,
      "step": 8520
    },
    {
      "epoch": 1.2697348823354186,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.0190437890974084,
      "loss": 0.831,
      "num_input_tokens_seen": 4944128,
      "step": 8525
    },
    {
      "epoch": 1.2704795948763778,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.01905495978552279,
      "loss": 0.7806,
      "num_input_tokens_seen": 4947072,
      "step": 8530
    },
    {
      "epoch": 1.2712243074173368,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.019066130473637175,
      "loss": 0.8111,
      "num_input_tokens_seen": 4949760,
      "step": 8535
    },
    {
      "epoch": 1.271969019958296,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.019077301161751565,
      "loss": 0.8269,
      "num_input_tokens_seen": 4952544,
      "step": 8540
    },
    {
      "epoch": 1.2727137324992552,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01908847184986595,
      "loss": 0.8218,
      "num_input_tokens_seen": 4955520,
      "step": 8545
    },
    {
      "epoch": 1.2734584450402144,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.01909964253798034,
      "loss": 0.824,
      "num_input_tokens_seen": 4958272,
      "step": 8550
    },
    {
      "epoch": 1.2742031575811736,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.019110813226094727,
      "loss": 0.7906,
      "num_input_tokens_seen": 4961280,
      "step": 8555
    },
    {
      "epoch": 1.2749478701221328,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.019121983914209117,
      "loss": 0.8156,
      "num_input_tokens_seen": 4964032,
      "step": 8560
    },
    {
      "epoch": 1.275692582663092,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.019133154602323503,
      "loss": 0.7996,
      "num_input_tokens_seen": 4967008,
      "step": 8565
    },
    {
      "epoch": 1.2764372952040512,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.019144325290437893,
      "loss": 0.7925,
      "num_input_tokens_seen": 4969696,
      "step": 8570
    },
    {
      "epoch": 1.2771820077450105,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.01915549597855228,
      "loss": 0.8159,
      "num_input_tokens_seen": 4973280,
      "step": 8575
    },
    {
      "epoch": 1.2779267202859697,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.019166666666666665,
      "loss": 0.8047,
      "num_input_tokens_seen": 4976032,
      "step": 8580
    },
    {
      "epoch": 1.2786714328269289,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01917783735478105,
      "loss": 0.8126,
      "num_input_tokens_seen": 4979168,
      "step": 8585
    },
    {
      "epoch": 1.279416145367888,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01918900804289544,
      "loss": 0.7915,
      "num_input_tokens_seen": 4982112,
      "step": 8590
    },
    {
      "epoch": 1.2801608579088473,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.019200178731009827,
      "loss": 0.8151,
      "num_input_tokens_seen": 4984800,
      "step": 8595
    },
    {
      "epoch": 1.2809055704498062,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.019211349419124217,
      "loss": 0.7866,
      "num_input_tokens_seen": 4987584,
      "step": 8600
    },
    {
      "epoch": 1.2816502829907654,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.019222520107238603,
      "loss": 0.8034,
      "num_input_tokens_seen": 4990592,
      "step": 8605
    },
    {
      "epoch": 1.2823949955317246,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.019233690795352993,
      "loss": 0.8082,
      "num_input_tokens_seen": 4993696,
      "step": 8610
    },
    {
      "epoch": 1.2831397080726838,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.01924486148346738,
      "loss": 0.8092,
      "num_input_tokens_seen": 4996864,
      "step": 8615
    },
    {
      "epoch": 1.283884420613643,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01925603217158177,
      "loss": 0.7959,
      "num_input_tokens_seen": 4999744,
      "step": 8620
    },
    {
      "epoch": 1.2846291331546023,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.019267202859696155,
      "loss": 0.8046,
      "num_input_tokens_seen": 5002656,
      "step": 8625
    },
    {
      "epoch": 1.2853738456955615,
      "grad_norm": 0.048828125,
      "learning_rate": 0.019278373547810545,
      "loss": 0.8014,
      "num_input_tokens_seen": 5005280,
      "step": 8630
    },
    {
      "epoch": 1.2861185582365207,
      "grad_norm": 0.033203125,
      "learning_rate": 0.019289544235924935,
      "loss": 0.809,
      "num_input_tokens_seen": 5008352,
      "step": 8635
    },
    {
      "epoch": 1.2868632707774799,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.01930071492403932,
      "loss": 0.8048,
      "num_input_tokens_seen": 5011200,
      "step": 8640
    },
    {
      "epoch": 1.287607983318439,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.01931188561215371,
      "loss": 0.7993,
      "num_input_tokens_seen": 5014016,
      "step": 8645
    },
    {
      "epoch": 1.2883526958593983,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.019323056300268097,
      "loss": 0.807,
      "num_input_tokens_seen": 5016832,
      "step": 8650
    },
    {
      "epoch": 1.2890974084003575,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.019334226988382486,
      "loss": 0.7973,
      "num_input_tokens_seen": 5019904,
      "step": 8655
    },
    {
      "epoch": 1.2898421209413167,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.019345397676496873,
      "loss": 0.8205,
      "num_input_tokens_seen": 5022784,
      "step": 8660
    },
    {
      "epoch": 1.2905868334822759,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.01935656836461126,
      "loss": 0.7961,
      "num_input_tokens_seen": 5025696,
      "step": 8665
    },
    {
      "epoch": 1.291331546023235,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.019367739052725645,
      "loss": 0.8103,
      "num_input_tokens_seen": 5028704,
      "step": 8670
    },
    {
      "epoch": 1.2920762585641943,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.019378909740840035,
      "loss": 0.8216,
      "num_input_tokens_seen": 5031648,
      "step": 8675
    },
    {
      "epoch": 1.2928209711051535,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01939008042895442,
      "loss": 0.7888,
      "num_input_tokens_seen": 5034752,
      "step": 8680
    },
    {
      "epoch": 1.2935656836461127,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.01940125111706881,
      "loss": 0.8106,
      "num_input_tokens_seen": 5037568,
      "step": 8685
    },
    {
      "epoch": 1.2943103961870719,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.019412421805183197,
      "loss": 0.8073,
      "num_input_tokens_seen": 5040000,
      "step": 8690
    },
    {
      "epoch": 1.295055108728031,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.019423592493297587,
      "loss": 0.7804,
      "num_input_tokens_seen": 5042976,
      "step": 8695
    },
    {
      "epoch": 1.2957998212689903,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.019434763181411973,
      "loss": 0.8031,
      "num_input_tokens_seen": 5045792,
      "step": 8700
    },
    {
      "epoch": 1.2965445338099495,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.019445933869526363,
      "loss": 0.8071,
      "num_input_tokens_seen": 5048288,
      "step": 8705
    },
    {
      "epoch": 1.2972892463509085,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01945710455764075,
      "loss": 0.812,
      "num_input_tokens_seen": 5051200,
      "step": 8710
    },
    {
      "epoch": 1.2980339588918677,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.01946827524575514,
      "loss": 0.8101,
      "num_input_tokens_seen": 5054048,
      "step": 8715
    },
    {
      "epoch": 1.2987786714328269,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.019479445933869525,
      "loss": 0.7875,
      "num_input_tokens_seen": 5056736,
      "step": 8720
    },
    {
      "epoch": 1.299523383973786,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.019490616621983915,
      "loss": 0.8112,
      "num_input_tokens_seen": 5059936,
      "step": 8725
    },
    {
      "epoch": 1.3002680965147453,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0195017873100983,
      "loss": 0.7896,
      "num_input_tokens_seen": 5062688,
      "step": 8730
    },
    {
      "epoch": 1.3010128090557045,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.01951295799821269,
      "loss": 0.797,
      "num_input_tokens_seen": 5065792,
      "step": 8735
    },
    {
      "epoch": 1.3017575215966637,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.019524128686327077,
      "loss": 0.791,
      "num_input_tokens_seen": 5068768,
      "step": 8740
    },
    {
      "epoch": 1.302502234137623,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.019535299374441466,
      "loss": 0.8038,
      "num_input_tokens_seen": 5071584,
      "step": 8745
    },
    {
      "epoch": 1.303246946678582,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.019546470062555853,
      "loss": 0.8173,
      "num_input_tokens_seen": 5074400,
      "step": 8750
    },
    {
      "epoch": 1.3039916592195413,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01955764075067024,
      "loss": 0.821,
      "num_input_tokens_seen": 5077120,
      "step": 8755
    },
    {
      "epoch": 1.3047363717605005,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.01956881143878463,
      "loss": 0.7973,
      "num_input_tokens_seen": 5079872,
      "step": 8760
    },
    {
      "epoch": 1.3054810843014597,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.019579982126899015,
      "loss": 0.8149,
      "num_input_tokens_seen": 5082720,
      "step": 8765
    },
    {
      "epoch": 1.306225796842419,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.019591152815013405,
      "loss": 0.8352,
      "num_input_tokens_seen": 5085504,
      "step": 8770
    },
    {
      "epoch": 1.3069705093833779,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.01960232350312779,
      "loss": 0.8177,
      "num_input_tokens_seen": 5088000,
      "step": 8775
    },
    {
      "epoch": 1.307715221924337,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.01961349419124218,
      "loss": 0.8196,
      "num_input_tokens_seen": 5090752,
      "step": 8780
    },
    {
      "epoch": 1.3084599344652963,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.019624664879356567,
      "loss": 0.8319,
      "num_input_tokens_seen": 5093440,
      "step": 8785
    },
    {
      "epoch": 1.3092046470062555,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.019635835567470956,
      "loss": 0.8065,
      "num_input_tokens_seen": 5096288,
      "step": 8790
    },
    {
      "epoch": 1.3099493595472147,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.019647006255585343,
      "loss": 0.8145,
      "num_input_tokens_seen": 5099072,
      "step": 8795
    },
    {
      "epoch": 1.310694072088174,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.019658176943699732,
      "loss": 0.7977,
      "num_input_tokens_seen": 5102080,
      "step": 8800
    },
    {
      "epoch": 1.311438784629133,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.01966934763181412,
      "loss": 0.7927,
      "num_input_tokens_seen": 5105024,
      "step": 8805
    },
    {
      "epoch": 1.3121834971700923,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01968051831992851,
      "loss": 0.7866,
      "num_input_tokens_seen": 5107904,
      "step": 8810
    },
    {
      "epoch": 1.3129282097110515,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.019691689008042895,
      "loss": 0.8044,
      "num_input_tokens_seen": 5110880,
      "step": 8815
    },
    {
      "epoch": 1.3136729222520107,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.019702859696157284,
      "loss": 0.8088,
      "num_input_tokens_seen": 5113472,
      "step": 8820
    },
    {
      "epoch": 1.31441763479297,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.01971403038427167,
      "loss": 0.7942,
      "num_input_tokens_seen": 5116544,
      "step": 8825
    },
    {
      "epoch": 1.3151623473339291,
      "grad_norm": 0.01953125,
      "learning_rate": 0.01972520107238606,
      "loss": 0.8029,
      "num_input_tokens_seen": 5119136,
      "step": 8830
    },
    {
      "epoch": 1.3159070598748883,
      "grad_norm": 0.044921875,
      "learning_rate": 0.019736371760500443,
      "loss": 0.796,
      "num_input_tokens_seen": 5122016,
      "step": 8835
    },
    {
      "epoch": 1.3166517724158475,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.019747542448614833,
      "loss": 0.8188,
      "num_input_tokens_seen": 5124960,
      "step": 8840
    },
    {
      "epoch": 1.3173964849568067,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.01975871313672922,
      "loss": 0.795,
      "num_input_tokens_seen": 5128000,
      "step": 8845
    },
    {
      "epoch": 1.318141197497766,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.01976988382484361,
      "loss": 0.7871,
      "num_input_tokens_seen": 5131008,
      "step": 8850
    },
    {
      "epoch": 1.3188859100387251,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.019781054512958,
      "loss": 0.7987,
      "num_input_tokens_seen": 5133824,
      "step": 8855
    },
    {
      "epoch": 1.3196306225796843,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.019792225201072385,
      "loss": 0.7866,
      "num_input_tokens_seen": 5136608,
      "step": 8860
    },
    {
      "epoch": 1.3203753351206435,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.019803395889186774,
      "loss": 0.8155,
      "num_input_tokens_seen": 5139360,
      "step": 8865
    },
    {
      "epoch": 1.3211200476616027,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.01981456657730116,
      "loss": 0.7787,
      "num_input_tokens_seen": 5142208,
      "step": 8870
    },
    {
      "epoch": 1.321864760202562,
      "grad_norm": 0.046875,
      "learning_rate": 0.01982573726541555,
      "loss": 0.8186,
      "num_input_tokens_seen": 5145280,
      "step": 8875
    },
    {
      "epoch": 1.322609472743521,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.019836907953529936,
      "loss": 0.7946,
      "num_input_tokens_seen": 5148224,
      "step": 8880
    },
    {
      "epoch": 1.3233541852844801,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.019848078641644326,
      "loss": 0.8042,
      "num_input_tokens_seen": 5150976,
      "step": 8885
    },
    {
      "epoch": 1.3240988978254393,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.019859249329758712,
      "loss": 0.8023,
      "num_input_tokens_seen": 5153504,
      "step": 8890
    },
    {
      "epoch": 1.3248436103663985,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.019870420017873102,
      "loss": 0.8286,
      "num_input_tokens_seen": 5156512,
      "step": 8895
    },
    {
      "epoch": 1.3255883229073577,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01988159070598749,
      "loss": 0.8347,
      "num_input_tokens_seen": 5159328,
      "step": 8900
    },
    {
      "epoch": 1.326333035448317,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.019892761394101878,
      "loss": 0.8058,
      "num_input_tokens_seen": 5162048,
      "step": 8905
    },
    {
      "epoch": 1.3270777479892761,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.019903932082216264,
      "loss": 0.7808,
      "num_input_tokens_seen": 5165056,
      "step": 8910
    },
    {
      "epoch": 1.3278224605302353,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.019915102770330654,
      "loss": 0.8163,
      "num_input_tokens_seen": 5168000,
      "step": 8915
    },
    {
      "epoch": 1.3285671730711945,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.01992627345844504,
      "loss": 0.8305,
      "num_input_tokens_seen": 5170784,
      "step": 8920
    },
    {
      "epoch": 1.3293118856121537,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.019937444146559426,
      "loss": 0.7843,
      "num_input_tokens_seen": 5174016,
      "step": 8925
    },
    {
      "epoch": 1.330056598153113,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.019948614834673813,
      "loss": 0.8133,
      "num_input_tokens_seen": 5177120,
      "step": 8930
    },
    {
      "epoch": 1.3308013106940721,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.019959785522788202,
      "loss": 0.7967,
      "num_input_tokens_seen": 5179904,
      "step": 8935
    },
    {
      "epoch": 1.3315460232350314,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.01997095621090259,
      "loss": 0.7754,
      "num_input_tokens_seen": 5182912,
      "step": 8940
    },
    {
      "epoch": 1.3322907357759903,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.01998212689901698,
      "loss": 0.8011,
      "num_input_tokens_seen": 5185888,
      "step": 8945
    },
    {
      "epoch": 1.3330354483169495,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.019993297587131365,
      "loss": 0.8013,
      "num_input_tokens_seen": 5188864,
      "step": 8950
    },
    {
      "epoch": 1.3337801608579087,
      "grad_norm": 0.037109375,
      "learning_rate": 0.020004468275245754,
      "loss": 0.7879,
      "num_input_tokens_seen": 5191392,
      "step": 8955
    },
    {
      "epoch": 1.334524873398868,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02001563896336014,
      "loss": 0.7738,
      "num_input_tokens_seen": 5193984,
      "step": 8960
    },
    {
      "epoch": 1.3352695859398271,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.02002680965147453,
      "loss": 0.799,
      "num_input_tokens_seen": 5196960,
      "step": 8965
    },
    {
      "epoch": 1.3360142984807863,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02003798033958892,
      "loss": 0.753,
      "num_input_tokens_seen": 5199776,
      "step": 8970
    },
    {
      "epoch": 1.3367590110217455,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.020049151027703306,
      "loss": 0.8538,
      "num_input_tokens_seen": 5203040,
      "step": 8975
    },
    {
      "epoch": 1.3375037235627047,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.020060321715817696,
      "loss": 0.7857,
      "num_input_tokens_seen": 5205952,
      "step": 8980
    },
    {
      "epoch": 1.338248436103664,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.020071492403932082,
      "loss": 0.7913,
      "num_input_tokens_seen": 5208768,
      "step": 8985
    },
    {
      "epoch": 1.3389931486446232,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.020082663092046472,
      "loss": 0.8059,
      "num_input_tokens_seen": 5211584,
      "step": 8990
    },
    {
      "epoch": 1.3397378611855824,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.020093833780160858,
      "loss": 0.8582,
      "num_input_tokens_seen": 5214752,
      "step": 8995
    },
    {
      "epoch": 1.3404825737265416,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.020105004468275248,
      "loss": 0.8108,
      "num_input_tokens_seen": 5217536,
      "step": 9000
    },
    {
      "epoch": 1.3412272862675008,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.020116175156389634,
      "loss": 0.787,
      "num_input_tokens_seen": 5220480,
      "step": 9005
    },
    {
      "epoch": 1.34197199880846,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02012734584450402,
      "loss": 0.8246,
      "num_input_tokens_seen": 5223520,
      "step": 9010
    },
    {
      "epoch": 1.3427167113494192,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.020138516532618406,
      "loss": 0.7788,
      "num_input_tokens_seen": 5227744,
      "step": 9015
    },
    {
      "epoch": 1.3434614238903784,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.020149687220732796,
      "loss": 0.8091,
      "num_input_tokens_seen": 5230400,
      "step": 9020
    },
    {
      "epoch": 1.3442061364313376,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.020160857908847182,
      "loss": 0.8237,
      "num_input_tokens_seen": 5233312,
      "step": 9025
    },
    {
      "epoch": 1.3449508489722968,
      "grad_norm": 0.03515625,
      "learning_rate": 0.020172028596961572,
      "loss": 0.8144,
      "num_input_tokens_seen": 5236256,
      "step": 9030
    },
    {
      "epoch": 1.345695561513256,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.020183199285075958,
      "loss": 0.8097,
      "num_input_tokens_seen": 5239296,
      "step": 9035
    },
    {
      "epoch": 1.3464402740542152,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.020194369973190348,
      "loss": 0.8111,
      "num_input_tokens_seen": 5242560,
      "step": 9040
    },
    {
      "epoch": 1.3471849865951744,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.020205540661304734,
      "loss": 0.8068,
      "num_input_tokens_seen": 5245408,
      "step": 9045
    },
    {
      "epoch": 1.3479296991361336,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.020216711349419124,
      "loss": 0.804,
      "num_input_tokens_seen": 5248448,
      "step": 9050
    },
    {
      "epoch": 1.3486744116770926,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02022788203753351,
      "loss": 0.8207,
      "num_input_tokens_seen": 5251616,
      "step": 9055
    },
    {
      "epoch": 1.3494191242180518,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0202390527256479,
      "loss": 0.8011,
      "num_input_tokens_seen": 5254752,
      "step": 9060
    },
    {
      "epoch": 1.350163836759011,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.020250223413762286,
      "loss": 0.8051,
      "num_input_tokens_seen": 5257696,
      "step": 9065
    },
    {
      "epoch": 1.3509085492999702,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.020261394101876676,
      "loss": 0.7975,
      "num_input_tokens_seen": 5260800,
      "step": 9070
    },
    {
      "epoch": 1.3516532618409294,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.020272564789991062,
      "loss": 0.8052,
      "num_input_tokens_seen": 5263488,
      "step": 9075
    },
    {
      "epoch": 1.3523979743818886,
      "grad_norm": 0.046875,
      "learning_rate": 0.02028373547810545,
      "loss": 0.801,
      "num_input_tokens_seen": 5266272,
      "step": 9080
    },
    {
      "epoch": 1.3531426869228478,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02029490616621984,
      "loss": 0.8049,
      "num_input_tokens_seen": 5269248,
      "step": 9085
    },
    {
      "epoch": 1.353887399463807,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.020306076854334228,
      "loss": 0.8055,
      "num_input_tokens_seen": 5272448,
      "step": 9090
    },
    {
      "epoch": 1.3546321120047662,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.020317247542448614,
      "loss": 0.8148,
      "num_input_tokens_seen": 5275200,
      "step": 9095
    },
    {
      "epoch": 1.3553768245457254,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.020328418230563,
      "loss": 0.7711,
      "num_input_tokens_seen": 5277984,
      "step": 9100
    },
    {
      "epoch": 1.3561215370866846,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02033958891867739,
      "loss": 0.7912,
      "num_input_tokens_seen": 5281024,
      "step": 9105
    },
    {
      "epoch": 1.3568662496276438,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.020350759606791776,
      "loss": 0.8236,
      "num_input_tokens_seen": 5284000,
      "step": 9110
    },
    {
      "epoch": 1.357610962168603,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.020361930294906166,
      "loss": 0.8221,
      "num_input_tokens_seen": 5286624,
      "step": 9115
    },
    {
      "epoch": 1.358355674709562,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.020373100983020552,
      "loss": 0.7948,
      "num_input_tokens_seen": 5289408,
      "step": 9120
    },
    {
      "epoch": 1.3591003872505212,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02038427167113494,
      "loss": 0.8165,
      "num_input_tokens_seen": 5292480,
      "step": 9125
    },
    {
      "epoch": 1.3598450997914804,
      "grad_norm": 0.03515625,
      "learning_rate": 0.020395442359249328,
      "loss": 0.8064,
      "num_input_tokens_seen": 5295776,
      "step": 9130
    },
    {
      "epoch": 1.3605898123324396,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.020406613047363718,
      "loss": 0.8289,
      "num_input_tokens_seen": 5298400,
      "step": 9135
    },
    {
      "epoch": 1.3613345248733988,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.020417783735478104,
      "loss": 0.8162,
      "num_input_tokens_seen": 5301152,
      "step": 9140
    },
    {
      "epoch": 1.362079237414358,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.020428954423592494,
      "loss": 0.8081,
      "num_input_tokens_seen": 5304480,
      "step": 9145
    },
    {
      "epoch": 1.3628239499553172,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02044012511170688,
      "loss": 0.7994,
      "num_input_tokens_seen": 5307520,
      "step": 9150
    },
    {
      "epoch": 1.3635686624962764,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02045129579982127,
      "loss": 0.8007,
      "num_input_tokens_seen": 5310240,
      "step": 9155
    },
    {
      "epoch": 1.3643133750372356,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.020462466487935656,
      "loss": 0.8088,
      "num_input_tokens_seen": 5312928,
      "step": 9160
    },
    {
      "epoch": 1.3650580875781948,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.020473637176050045,
      "loss": 0.8033,
      "num_input_tokens_seen": 5316000,
      "step": 9165
    },
    {
      "epoch": 1.365802800119154,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02048480786416443,
      "loss": 0.803,
      "num_input_tokens_seen": 5319232,
      "step": 9170
    },
    {
      "epoch": 1.3665475126601132,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02049597855227882,
      "loss": 0.8124,
      "num_input_tokens_seen": 5322080,
      "step": 9175
    },
    {
      "epoch": 1.3672922252010724,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.020507149240393208,
      "loss": 0.8075,
      "num_input_tokens_seen": 5324928,
      "step": 9180
    },
    {
      "epoch": 1.3680369377420316,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.020518319928507594,
      "loss": 0.8081,
      "num_input_tokens_seen": 5327744,
      "step": 9185
    },
    {
      "epoch": 1.3687816502829908,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.020529490616621984,
      "loss": 0.7946,
      "num_input_tokens_seen": 5330400,
      "step": 9190
    },
    {
      "epoch": 1.36952636282395,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02054066130473637,
      "loss": 0.799,
      "num_input_tokens_seen": 5333248,
      "step": 9195
    },
    {
      "epoch": 1.3702710753649092,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02055183199285076,
      "loss": 0.7926,
      "num_input_tokens_seen": 5336064,
      "step": 9200
    },
    {
      "epoch": 1.3710157879058684,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.020563002680965146,
      "loss": 0.8,
      "num_input_tokens_seen": 5339072,
      "step": 9205
    },
    {
      "epoch": 1.3717605004468276,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.020574173369079535,
      "loss": 0.8,
      "num_input_tokens_seen": 5342016,
      "step": 9210
    },
    {
      "epoch": 1.3725052129877868,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.02058534405719392,
      "loss": 0.8038,
      "num_input_tokens_seen": 5344896,
      "step": 9215
    },
    {
      "epoch": 1.373249925528746,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02059651474530831,
      "loss": 0.8146,
      "num_input_tokens_seen": 5347648,
      "step": 9220
    },
    {
      "epoch": 1.3739946380697052,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.020607685433422698,
      "loss": 0.8031,
      "num_input_tokens_seen": 5350624,
      "step": 9225
    },
    {
      "epoch": 1.3747393506106642,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.020618856121537087,
      "loss": 0.8098,
      "num_input_tokens_seen": 5353728,
      "step": 9230
    },
    {
      "epoch": 1.3754840631516234,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.020630026809651474,
      "loss": 0.7693,
      "num_input_tokens_seen": 5356640,
      "step": 9235
    },
    {
      "epoch": 1.3762287756925826,
      "grad_norm": 0.017578125,
      "learning_rate": 0.020641197497765863,
      "loss": 0.8064,
      "num_input_tokens_seen": 5359552,
      "step": 9240
    },
    {
      "epoch": 1.3769734882335418,
      "grad_norm": 0.03125,
      "learning_rate": 0.02065236818588025,
      "loss": 0.7906,
      "num_input_tokens_seen": 5362528,
      "step": 9245
    },
    {
      "epoch": 1.377718200774501,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02066353887399464,
      "loss": 0.8307,
      "num_input_tokens_seen": 5365600,
      "step": 9250
    },
    {
      "epoch": 1.3784629133154602,
      "grad_norm": 0.02734375,
      "learning_rate": 0.020674709562109025,
      "loss": 0.8043,
      "num_input_tokens_seen": 5368288,
      "step": 9255
    },
    {
      "epoch": 1.3792076258564194,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.020685880250223415,
      "loss": 0.8125,
      "num_input_tokens_seen": 5371296,
      "step": 9260
    },
    {
      "epoch": 1.3799523383973786,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.0206970509383378,
      "loss": 0.7883,
      "num_input_tokens_seen": 5374240,
      "step": 9265
    },
    {
      "epoch": 1.3806970509383378,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.020708221626452188,
      "loss": 0.8076,
      "num_input_tokens_seen": 5377312,
      "step": 9270
    },
    {
      "epoch": 1.381441763479297,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.020719392314566574,
      "loss": 0.8076,
      "num_input_tokens_seen": 5380064,
      "step": 9275
    },
    {
      "epoch": 1.3821864760202562,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.020730563002680964,
      "loss": 0.8206,
      "num_input_tokens_seen": 5383072,
      "step": 9280
    },
    {
      "epoch": 1.3829311885612154,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02074173369079535,
      "loss": 0.8166,
      "num_input_tokens_seen": 5385792,
      "step": 9285
    },
    {
      "epoch": 1.3836759011021746,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02075290437890974,
      "loss": 0.8103,
      "num_input_tokens_seen": 5388640,
      "step": 9290
    },
    {
      "epoch": 1.3844206136431336,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.020764075067024126,
      "loss": 0.7924,
      "num_input_tokens_seen": 5391360,
      "step": 9295
    },
    {
      "epoch": 1.3851653261840928,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.020775245755138515,
      "loss": 0.7954,
      "num_input_tokens_seen": 5394240,
      "step": 9300
    },
    {
      "epoch": 1.385910038725052,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.0207864164432529,
      "loss": 0.8171,
      "num_input_tokens_seen": 5396960,
      "step": 9305
    },
    {
      "epoch": 1.3866547512660112,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02079758713136729,
      "loss": 0.7909,
      "num_input_tokens_seen": 5399616,
      "step": 9310
    },
    {
      "epoch": 1.3873994638069704,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.02080875781948168,
      "loss": 0.804,
      "num_input_tokens_seen": 5402144,
      "step": 9315
    },
    {
      "epoch": 1.3881441763479296,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.020819928507596067,
      "loss": 0.7887,
      "num_input_tokens_seen": 5404992,
      "step": 9320
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 0.0390625,
      "learning_rate": 0.020831099195710457,
      "loss": 0.8093,
      "num_input_tokens_seen": 5407808,
      "step": 9325
    },
    {
      "epoch": 1.389633601429848,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.020842269883824843,
      "loss": 0.8135,
      "num_input_tokens_seen": 5410752,
      "step": 9330
    },
    {
      "epoch": 1.3903783139708072,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.020853440571939233,
      "loss": 0.8262,
      "num_input_tokens_seen": 5413504,
      "step": 9335
    },
    {
      "epoch": 1.3911230265117664,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02086461126005362,
      "loss": 0.8177,
      "num_input_tokens_seen": 5416672,
      "step": 9340
    },
    {
      "epoch": 1.3918677390527256,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02087578194816801,
      "loss": 0.8126,
      "num_input_tokens_seen": 5419840,
      "step": 9345
    },
    {
      "epoch": 1.3926124515936849,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.020886952636282395,
      "loss": 0.7995,
      "num_input_tokens_seen": 5422624,
      "step": 9350
    },
    {
      "epoch": 1.393357164134644,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02089812332439678,
      "loss": 0.8233,
      "num_input_tokens_seen": 5425568,
      "step": 9355
    },
    {
      "epoch": 1.3941018766756033,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.020909294012511168,
      "loss": 0.7996,
      "num_input_tokens_seen": 5428416,
      "step": 9360
    },
    {
      "epoch": 1.3948465892165625,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.020920464700625557,
      "loss": 0.8144,
      "num_input_tokens_seen": 5431264,
      "step": 9365
    },
    {
      "epoch": 1.3955913017575217,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.020931635388739944,
      "loss": 0.8034,
      "num_input_tokens_seen": 5434048,
      "step": 9370
    },
    {
      "epoch": 1.3963360142984809,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.020942806076854333,
      "loss": 0.7984,
      "num_input_tokens_seen": 5437056,
      "step": 9375
    },
    {
      "epoch": 1.39708072683944,
      "grad_norm": 0.04296875,
      "learning_rate": 0.02095397676496872,
      "loss": 0.8214,
      "num_input_tokens_seen": 5439840,
      "step": 9380
    },
    {
      "epoch": 1.3978254393803993,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02096514745308311,
      "loss": 0.7973,
      "num_input_tokens_seen": 5442816,
      "step": 9385
    },
    {
      "epoch": 1.3985701519213585,
      "grad_norm": 0.029296875,
      "learning_rate": 0.020976318141197495,
      "loss": 0.7971,
      "num_input_tokens_seen": 5445664,
      "step": 9390
    },
    {
      "epoch": 1.3993148644623177,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.020987488829311885,
      "loss": 0.7844,
      "num_input_tokens_seen": 5448704,
      "step": 9395
    },
    {
      "epoch": 1.4000595770032767,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02099865951742627,
      "loss": 0.776,
      "num_input_tokens_seen": 5451648,
      "step": 9400
    },
    {
      "epoch": 1.4008042895442359,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02100983020554066,
      "loss": 0.779,
      "num_input_tokens_seen": 5454624,
      "step": 9405
    },
    {
      "epoch": 1.401549002085195,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.021021000893655047,
      "loss": 0.8175,
      "num_input_tokens_seen": 5457248,
      "step": 9410
    },
    {
      "epoch": 1.4022937146261543,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.021032171581769437,
      "loss": 0.7906,
      "num_input_tokens_seen": 5460032,
      "step": 9415
    },
    {
      "epoch": 1.4030384271671135,
      "grad_norm": 0.046875,
      "learning_rate": 0.021043342269883823,
      "loss": 0.8023,
      "num_input_tokens_seen": 5462816,
      "step": 9420
    },
    {
      "epoch": 1.4037831397080727,
      "grad_norm": 0.0703125,
      "learning_rate": 0.021054512957998213,
      "loss": 0.8601,
      "num_input_tokens_seen": 5465632,
      "step": 9425
    },
    {
      "epoch": 1.4045278522490319,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.021065683646112603,
      "loss": 0.7979,
      "num_input_tokens_seen": 5468544,
      "step": 9430
    },
    {
      "epoch": 1.405272564789991,
      "grad_norm": 0.197265625,
      "learning_rate": 0.02107685433422699,
      "loss": 0.7797,
      "num_input_tokens_seen": 5471360,
      "step": 9435
    },
    {
      "epoch": 1.4060172773309503,
      "grad_norm": 0.06640625,
      "learning_rate": 0.02108802502234138,
      "loss": 0.8382,
      "num_input_tokens_seen": 5474080,
      "step": 9440
    },
    {
      "epoch": 1.4067619898719095,
      "grad_norm": 0.359375,
      "learning_rate": 0.02109919571045576,
      "loss": 0.8207,
      "num_input_tokens_seen": 5477152,
      "step": 9445
    },
    {
      "epoch": 1.4075067024128687,
      "grad_norm": 0.625,
      "learning_rate": 0.02111036639857015,
      "loss": 0.9157,
      "num_input_tokens_seen": 5480064,
      "step": 9450
    },
    {
      "epoch": 1.4082514149538279,
      "grad_norm": 9.5625,
      "learning_rate": 0.021121537086684537,
      "loss": 0.8139,
      "num_input_tokens_seen": 5483136,
      "step": 9455
    },
    {
      "epoch": 1.408996127494787,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.021132707774798927,
      "loss": 0.8108,
      "num_input_tokens_seen": 5486368,
      "step": 9460
    },
    {
      "epoch": 1.409740840035746,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.021143878462913313,
      "loss": 0.8154,
      "num_input_tokens_seen": 5489024,
      "step": 9465
    },
    {
      "epoch": 1.4104855525767053,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.021155049151027703,
      "loss": 0.8034,
      "num_input_tokens_seen": 5491616,
      "step": 9470
    },
    {
      "epoch": 1.4112302651176645,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.02116621983914209,
      "loss": 0.7808,
      "num_input_tokens_seen": 5494464,
      "step": 9475
    },
    {
      "epoch": 1.4119749776586237,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.02117739052725648,
      "loss": 0.7728,
      "num_input_tokens_seen": 5497376,
      "step": 9480
    },
    {
      "epoch": 1.4127196901995829,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.021188561215370865,
      "loss": 0.7847,
      "num_input_tokens_seen": 5500128,
      "step": 9485
    },
    {
      "epoch": 1.413464402740542,
      "grad_norm": 0.6484375,
      "learning_rate": 0.021199731903485255,
      "loss": 0.8152,
      "num_input_tokens_seen": 5503136,
      "step": 9490
    },
    {
      "epoch": 1.4142091152815013,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02121090259159964,
      "loss": 0.8548,
      "num_input_tokens_seen": 5506048,
      "step": 9495
    },
    {
      "epoch": 1.4149538278224605,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.02122207327971403,
      "loss": 0.789,
      "num_input_tokens_seen": 5508736,
      "step": 9500
    },
    {
      "epoch": 1.4156985403634197,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.021233243967828417,
      "loss": 0.8385,
      "num_input_tokens_seen": 5511776,
      "step": 9505
    },
    {
      "epoch": 1.416443252904379,
      "grad_norm": 0.11328125,
      "learning_rate": 0.021244414655942807,
      "loss": 0.7992,
      "num_input_tokens_seen": 5514848,
      "step": 9510
    },
    {
      "epoch": 1.417187965445338,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.021255585344057193,
      "loss": 1.1079,
      "num_input_tokens_seen": 5517600,
      "step": 9515
    },
    {
      "epoch": 1.4179326779862973,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.021266756032171583,
      "loss": 0.805,
      "num_input_tokens_seen": 5520384,
      "step": 9520
    },
    {
      "epoch": 1.4186773905272565,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.02127792672028597,
      "loss": 0.8075,
      "num_input_tokens_seen": 5523392,
      "step": 9525
    },
    {
      "epoch": 1.4194221030682157,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.021289097408400355,
      "loss": 0.8202,
      "num_input_tokens_seen": 5526496,
      "step": 9530
    },
    {
      "epoch": 1.420166815609175,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.021300268096514745,
      "loss": 0.7948,
      "num_input_tokens_seen": 5529600,
      "step": 9535
    },
    {
      "epoch": 1.420911528150134,
      "grad_norm": 0.142578125,
      "learning_rate": 0.02131143878462913,
      "loss": 0.8255,
      "num_input_tokens_seen": 5532352,
      "step": 9540
    },
    {
      "epoch": 1.4216562406910933,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.02132260947274352,
      "loss": 0.8182,
      "num_input_tokens_seen": 5535456,
      "step": 9545
    },
    {
      "epoch": 1.4224009532320525,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.021333780160857907,
      "loss": 0.8046,
      "num_input_tokens_seen": 5538464,
      "step": 9550
    },
    {
      "epoch": 1.4231456657730117,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.021344950848972297,
      "loss": 0.8067,
      "num_input_tokens_seen": 5541344,
      "step": 9555
    },
    {
      "epoch": 1.423890378313971,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.021356121537086683,
      "loss": 0.8266,
      "num_input_tokens_seen": 5544320,
      "step": 9560
    },
    {
      "epoch": 1.4246350908549301,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.021367292225201073,
      "loss": 0.7949,
      "num_input_tokens_seen": 5547008,
      "step": 9565
    },
    {
      "epoch": 1.4253798033958893,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02137846291331546,
      "loss": 0.8122,
      "num_input_tokens_seen": 5549920,
      "step": 9570
    },
    {
      "epoch": 1.4261245159368483,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.02138963360142985,
      "loss": 0.7807,
      "num_input_tokens_seen": 5552960,
      "step": 9575
    },
    {
      "epoch": 1.4268692284778075,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.021400804289544235,
      "loss": 0.794,
      "num_input_tokens_seen": 5555840,
      "step": 9580
    },
    {
      "epoch": 1.4276139410187667,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.021411974977658624,
      "loss": 0.7859,
      "num_input_tokens_seen": 5558912,
      "step": 9585
    },
    {
      "epoch": 1.428358653559726,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.02142314566577301,
      "loss": 0.7643,
      "num_input_tokens_seen": 5561792,
      "step": 9590
    },
    {
      "epoch": 1.4291033661006851,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0214343163538874,
      "loss": 0.8261,
      "num_input_tokens_seen": 5564480,
      "step": 9595
    },
    {
      "epoch": 1.4298480786416443,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.021445487042001787,
      "loss": 0.8146,
      "num_input_tokens_seen": 5567264,
      "step": 9600
    },
    {
      "epoch": 1.4305927911826035,
      "grad_norm": 0.03125,
      "learning_rate": 0.021456657730116176,
      "loss": 0.8052,
      "num_input_tokens_seen": 5569984,
      "step": 9605
    },
    {
      "epoch": 1.4313375037235627,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.021467828418230563,
      "loss": 0.8312,
      "num_input_tokens_seen": 5573472,
      "step": 9610
    },
    {
      "epoch": 1.432082216264522,
      "grad_norm": 0.046875,
      "learning_rate": 0.02147899910634495,
      "loss": 0.8076,
      "num_input_tokens_seen": 5576416,
      "step": 9615
    },
    {
      "epoch": 1.4328269288054811,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.021490169794459335,
      "loss": 0.7883,
      "num_input_tokens_seen": 5579584,
      "step": 9620
    },
    {
      "epoch": 1.4335716413464403,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.021501340482573725,
      "loss": 0.8022,
      "num_input_tokens_seen": 5582464,
      "step": 9625
    },
    {
      "epoch": 1.4343163538873995,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.02151251117068811,
      "loss": 0.7902,
      "num_input_tokens_seen": 5585056,
      "step": 9630
    },
    {
      "epoch": 1.4350610664283587,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0215236818588025,
      "loss": 0.8092,
      "num_input_tokens_seen": 5587968,
      "step": 9635
    },
    {
      "epoch": 1.4358057789693177,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.021534852546916887,
      "loss": 0.8146,
      "num_input_tokens_seen": 5591136,
      "step": 9640
    },
    {
      "epoch": 1.436550491510277,
      "grad_norm": 0.02734375,
      "learning_rate": 0.021546023235031277,
      "loss": 0.7889,
      "num_input_tokens_seen": 5594016,
      "step": 9645
    },
    {
      "epoch": 1.4372952040512361,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.021557193923145666,
      "loss": 0.7713,
      "num_input_tokens_seen": 5596736,
      "step": 9650
    },
    {
      "epoch": 1.4380399165921953,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.021568364611260053,
      "loss": 0.8128,
      "num_input_tokens_seen": 5599552,
      "step": 9655
    },
    {
      "epoch": 1.4387846291331545,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.021579535299374442,
      "loss": 0.7509,
      "num_input_tokens_seen": 5602752,
      "step": 9660
    },
    {
      "epoch": 1.4395293416741137,
      "grad_norm": 0.06640625,
      "learning_rate": 0.02159070598748883,
      "loss": 0.816,
      "num_input_tokens_seen": 5605568,
      "step": 9665
    },
    {
      "epoch": 1.440274054215073,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.021601876675603218,
      "loss": 0.7872,
      "num_input_tokens_seen": 5608448,
      "step": 9670
    },
    {
      "epoch": 1.4410187667560321,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.021613047363717604,
      "loss": 0.8099,
      "num_input_tokens_seen": 5611584,
      "step": 9675
    },
    {
      "epoch": 1.4417634792969913,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.021624218051831994,
      "loss": 0.7527,
      "num_input_tokens_seen": 5614080,
      "step": 9680
    },
    {
      "epoch": 1.4425081918379505,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.02163538873994638,
      "loss": 0.8037,
      "num_input_tokens_seen": 5617088,
      "step": 9685
    },
    {
      "epoch": 1.4432529043789097,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.02164655942806077,
      "loss": 0.7943,
      "num_input_tokens_seen": 5619872,
      "step": 9690
    },
    {
      "epoch": 1.443997616919869,
      "grad_norm": 0.033203125,
      "learning_rate": 0.021657730116175156,
      "loss": 0.7797,
      "num_input_tokens_seen": 5622496,
      "step": 9695
    },
    {
      "epoch": 1.4447423294608281,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.021668900804289543,
      "loss": 0.8429,
      "num_input_tokens_seen": 5625408,
      "step": 9700
    },
    {
      "epoch": 1.4454870420017873,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02168007149240393,
      "loss": 0.8188,
      "num_input_tokens_seen": 5628064,
      "step": 9705
    },
    {
      "epoch": 1.4462317545427466,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02169124218051832,
      "loss": 0.7953,
      "num_input_tokens_seen": 5631104,
      "step": 9710
    },
    {
      "epoch": 1.4469764670837058,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.021702412868632705,
      "loss": 0.8216,
      "num_input_tokens_seen": 5633920,
      "step": 9715
    },
    {
      "epoch": 1.447721179624665,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.021713583556747094,
      "loss": 0.7726,
      "num_input_tokens_seen": 5636672,
      "step": 9720
    },
    {
      "epoch": 1.4484658921656242,
      "grad_norm": 0.037109375,
      "learning_rate": 0.02172475424486148,
      "loss": 0.7927,
      "num_input_tokens_seen": 5639616,
      "step": 9725
    },
    {
      "epoch": 1.4492106047065834,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02173592493297587,
      "loss": 0.836,
      "num_input_tokens_seen": 5642880,
      "step": 9730
    },
    {
      "epoch": 1.4499553172475426,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.021747095621090257,
      "loss": 0.806,
      "num_input_tokens_seen": 5645728,
      "step": 9735
    },
    {
      "epoch": 1.4507000297885018,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.021758266309204646,
      "loss": 0.8072,
      "num_input_tokens_seen": 5648928,
      "step": 9740
    },
    {
      "epoch": 1.4514447423294607,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.021769436997319033,
      "loss": 0.803,
      "num_input_tokens_seen": 5651552,
      "step": 9745
    },
    {
      "epoch": 1.45218945487042,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.021780607685433422,
      "loss": 0.7837,
      "num_input_tokens_seen": 5654496,
      "step": 9750
    },
    {
      "epoch": 1.4529341674113792,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02179177837354781,
      "loss": 0.8001,
      "num_input_tokens_seen": 5657280,
      "step": 9755
    },
    {
      "epoch": 1.4536788799523384,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.021802949061662198,
      "loss": 0.8069,
      "num_input_tokens_seen": 5659872,
      "step": 9760
    },
    {
      "epoch": 1.4544235924932976,
      "grad_norm": 0.060546875,
      "learning_rate": 0.021814119749776588,
      "loss": 0.817,
      "num_input_tokens_seen": 5662752,
      "step": 9765
    },
    {
      "epoch": 1.4551683050342568,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.021825290437890974,
      "loss": 0.7762,
      "num_input_tokens_seen": 5665568,
      "step": 9770
    },
    {
      "epoch": 1.455913017575216,
      "grad_norm": 0.033203125,
      "learning_rate": 0.021836461126005364,
      "loss": 0.8044,
      "num_input_tokens_seen": 5668448,
      "step": 9775
    },
    {
      "epoch": 1.4566577301161752,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.02184763181411975,
      "loss": 0.8387,
      "num_input_tokens_seen": 5671200,
      "step": 9780
    },
    {
      "epoch": 1.4574024426571344,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02185880250223414,
      "loss": 0.7931,
      "num_input_tokens_seen": 5674144,
      "step": 9785
    },
    {
      "epoch": 1.4581471551980936,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.021869973190348523,
      "loss": 0.8012,
      "num_input_tokens_seen": 5677152,
      "step": 9790
    },
    {
      "epoch": 1.4588918677390528,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.021881143878462912,
      "loss": 0.809,
      "num_input_tokens_seen": 5680448,
      "step": 9795
    },
    {
      "epoch": 1.459636580280012,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0218923145665773,
      "loss": 0.7979,
      "num_input_tokens_seen": 5683424,
      "step": 9800
    },
    {
      "epoch": 1.4603812928209712,
      "grad_norm": 0.03515625,
      "learning_rate": 0.021903485254691688,
      "loss": 0.7865,
      "num_input_tokens_seen": 5686336,
      "step": 9805
    },
    {
      "epoch": 1.4611260053619302,
      "grad_norm": 0.037109375,
      "learning_rate": 0.021914655942806074,
      "loss": 0.8182,
      "num_input_tokens_seen": 5689120,
      "step": 9810
    },
    {
      "epoch": 1.4618707179028894,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.021925826630920464,
      "loss": 0.8129,
      "num_input_tokens_seen": 5691648,
      "step": 9815
    },
    {
      "epoch": 1.4626154304438486,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02193699731903485,
      "loss": 0.7991,
      "num_input_tokens_seen": 5694368,
      "step": 9820
    },
    {
      "epoch": 1.4633601429848078,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02194816800714924,
      "loss": 0.8092,
      "num_input_tokens_seen": 5697280,
      "step": 9825
    },
    {
      "epoch": 1.464104855525767,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.021959338695263626,
      "loss": 0.7967,
      "num_input_tokens_seen": 5700448,
      "step": 9830
    },
    {
      "epoch": 1.4648495680667262,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.021970509383378016,
      "loss": 0.8058,
      "num_input_tokens_seen": 5703584,
      "step": 9835
    },
    {
      "epoch": 1.4655942806076854,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.021981680071492402,
      "loss": 0.8067,
      "num_input_tokens_seen": 5706400,
      "step": 9840
    },
    {
      "epoch": 1.4663389931486446,
      "grad_norm": 0.052734375,
      "learning_rate": 0.021992850759606792,
      "loss": 0.8097,
      "num_input_tokens_seen": 5709184,
      "step": 9845
    },
    {
      "epoch": 1.4670837056896038,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.022004021447721178,
      "loss": 0.7842,
      "num_input_tokens_seen": 5711712,
      "step": 9850
    },
    {
      "epoch": 1.467828418230563,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.022015192135835568,
      "loss": 0.8094,
      "num_input_tokens_seen": 5714464,
      "step": 9855
    },
    {
      "epoch": 1.4685731307715222,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.022026362823949954,
      "loss": 0.8076,
      "num_input_tokens_seen": 5717216,
      "step": 9860
    },
    {
      "epoch": 1.4693178433124814,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.022037533512064344,
      "loss": 0.7929,
      "num_input_tokens_seen": 5720608,
      "step": 9865
    },
    {
      "epoch": 1.4700625558534406,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.02204870420017873,
      "loss": 0.8132,
      "num_input_tokens_seen": 5723488,
      "step": 9870
    },
    {
      "epoch": 1.4708072683943998,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.022059874888293116,
      "loss": 0.8231,
      "num_input_tokens_seen": 5726624,
      "step": 9875
    },
    {
      "epoch": 1.471551980935359,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.022071045576407506,
      "loss": 0.809,
      "num_input_tokens_seen": 5729408,
      "step": 9880
    },
    {
      "epoch": 1.4722966934763182,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.022082216264521892,
      "loss": 0.7895,
      "num_input_tokens_seen": 5732352,
      "step": 9885
    },
    {
      "epoch": 1.4730414060172774,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.022093386952636282,
      "loss": 0.8233,
      "num_input_tokens_seen": 5735072,
      "step": 9890
    },
    {
      "epoch": 1.4737861185582366,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.022104557640750668,
      "loss": 0.7971,
      "num_input_tokens_seen": 5738272,
      "step": 9895
    },
    {
      "epoch": 1.4745308310991958,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.022115728328865058,
      "loss": 0.7795,
      "num_input_tokens_seen": 5741024,
      "step": 9900
    },
    {
      "epoch": 1.475275543640155,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.022126899016979444,
      "loss": 0.8002,
      "num_input_tokens_seen": 5743968,
      "step": 9905
    },
    {
      "epoch": 1.4760202561811142,
      "grad_norm": 0.0390625,
      "learning_rate": 0.022138069705093834,
      "loss": 0.8096,
      "num_input_tokens_seen": 5746816,
      "step": 9910
    },
    {
      "epoch": 1.4767649687220734,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.02214924039320822,
      "loss": 0.8151,
      "num_input_tokens_seen": 5749312,
      "step": 9915
    },
    {
      "epoch": 1.4775096812630324,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02216041108132261,
      "loss": 0.79,
      "num_input_tokens_seen": 5751776,
      "step": 9920
    },
    {
      "epoch": 1.4782543938039916,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.022171581769436996,
      "loss": 0.8015,
      "num_input_tokens_seen": 5754432,
      "step": 9925
    },
    {
      "epoch": 1.4789991063449508,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.022182752457551386,
      "loss": 0.7875,
      "num_input_tokens_seen": 5757152,
      "step": 9930
    },
    {
      "epoch": 1.47974381888591,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.022193923145665772,
      "loss": 0.8074,
      "num_input_tokens_seen": 5760096,
      "step": 9935
    },
    {
      "epoch": 1.4804885314268692,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02220509383378016,
      "loss": 0.7856,
      "num_input_tokens_seen": 5762880,
      "step": 9940
    },
    {
      "epoch": 1.4812332439678284,
      "grad_norm": 0.033203125,
      "learning_rate": 0.022216264521894548,
      "loss": 0.7793,
      "num_input_tokens_seen": 5765696,
      "step": 9945
    },
    {
      "epoch": 1.4819779565087876,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.022227435210008938,
      "loss": 0.8054,
      "num_input_tokens_seen": 5768448,
      "step": 9950
    },
    {
      "epoch": 1.4827226690497468,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.022238605898123324,
      "loss": 0.8354,
      "num_input_tokens_seen": 5771520,
      "step": 9955
    },
    {
      "epoch": 1.483467381590706,
      "grad_norm": 0.056640625,
      "learning_rate": 0.02224977658623771,
      "loss": 0.8282,
      "num_input_tokens_seen": 5774400,
      "step": 9960
    },
    {
      "epoch": 1.4842120941316652,
      "grad_norm": 0.037109375,
      "learning_rate": 0.022260947274352096,
      "loss": 0.7934,
      "num_input_tokens_seen": 5777216,
      "step": 9965
    },
    {
      "epoch": 1.4849568066726244,
      "grad_norm": 0.044921875,
      "learning_rate": 0.022272117962466486,
      "loss": 0.7901,
      "num_input_tokens_seen": 5779744,
      "step": 9970
    },
    {
      "epoch": 1.4857015192135836,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.022283288650580872,
      "loss": 0.8087,
      "num_input_tokens_seen": 5782720,
      "step": 9975
    },
    {
      "epoch": 1.4864462317545428,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.022294459338695262,
      "loss": 0.8176,
      "num_input_tokens_seen": 5785344,
      "step": 9980
    },
    {
      "epoch": 1.4871909442955018,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02230563002680965,
      "loss": 0.7723,
      "num_input_tokens_seen": 5787936,
      "step": 9985
    },
    {
      "epoch": 1.487935656836461,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.022316800714924038,
      "loss": 0.8339,
      "num_input_tokens_seen": 5790880,
      "step": 9990
    },
    {
      "epoch": 1.4886803693774202,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.022327971403038428,
      "loss": 0.8205,
      "num_input_tokens_seen": 5793920,
      "step": 9995
    },
    {
      "epoch": 1.4894250819183794,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.022339142091152814,
      "loss": 0.8047,
      "num_input_tokens_seen": 5796768,
      "step": 10000
    },
    {
      "epoch": 1.4901697944593386,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.022350312779267204,
      "loss": 0.8058,
      "num_input_tokens_seen": 5799520,
      "step": 10005
    },
    {
      "epoch": 1.4909145070002978,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02236148346738159,
      "loss": 0.8062,
      "num_input_tokens_seen": 5802336,
      "step": 10010
    },
    {
      "epoch": 1.491659219541257,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02237265415549598,
      "loss": 0.7878,
      "num_input_tokens_seen": 5805600,
      "step": 10015
    },
    {
      "epoch": 1.4924039320822162,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.022383824843610366,
      "loss": 0.8056,
      "num_input_tokens_seen": 5808320,
      "step": 10020
    },
    {
      "epoch": 1.4931486446231754,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.022394995531724755,
      "loss": 0.8029,
      "num_input_tokens_seen": 5811264,
      "step": 10025
    },
    {
      "epoch": 1.4938933571641346,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02240616621983914,
      "loss": 0.8179,
      "num_input_tokens_seen": 5813920,
      "step": 10030
    },
    {
      "epoch": 1.4946380697050938,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02241733690795353,
      "loss": 0.8145,
      "num_input_tokens_seen": 5817024,
      "step": 10035
    },
    {
      "epoch": 1.495382782246053,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.022428507596067918,
      "loss": 0.8031,
      "num_input_tokens_seen": 5820000,
      "step": 10040
    },
    {
      "epoch": 1.4961274947870122,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.022439678284182307,
      "loss": 0.7921,
      "num_input_tokens_seen": 5822784,
      "step": 10045
    },
    {
      "epoch": 1.4968722073279714,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02245084897229669,
      "loss": 0.7767,
      "num_input_tokens_seen": 5825568,
      "step": 10050
    },
    {
      "epoch": 1.4976169198689306,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02246201966041108,
      "loss": 0.7997,
      "num_input_tokens_seen": 5828288,
      "step": 10055
    },
    {
      "epoch": 1.4983616324098898,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.022473190348525466,
      "loss": 0.7986,
      "num_input_tokens_seen": 5831360,
      "step": 10060
    },
    {
      "epoch": 1.499106344950849,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.022484361036639856,
      "loss": 0.8334,
      "num_input_tokens_seen": 5834208,
      "step": 10065
    },
    {
      "epoch": 1.4998510574918082,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.022495531724754242,
      "loss": 0.8209,
      "num_input_tokens_seen": 5837184,
      "step": 10070
    },
    {
      "epoch": 1.5005957700327675,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02250670241286863,
      "loss": 0.8168,
      "num_input_tokens_seen": 5840096,
      "step": 10075
    },
    {
      "epoch": 1.5013404825737267,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.022517873100983018,
      "loss": 0.8106,
      "num_input_tokens_seen": 5843584,
      "step": 10080
    },
    {
      "epoch": 1.5020851951146859,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.022529043789097408,
      "loss": 0.7996,
      "num_input_tokens_seen": 5846144,
      "step": 10085
    },
    {
      "epoch": 1.502829907655645,
      "grad_norm": 0.017578125,
      "learning_rate": 0.022540214477211794,
      "loss": 0.8053,
      "num_input_tokens_seen": 5848960,
      "step": 10090
    },
    {
      "epoch": 1.5035746201966043,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.022551385165326183,
      "loss": 0.8197,
      "num_input_tokens_seen": 5851744,
      "step": 10095
    },
    {
      "epoch": 1.5043193327375635,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.022562555853440573,
      "loss": 0.8019,
      "num_input_tokens_seen": 5854848,
      "step": 10100
    },
    {
      "epoch": 1.5050640452785224,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02257372654155496,
      "loss": 0.8079,
      "num_input_tokens_seen": 5857792,
      "step": 10105
    },
    {
      "epoch": 1.5058087578194816,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02258489722966935,
      "loss": 0.8005,
      "num_input_tokens_seen": 5860672,
      "step": 10110
    },
    {
      "epoch": 1.5065534703604408,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.022596067917783735,
      "loss": 0.8121,
      "num_input_tokens_seen": 5863552,
      "step": 10115
    },
    {
      "epoch": 1.5072981829014,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.022607238605898125,
      "loss": 0.8057,
      "num_input_tokens_seen": 5866592,
      "step": 10120
    },
    {
      "epoch": 1.5080428954423593,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02261840929401251,
      "loss": 0.7915,
      "num_input_tokens_seen": 5869408,
      "step": 10125
    },
    {
      "epoch": 1.5087876079833185,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.0226295799821269,
      "loss": 0.8001,
      "num_input_tokens_seen": 5872128,
      "step": 10130
    },
    {
      "epoch": 1.5095323205242777,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.022640750670241284,
      "loss": 0.7877,
      "num_input_tokens_seen": 5875104,
      "step": 10135
    },
    {
      "epoch": 1.5102770330652369,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.022651921358355673,
      "loss": 0.7987,
      "num_input_tokens_seen": 5878176,
      "step": 10140
    },
    {
      "epoch": 1.5110217456061958,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02266309204647006,
      "loss": 0.8177,
      "num_input_tokens_seen": 5881088,
      "step": 10145
    },
    {
      "epoch": 1.511766458147155,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.02267426273458445,
      "loss": 0.7949,
      "num_input_tokens_seen": 5883968,
      "step": 10150
    },
    {
      "epoch": 1.5125111706881142,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.022685433422698836,
      "loss": 0.7735,
      "num_input_tokens_seen": 5887104,
      "step": 10155
    },
    {
      "epoch": 1.5132558832290735,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.022696604110813225,
      "loss": 0.7843,
      "num_input_tokens_seen": 5889952,
      "step": 10160
    },
    {
      "epoch": 1.5140005957700327,
      "grad_norm": 0.056640625,
      "learning_rate": 0.02270777479892761,
      "loss": 0.8286,
      "num_input_tokens_seen": 5892896,
      "step": 10165
    },
    {
      "epoch": 1.5147453083109919,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.022718945487042,
      "loss": 0.7698,
      "num_input_tokens_seen": 5895680,
      "step": 10170
    },
    {
      "epoch": 1.515490020851951,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.022730116175156388,
      "loss": 0.7947,
      "num_input_tokens_seen": 5898560,
      "step": 10175
    },
    {
      "epoch": 1.5162347333929103,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.022741286863270777,
      "loss": 0.7885,
      "num_input_tokens_seen": 5901696,
      "step": 10180
    },
    {
      "epoch": 1.5169794459338695,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.022752457551385163,
      "loss": 0.8151,
      "num_input_tokens_seen": 5904384,
      "step": 10185
    },
    {
      "epoch": 1.5177241584748287,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.022763628239499553,
      "loss": 0.7851,
      "num_input_tokens_seen": 5907424,
      "step": 10190
    },
    {
      "epoch": 1.5184688710157879,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.02277479892761394,
      "loss": 0.7708,
      "num_input_tokens_seen": 5910464,
      "step": 10195
    },
    {
      "epoch": 1.519213583556747,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.02278596961572833,
      "loss": 0.8063,
      "num_input_tokens_seen": 5913504,
      "step": 10200
    },
    {
      "epoch": 1.5199582960977063,
      "grad_norm": 0.048828125,
      "learning_rate": 0.022797140303842715,
      "loss": 0.7989,
      "num_input_tokens_seen": 5916544,
      "step": 10205
    },
    {
      "epoch": 1.5207030086386655,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.022808310991957105,
      "loss": 0.8429,
      "num_input_tokens_seen": 5919744,
      "step": 10210
    },
    {
      "epoch": 1.5214477211796247,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.022819481680071495,
      "loss": 0.8228,
      "num_input_tokens_seen": 5922656,
      "step": 10215
    },
    {
      "epoch": 1.5221924337205839,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.022830652368185878,
      "loss": 0.7976,
      "num_input_tokens_seen": 5925376,
      "step": 10220
    },
    {
      "epoch": 1.522937146261543,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.022841823056300267,
      "loss": 0.7982,
      "num_input_tokens_seen": 5928160,
      "step": 10225
    },
    {
      "epoch": 1.5236818588025023,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.022852993744414653,
      "loss": 0.799,
      "num_input_tokens_seen": 5930848,
      "step": 10230
    },
    {
      "epoch": 1.5244265713434615,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.022864164432529043,
      "loss": 0.8075,
      "num_input_tokens_seen": 5933760,
      "step": 10235
    },
    {
      "epoch": 1.5251712838844207,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02287533512064343,
      "loss": 0.7915,
      "num_input_tokens_seen": 5936544,
      "step": 10240
    },
    {
      "epoch": 1.52591599642538,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.02288650580875782,
      "loss": 0.7914,
      "num_input_tokens_seen": 5939520,
      "step": 10245
    },
    {
      "epoch": 1.526660708966339,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.022897676496872205,
      "loss": 0.7904,
      "num_input_tokens_seen": 5942528,
      "step": 10250
    },
    {
      "epoch": 1.5274054215072983,
      "grad_norm": 0.03515625,
      "learning_rate": 0.022908847184986595,
      "loss": 0.7925,
      "num_input_tokens_seen": 5945888,
      "step": 10255
    },
    {
      "epoch": 1.5281501340482575,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02292001787310098,
      "loss": 0.783,
      "num_input_tokens_seen": 5949088,
      "step": 10260
    },
    {
      "epoch": 1.5288948465892167,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02293118856121537,
      "loss": 0.8166,
      "num_input_tokens_seen": 5952064,
      "step": 10265
    },
    {
      "epoch": 1.529639559130176,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.022942359249329757,
      "loss": 0.8442,
      "num_input_tokens_seen": 5954944,
      "step": 10270
    },
    {
      "epoch": 1.5303842716711349,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.022953529937444147,
      "loss": 0.7922,
      "num_input_tokens_seen": 5957728,
      "step": 10275
    },
    {
      "epoch": 1.531128984212094,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.022964700625558533,
      "loss": 0.8029,
      "num_input_tokens_seen": 5960736,
      "step": 10280
    },
    {
      "epoch": 1.5318736967530533,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.022975871313672923,
      "loss": 0.8142,
      "num_input_tokens_seen": 5963584,
      "step": 10285
    },
    {
      "epoch": 1.5326184092940125,
      "grad_norm": 0.056640625,
      "learning_rate": 0.02298704200178731,
      "loss": 0.7864,
      "num_input_tokens_seen": 5966432,
      "step": 10290
    },
    {
      "epoch": 1.5333631218349717,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.0229982126899017,
      "loss": 0.7923,
      "num_input_tokens_seen": 5969248,
      "step": 10295
    },
    {
      "epoch": 1.534107834375931,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.023009383378016085,
      "loss": 0.7892,
      "num_input_tokens_seen": 5972064,
      "step": 10300
    },
    {
      "epoch": 1.53485254691689,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.023020554066130475,
      "loss": 0.7987,
      "num_input_tokens_seen": 5974976,
      "step": 10305
    },
    {
      "epoch": 1.5355972594578493,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.023031724754244857,
      "loss": 0.8282,
      "num_input_tokens_seen": 5977728,
      "step": 10310
    },
    {
      "epoch": 1.5363419719988085,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.023042895442359247,
      "loss": 0.8141,
      "num_input_tokens_seen": 5980512,
      "step": 10315
    },
    {
      "epoch": 1.5370866845397675,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.023054066130473637,
      "loss": 0.7931,
      "num_input_tokens_seen": 5983616,
      "step": 10320
    },
    {
      "epoch": 1.5378313970807267,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.023065236818588023,
      "loss": 0.7989,
      "num_input_tokens_seen": 5986464,
      "step": 10325
    },
    {
      "epoch": 1.538576109621686,
      "grad_norm": 0.03125,
      "learning_rate": 0.023076407506702413,
      "loss": 0.8023,
      "num_input_tokens_seen": 5989632,
      "step": 10330
    },
    {
      "epoch": 1.539320822162645,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.0230875781948168,
      "loss": 0.8054,
      "num_input_tokens_seen": 5992736,
      "step": 10335
    },
    {
      "epoch": 1.5400655347036043,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02309874888293119,
      "loss": 0.787,
      "num_input_tokens_seen": 5995424,
      "step": 10340
    },
    {
      "epoch": 1.5408102472445635,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.023109919571045575,
      "loss": 0.8307,
      "num_input_tokens_seen": 5998176,
      "step": 10345
    },
    {
      "epoch": 1.5415549597855227,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.023121090259159965,
      "loss": 0.7937,
      "num_input_tokens_seen": 6001408,
      "step": 10350
    },
    {
      "epoch": 1.542299672326482,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.02313226094727435,
      "loss": 0.819,
      "num_input_tokens_seen": 6004064,
      "step": 10355
    },
    {
      "epoch": 1.543044384867441,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02314343163538874,
      "loss": 0.8026,
      "num_input_tokens_seen": 6006944,
      "step": 10360
    },
    {
      "epoch": 1.5437890974084003,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.023154602323503127,
      "loss": 0.8106,
      "num_input_tokens_seen": 6009984,
      "step": 10365
    },
    {
      "epoch": 1.5445338099493595,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.023165773011617517,
      "loss": 0.7798,
      "num_input_tokens_seen": 6012864,
      "step": 10370
    },
    {
      "epoch": 1.5452785224903187,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.023176943699731903,
      "loss": 0.8059,
      "num_input_tokens_seen": 6015680,
      "step": 10375
    },
    {
      "epoch": 1.546023235031278,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.023188114387846293,
      "loss": 0.7783,
      "num_input_tokens_seen": 6018496,
      "step": 10380
    },
    {
      "epoch": 1.5467679475722371,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02319928507596068,
      "loss": 0.7957,
      "num_input_tokens_seen": 6021600,
      "step": 10385
    },
    {
      "epoch": 1.5475126601131963,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02321045576407507,
      "loss": 0.8155,
      "num_input_tokens_seen": 6024480,
      "step": 10390
    },
    {
      "epoch": 1.5482573726541555,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02322162645218945,
      "loss": 0.7903,
      "num_input_tokens_seen": 6027456,
      "step": 10395
    },
    {
      "epoch": 1.5490020851951147,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02323279714030384,
      "loss": 0.806,
      "num_input_tokens_seen": 6030432,
      "step": 10400
    },
    {
      "epoch": 1.549746797736074,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.023243967828418227,
      "loss": 0.824,
      "num_input_tokens_seen": 6033376,
      "step": 10405
    },
    {
      "epoch": 1.5504915102770331,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.023255138516532617,
      "loss": 0.8097,
      "num_input_tokens_seen": 6036256,
      "step": 10410
    },
    {
      "epoch": 1.5512362228179923,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.023266309204647003,
      "loss": 0.7824,
      "num_input_tokens_seen": 6039232,
      "step": 10415
    },
    {
      "epoch": 1.5519809353589515,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.023277479892761393,
      "loss": 0.8344,
      "num_input_tokens_seen": 6042240,
      "step": 10420
    },
    {
      "epoch": 1.5527256478999107,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02328865058087578,
      "loss": 0.7857,
      "num_input_tokens_seen": 6045024,
      "step": 10425
    },
    {
      "epoch": 1.55347036044087,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.02329982126899017,
      "loss": 0.8067,
      "num_input_tokens_seen": 6047968,
      "step": 10430
    },
    {
      "epoch": 1.5542150729818291,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.02331099195710456,
      "loss": 0.7747,
      "num_input_tokens_seen": 6051360,
      "step": 10435
    },
    {
      "epoch": 1.5549597855227884,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.023322162645218945,
      "loss": 0.7967,
      "num_input_tokens_seen": 6054080,
      "step": 10440
    },
    {
      "epoch": 1.5557044980637476,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.023333333333333334,
      "loss": 0.8117,
      "num_input_tokens_seen": 6056672,
      "step": 10445
    },
    {
      "epoch": 1.5564492106047065,
      "grad_norm": 0.03125,
      "learning_rate": 0.02334450402144772,
      "loss": 0.8189,
      "num_input_tokens_seen": 6059648,
      "step": 10450
    },
    {
      "epoch": 1.5571939231456657,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02335567470956211,
      "loss": 0.8168,
      "num_input_tokens_seen": 6062496,
      "step": 10455
    },
    {
      "epoch": 1.557938635686625,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.023366845397676497,
      "loss": 0.8028,
      "num_input_tokens_seen": 6065344,
      "step": 10460
    },
    {
      "epoch": 1.5586833482275841,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.023378016085790886,
      "loss": 0.806,
      "num_input_tokens_seen": 6068064,
      "step": 10465
    },
    {
      "epoch": 1.5594280607685433,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.023389186773905273,
      "loss": 0.7943,
      "num_input_tokens_seen": 6070944,
      "step": 10470
    },
    {
      "epoch": 1.5601727733095025,
      "grad_norm": 0.05859375,
      "learning_rate": 0.023400357462019662,
      "loss": 0.7999,
      "num_input_tokens_seen": 6073856,
      "step": 10475
    },
    {
      "epoch": 1.5609174858504618,
      "grad_norm": 0.0390625,
      "learning_rate": 0.023411528150134045,
      "loss": 0.8105,
      "num_input_tokens_seen": 6076608,
      "step": 10480
    },
    {
      "epoch": 1.561662198391421,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.023422698838248435,
      "loss": 0.7808,
      "num_input_tokens_seen": 6079424,
      "step": 10485
    },
    {
      "epoch": 1.5624069109323802,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.02343386952636282,
      "loss": 0.7896,
      "num_input_tokens_seen": 6082432,
      "step": 10490
    },
    {
      "epoch": 1.5631516234733391,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.02344504021447721,
      "loss": 0.804,
      "num_input_tokens_seen": 6085472,
      "step": 10495
    },
    {
      "epoch": 1.5638963360142983,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.023456210902591597,
      "loss": 0.7699,
      "num_input_tokens_seen": 6088448,
      "step": 10500
    },
    {
      "epoch": 1.5646410485552575,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.023467381590705987,
      "loss": 0.8267,
      "num_input_tokens_seen": 6091552,
      "step": 10505
    },
    {
      "epoch": 1.5653857610962167,
      "grad_norm": 0.064453125,
      "learning_rate": 0.023478552278820373,
      "loss": 0.8066,
      "num_input_tokens_seen": 6094496,
      "step": 10510
    },
    {
      "epoch": 1.566130473637176,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.023489722966934762,
      "loss": 0.7898,
      "num_input_tokens_seen": 6097248,
      "step": 10515
    },
    {
      "epoch": 1.5668751861781351,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02350089365504915,
      "loss": 0.8241,
      "num_input_tokens_seen": 6100000,
      "step": 10520
    },
    {
      "epoch": 1.5676198987190944,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02351206434316354,
      "loss": 0.8103,
      "num_input_tokens_seen": 6103104,
      "step": 10525
    },
    {
      "epoch": 1.5683646112600536,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.023523235031277925,
      "loss": 0.7883,
      "num_input_tokens_seen": 6105984,
      "step": 10530
    },
    {
      "epoch": 1.5691093238010128,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.023534405719392314,
      "loss": 0.7994,
      "num_input_tokens_seen": 6108736,
      "step": 10535
    },
    {
      "epoch": 1.569854036341972,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.0235455764075067,
      "loss": 0.7936,
      "num_input_tokens_seen": 6111840,
      "step": 10540
    },
    {
      "epoch": 1.5705987488829312,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02355674709562109,
      "loss": 0.772,
      "num_input_tokens_seen": 6114848,
      "step": 10545
    },
    {
      "epoch": 1.5713434614238904,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02356791778373548,
      "loss": 0.7828,
      "num_input_tokens_seen": 6117600,
      "step": 10550
    },
    {
      "epoch": 1.5720881739648496,
      "grad_norm": 0.041015625,
      "learning_rate": 0.023579088471849866,
      "loss": 0.7732,
      "num_input_tokens_seen": 6120160,
      "step": 10555
    },
    {
      "epoch": 1.5728328865058088,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.023590259159964256,
      "loss": 0.7904,
      "num_input_tokens_seen": 6122848,
      "step": 10560
    },
    {
      "epoch": 1.573577599046768,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.023601429848078642,
      "loss": 0.8061,
      "num_input_tokens_seen": 6125728,
      "step": 10565
    },
    {
      "epoch": 1.5743223115877272,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02361260053619303,
      "loss": 0.7849,
      "num_input_tokens_seen": 6128448,
      "step": 10570
    },
    {
      "epoch": 1.5750670241286864,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.023623771224307415,
      "loss": 0.8014,
      "num_input_tokens_seen": 6131584,
      "step": 10575
    },
    {
      "epoch": 1.5758117366696456,
      "grad_norm": 0.029296875,
      "learning_rate": 0.023634941912421804,
      "loss": 0.7901,
      "num_input_tokens_seen": 6134400,
      "step": 10580
    },
    {
      "epoch": 1.5765564492106048,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02364611260053619,
      "loss": 0.7793,
      "num_input_tokens_seen": 6137056,
      "step": 10585
    },
    {
      "epoch": 1.577301161751564,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02365728328865058,
      "loss": 0.7644,
      "num_input_tokens_seen": 6139744,
      "step": 10590
    },
    {
      "epoch": 1.5780458742925232,
      "grad_norm": 0.033203125,
      "learning_rate": 0.023668453976764967,
      "loss": 0.8439,
      "num_input_tokens_seen": 6142432,
      "step": 10595
    },
    {
      "epoch": 1.5787905868334824,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.023679624664879356,
      "loss": 0.8101,
      "num_input_tokens_seen": 6145248,
      "step": 10600
    },
    {
      "epoch": 1.5795352993744416,
      "grad_norm": 0.04296875,
      "learning_rate": 0.023690795352993742,
      "loss": 0.7701,
      "num_input_tokens_seen": 6148256,
      "step": 10605
    },
    {
      "epoch": 1.5802800119154008,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.023701966041108132,
      "loss": 0.7709,
      "num_input_tokens_seen": 6151232,
      "step": 10610
    },
    {
      "epoch": 1.58102472445636,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02371313672922252,
      "loss": 0.8297,
      "num_input_tokens_seen": 6154144,
      "step": 10615
    },
    {
      "epoch": 1.5817694369973192,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.023724307417336908,
      "loss": 0.8267,
      "num_input_tokens_seen": 6157216,
      "step": 10620
    },
    {
      "epoch": 1.5825141495382782,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.023735478105451294,
      "loss": 0.8087,
      "num_input_tokens_seen": 6160128,
      "step": 10625
    },
    {
      "epoch": 1.5832588620792374,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.023746648793565684,
      "loss": 0.8305,
      "num_input_tokens_seen": 6163040,
      "step": 10630
    },
    {
      "epoch": 1.5840035746201966,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02375781948168007,
      "loss": 0.8215,
      "num_input_tokens_seen": 6165824,
      "step": 10635
    },
    {
      "epoch": 1.5847482871611558,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02376899016979446,
      "loss": 0.7737,
      "num_input_tokens_seen": 6168800,
      "step": 10640
    },
    {
      "epoch": 1.585492999702115,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.023780160857908846,
      "loss": 0.8204,
      "num_input_tokens_seen": 6171648,
      "step": 10645
    },
    {
      "epoch": 1.5862377122430742,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.023791331546023236,
      "loss": 0.7875,
      "num_input_tokens_seen": 6174688,
      "step": 10650
    },
    {
      "epoch": 1.5869824247840334,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.023802502234137622,
      "loss": 0.8316,
      "num_input_tokens_seen": 6177664,
      "step": 10655
    },
    {
      "epoch": 1.5877271373249926,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02381367292225201,
      "loss": 0.7973,
      "num_input_tokens_seen": 6180672,
      "step": 10660
    },
    {
      "epoch": 1.5884718498659516,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.023824843610366398,
      "loss": 0.8004,
      "num_input_tokens_seen": 6183616,
      "step": 10665
    },
    {
      "epoch": 1.5892165624069108,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.023836014298480784,
      "loss": 0.7961,
      "num_input_tokens_seen": 6186368,
      "step": 10670
    },
    {
      "epoch": 1.58996127494787,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.023847184986595174,
      "loss": 0.7928,
      "num_input_tokens_seen": 6189344,
      "step": 10675
    },
    {
      "epoch": 1.5907059874888292,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02385835567470956,
      "loss": 0.8009,
      "num_input_tokens_seen": 6192128,
      "step": 10680
    },
    {
      "epoch": 1.5914507000297884,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02386952636282395,
      "loss": 0.8063,
      "num_input_tokens_seen": 6195264,
      "step": 10685
    },
    {
      "epoch": 1.5921954125707476,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.023880697050938336,
      "loss": 0.8136,
      "num_input_tokens_seen": 6197920,
      "step": 10690
    },
    {
      "epoch": 1.5929401251117068,
      "grad_norm": 0.03125,
      "learning_rate": 0.023891867739052726,
      "loss": 0.8191,
      "num_input_tokens_seen": 6201088,
      "step": 10695
    },
    {
      "epoch": 1.593684837652666,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.023903038427167112,
      "loss": 0.8257,
      "num_input_tokens_seen": 6204224,
      "step": 10700
    },
    {
      "epoch": 1.5944295501936252,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.023914209115281502,
      "loss": 0.8079,
      "num_input_tokens_seen": 6207040,
      "step": 10705
    },
    {
      "epoch": 1.5951742627345844,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.023925379803395888,
      "loss": 0.816,
      "num_input_tokens_seen": 6209952,
      "step": 10710
    },
    {
      "epoch": 1.5959189752755436,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.023936550491510278,
      "loss": 0.8112,
      "num_input_tokens_seen": 6212896,
      "step": 10715
    },
    {
      "epoch": 1.5966636878165028,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.023947721179624664,
      "loss": 0.82,
      "num_input_tokens_seen": 6216448,
      "step": 10720
    },
    {
      "epoch": 1.597408400357462,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.023958891867739054,
      "loss": 0.7993,
      "num_input_tokens_seen": 6219296,
      "step": 10725
    },
    {
      "epoch": 1.5981531128984212,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02397006255585344,
      "loss": 0.7961,
      "num_input_tokens_seen": 6222528,
      "step": 10730
    },
    {
      "epoch": 1.5988978254393804,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02398123324396783,
      "loss": 0.7906,
      "num_input_tokens_seen": 6225184,
      "step": 10735
    },
    {
      "epoch": 1.5996425379803396,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.023992403932082212,
      "loss": 0.8065,
      "num_input_tokens_seen": 6227936,
      "step": 10740
    },
    {
      "epoch": 1.6003872505212988,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.024003574620196602,
      "loss": 0.8425,
      "num_input_tokens_seen": 6230688,
      "step": 10745
    },
    {
      "epoch": 1.601131963062258,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02401474530831099,
      "loss": 0.7937,
      "num_input_tokens_seen": 6233632,
      "step": 10750
    },
    {
      "epoch": 1.6018766756032172,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.024025915996425378,
      "loss": 0.7963,
      "num_input_tokens_seen": 6236480,
      "step": 10755
    },
    {
      "epoch": 1.6026213881441764,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.024037086684539764,
      "loss": 0.824,
      "num_input_tokens_seen": 6239456,
      "step": 10760
    },
    {
      "epoch": 1.6033661006851356,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.024048257372654154,
      "loss": 0.8091,
      "num_input_tokens_seen": 6242368,
      "step": 10765
    },
    {
      "epoch": 1.6041108132260948,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.024059428060768544,
      "loss": 0.7941,
      "num_input_tokens_seen": 6245152,
      "step": 10770
    },
    {
      "epoch": 1.604855525767054,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02407059874888293,
      "loss": 0.8091,
      "num_input_tokens_seen": 6248032,
      "step": 10775
    },
    {
      "epoch": 1.6056002383080132,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02408176943699732,
      "loss": 0.8025,
      "num_input_tokens_seen": 6251008,
      "step": 10780
    },
    {
      "epoch": 1.6063449508489724,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.024092940125111706,
      "loss": 0.7897,
      "num_input_tokens_seen": 6254240,
      "step": 10785
    },
    {
      "epoch": 1.6070896633899316,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.024104110813226096,
      "loss": 0.7923,
      "num_input_tokens_seen": 6257376,
      "step": 10790
    },
    {
      "epoch": 1.6078343759308906,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.024115281501340482,
      "loss": 0.8181,
      "num_input_tokens_seen": 6260608,
      "step": 10795
    },
    {
      "epoch": 1.6085790884718498,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.02412645218945487,
      "loss": 0.7893,
      "num_input_tokens_seen": 6263200,
      "step": 10800
    },
    {
      "epoch": 1.609323801012809,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.024137622877569258,
      "loss": 0.7931,
      "num_input_tokens_seen": 6266208,
      "step": 10805
    },
    {
      "epoch": 1.6100685135537682,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.024148793565683647,
      "loss": 0.7882,
      "num_input_tokens_seen": 6269024,
      "step": 10810
    },
    {
      "epoch": 1.6108132260947274,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.024159964253798034,
      "loss": 0.7912,
      "num_input_tokens_seen": 6272096,
      "step": 10815
    },
    {
      "epoch": 1.6115579386356866,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.024171134941912423,
      "loss": 0.794,
      "num_input_tokens_seen": 6274944,
      "step": 10820
    },
    {
      "epoch": 1.6123026511766458,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.024182305630026806,
      "loss": 0.8065,
      "num_input_tokens_seen": 6277856,
      "step": 10825
    },
    {
      "epoch": 1.613047363717605,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.024193476318141196,
      "loss": 0.8087,
      "num_input_tokens_seen": 6280544,
      "step": 10830
    },
    {
      "epoch": 1.6137920762585642,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.024204647006255582,
      "loss": 0.8172,
      "num_input_tokens_seen": 6283232,
      "step": 10835
    },
    {
      "epoch": 1.6145367887995232,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.024215817694369972,
      "loss": 0.805,
      "num_input_tokens_seen": 6286144,
      "step": 10840
    },
    {
      "epoch": 1.6152815013404824,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.024226988382484358,
      "loss": 0.8142,
      "num_input_tokens_seen": 6288864,
      "step": 10845
    },
    {
      "epoch": 1.6160262138814416,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.024238159070598748,
      "loss": 0.812,
      "num_input_tokens_seen": 6291872,
      "step": 10850
    },
    {
      "epoch": 1.6167709264224008,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.024249329758713134,
      "loss": 0.8114,
      "num_input_tokens_seen": 6294880,
      "step": 10855
    },
    {
      "epoch": 1.61751563896336,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.024260500446827524,
      "loss": 0.8147,
      "num_input_tokens_seen": 6297824,
      "step": 10860
    },
    {
      "epoch": 1.6182603515043192,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02427167113494191,
      "loss": 0.7999,
      "num_input_tokens_seen": 6300576,
      "step": 10865
    },
    {
      "epoch": 1.6190050640452784,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.0242828418230563,
      "loss": 0.8137,
      "num_input_tokens_seen": 6303616,
      "step": 10870
    },
    {
      "epoch": 1.6197497765862376,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.024294012511170686,
      "loss": 0.8174,
      "num_input_tokens_seen": 6306592,
      "step": 10875
    },
    {
      "epoch": 1.6204944891271968,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.024305183199285076,
      "loss": 0.7783,
      "num_input_tokens_seen": 6309536,
      "step": 10880
    },
    {
      "epoch": 1.621239201668156,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.024316353887399465,
      "loss": 0.782,
      "num_input_tokens_seen": 6312448,
      "step": 10885
    },
    {
      "epoch": 1.6219839142091153,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02432752457551385,
      "loss": 0.8037,
      "num_input_tokens_seen": 6315168,
      "step": 10890
    },
    {
      "epoch": 1.6227286267500745,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02433869526362824,
      "loss": 0.7965,
      "num_input_tokens_seen": 6317952,
      "step": 10895
    },
    {
      "epoch": 1.6234733392910337,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.024349865951742627,
      "loss": 0.795,
      "num_input_tokens_seen": 6320832,
      "step": 10900
    },
    {
      "epoch": 1.6242180518319929,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.024361036639857017,
      "loss": 0.8126,
      "num_input_tokens_seen": 6323872,
      "step": 10905
    },
    {
      "epoch": 1.624962764372952,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.024372207327971403,
      "loss": 0.7742,
      "num_input_tokens_seen": 6326560,
      "step": 10910
    },
    {
      "epoch": 1.6257074769139113,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02438337801608579,
      "loss": 0.8153,
      "num_input_tokens_seen": 6329504,
      "step": 10915
    },
    {
      "epoch": 1.6264521894548705,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.024394548704200176,
      "loss": 0.8113,
      "num_input_tokens_seen": 6332608,
      "step": 10920
    },
    {
      "epoch": 1.6271969019958297,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.024405719392314566,
      "loss": 0.8145,
      "num_input_tokens_seen": 6335328,
      "step": 10925
    },
    {
      "epoch": 1.6279416145367889,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.024416890080428952,
      "loss": 0.8144,
      "num_input_tokens_seen": 6338304,
      "step": 10930
    },
    {
      "epoch": 1.628686327077748,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02442806076854334,
      "loss": 0.8242,
      "num_input_tokens_seen": 6341632,
      "step": 10935
    },
    {
      "epoch": 1.6294310396187073,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.024439231456657728,
      "loss": 0.8008,
      "num_input_tokens_seen": 6344672,
      "step": 10940
    },
    {
      "epoch": 1.6301757521596665,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.024450402144772117,
      "loss": 0.8032,
      "num_input_tokens_seen": 6347520,
      "step": 10945
    },
    {
      "epoch": 1.6309204647006257,
      "grad_norm": 0.033203125,
      "learning_rate": 0.024461572832886504,
      "loss": 0.8135,
      "num_input_tokens_seen": 6350080,
      "step": 10950
    },
    {
      "epoch": 1.6316651772415849,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.024472743521000893,
      "loss": 0.7942,
      "num_input_tokens_seen": 6352768,
      "step": 10955
    },
    {
      "epoch": 1.632409889782544,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.02448391420911528,
      "loss": 0.7913,
      "num_input_tokens_seen": 6355776,
      "step": 10960
    },
    {
      "epoch": 1.6331546023235033,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.02449508489722967,
      "loss": 0.8149,
      "num_input_tokens_seen": 6358496,
      "step": 10965
    },
    {
      "epoch": 1.6338993148644623,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.024506255585344056,
      "loss": 0.7909,
      "num_input_tokens_seen": 6361408,
      "step": 10970
    },
    {
      "epoch": 1.6346440274054215,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.024517426273458445,
      "loss": 0.8083,
      "num_input_tokens_seen": 6364320,
      "step": 10975
    },
    {
      "epoch": 1.6353887399463807,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02452859696157283,
      "loss": 0.7982,
      "num_input_tokens_seen": 6366944,
      "step": 10980
    },
    {
      "epoch": 1.6361334524873399,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02453976764968722,
      "loss": 0.8241,
      "num_input_tokens_seen": 6369568,
      "step": 10985
    },
    {
      "epoch": 1.636878165028299,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.024550938337801607,
      "loss": 0.8178,
      "num_input_tokens_seen": 6372320,
      "step": 10990
    },
    {
      "epoch": 1.6376228775692583,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.024562109025915997,
      "loss": 0.7774,
      "num_input_tokens_seen": 6375264,
      "step": 10995
    },
    {
      "epoch": 1.6383675901102175,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.024573279714030383,
      "loss": 0.7954,
      "num_input_tokens_seen": 6378016,
      "step": 11000
    },
    {
      "epoch": 1.6391123026511767,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02458445040214477,
      "loss": 0.8012,
      "num_input_tokens_seen": 6380576,
      "step": 11005
    },
    {
      "epoch": 1.6398570151921357,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02459562109025916,
      "loss": 0.7844,
      "num_input_tokens_seen": 6383200,
      "step": 11010
    },
    {
      "epoch": 1.6406017277330949,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.024606791778373546,
      "loss": 0.8131,
      "num_input_tokens_seen": 6386016,
      "step": 11015
    },
    {
      "epoch": 1.641346440274054,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.024617962466487935,
      "loss": 0.8058,
      "num_input_tokens_seen": 6388768,
      "step": 11020
    },
    {
      "epoch": 1.6420911528150133,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02462913315460232,
      "loss": 0.8258,
      "num_input_tokens_seen": 6391616,
      "step": 11025
    },
    {
      "epoch": 1.6428358653559725,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02464030384271671,
      "loss": 0.8263,
      "num_input_tokens_seen": 6394240,
      "step": 11030
    },
    {
      "epoch": 1.6435805778969317,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.024651474530831097,
      "loss": 0.8148,
      "num_input_tokens_seen": 6397312,
      "step": 11035
    },
    {
      "epoch": 1.6443252904378909,
      "grad_norm": 0.025390625,
      "learning_rate": 0.024662645218945487,
      "loss": 0.801,
      "num_input_tokens_seen": 6399904,
      "step": 11040
    },
    {
      "epoch": 1.64507000297885,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.024673815907059873,
      "loss": 0.8215,
      "num_input_tokens_seen": 6402816,
      "step": 11045
    },
    {
      "epoch": 1.6458147155198093,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.024684986595174263,
      "loss": 0.8013,
      "num_input_tokens_seen": 6405792,
      "step": 11050
    },
    {
      "epoch": 1.6465594280607685,
      "grad_norm": 0.015625,
      "learning_rate": 0.02469615728328865,
      "loss": 0.824,
      "num_input_tokens_seen": 6408512,
      "step": 11055
    },
    {
      "epoch": 1.6473041406017277,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02470732797140304,
      "loss": 0.7942,
      "num_input_tokens_seen": 6411744,
      "step": 11060
    },
    {
      "epoch": 1.648048853142687,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.024718498659517425,
      "loss": 0.808,
      "num_input_tokens_seen": 6414720,
      "step": 11065
    },
    {
      "epoch": 1.648793565683646,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.024729669347631815,
      "loss": 0.8054,
      "num_input_tokens_seen": 6417536,
      "step": 11070
    },
    {
      "epoch": 1.6495382782246053,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.0247408400357462,
      "loss": 0.7996,
      "num_input_tokens_seen": 6420256,
      "step": 11075
    },
    {
      "epoch": 1.6502829907655645,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02475201072386059,
      "loss": 0.805,
      "num_input_tokens_seen": 6423296,
      "step": 11080
    },
    {
      "epoch": 1.6510277033065237,
      "grad_norm": 0.01953125,
      "learning_rate": 0.024763181411974974,
      "loss": 0.789,
      "num_input_tokens_seen": 6426080,
      "step": 11085
    },
    {
      "epoch": 1.651772415847483,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.024774352100089363,
      "loss": 0.8072,
      "num_input_tokens_seen": 6428864,
      "step": 11090
    },
    {
      "epoch": 1.6525171283884421,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02478552278820375,
      "loss": 0.8085,
      "num_input_tokens_seen": 6431584,
      "step": 11095
    },
    {
      "epoch": 1.6532618409294013,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.02479669347631814,
      "loss": 0.8017,
      "num_input_tokens_seen": 6434496,
      "step": 11100
    },
    {
      "epoch": 1.6540065534703605,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.024807864164432526,
      "loss": 0.7838,
      "num_input_tokens_seen": 6437280,
      "step": 11105
    },
    {
      "epoch": 1.6547512660113197,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.024819034852546915,
      "loss": 0.7912,
      "num_input_tokens_seen": 6440160,
      "step": 11110
    },
    {
      "epoch": 1.655495978552279,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.024830205540661305,
      "loss": 0.808,
      "num_input_tokens_seen": 6443328,
      "step": 11115
    },
    {
      "epoch": 1.6562406910932381,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02484137622877569,
      "loss": 0.8103,
      "num_input_tokens_seen": 6446112,
      "step": 11120
    },
    {
      "epoch": 1.6569854036341973,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02485254691689008,
      "loss": 0.8113,
      "num_input_tokens_seen": 6449152,
      "step": 11125
    },
    {
      "epoch": 1.6577301161751565,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.024863717605004467,
      "loss": 0.7873,
      "num_input_tokens_seen": 6451840,
      "step": 11130
    },
    {
      "epoch": 1.6584748287161157,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.024874888293118857,
      "loss": 0.8153,
      "num_input_tokens_seen": 6454784,
      "step": 11135
    },
    {
      "epoch": 1.6592195412570747,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.024886058981233243,
      "loss": 0.7772,
      "num_input_tokens_seen": 6457728,
      "step": 11140
    },
    {
      "epoch": 1.659964253798034,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.024897229669347633,
      "loss": 0.8129,
      "num_input_tokens_seen": 6460416,
      "step": 11145
    },
    {
      "epoch": 1.6607089663389931,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02490840035746202,
      "loss": 0.7906,
      "num_input_tokens_seen": 6463360,
      "step": 11150
    },
    {
      "epoch": 1.6614536788799523,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.02491957104557641,
      "loss": 0.7948,
      "num_input_tokens_seen": 6466144,
      "step": 11155
    },
    {
      "epoch": 1.6621983914209115,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.024930741733690795,
      "loss": 0.8011,
      "num_input_tokens_seen": 6469120,
      "step": 11160
    },
    {
      "epoch": 1.6629431039618707,
      "grad_norm": 0.03125,
      "learning_rate": 0.024941912421805185,
      "loss": 0.7995,
      "num_input_tokens_seen": 6472320,
      "step": 11165
    },
    {
      "epoch": 1.66368781650283,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02495308310991957,
      "loss": 0.7716,
      "num_input_tokens_seen": 6475296,
      "step": 11170
    },
    {
      "epoch": 1.6644325290437891,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.024964253798033957,
      "loss": 0.8371,
      "num_input_tokens_seen": 6478496,
      "step": 11175
    },
    {
      "epoch": 1.6651772415847483,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.024975424486148343,
      "loss": 0.8103,
      "num_input_tokens_seen": 6481440,
      "step": 11180
    },
    {
      "epoch": 1.6659219541257073,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.024986595174262733,
      "loss": 0.7833,
      "num_input_tokens_seen": 6484800,
      "step": 11185
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02499776586237712,
      "loss": 0.8148,
      "num_input_tokens_seen": 6487488,
      "step": 11190
    },
    {
      "epoch": 1.6674113792076257,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02500893655049151,
      "loss": 0.8082,
      "num_input_tokens_seen": 6490048,
      "step": 11195
    },
    {
      "epoch": 1.668156091748585,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.025020107238605895,
      "loss": 0.8219,
      "num_input_tokens_seen": 6493088,
      "step": 11200
    },
    {
      "epoch": 1.6689008042895441,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.025031277926720285,
      "loss": 0.7929,
      "num_input_tokens_seen": 6496128,
      "step": 11205
    },
    {
      "epoch": 1.6696455168305033,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02504244861483467,
      "loss": 0.82,
      "num_input_tokens_seen": 6498976,
      "step": 11210
    },
    {
      "epoch": 1.6703902293714625,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02505361930294906,
      "loss": 0.8304,
      "num_input_tokens_seen": 6502208,
      "step": 11215
    },
    {
      "epoch": 1.6711349419124217,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.025064789991063447,
      "loss": 0.7828,
      "num_input_tokens_seen": 6505408,
      "step": 11220
    },
    {
      "epoch": 1.671879654453381,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.025075960679177837,
      "loss": 0.7947,
      "num_input_tokens_seen": 6508416,
      "step": 11225
    },
    {
      "epoch": 1.6726243669943401,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.025087131367292227,
      "loss": 0.8031,
      "num_input_tokens_seen": 6511232,
      "step": 11230
    },
    {
      "epoch": 1.6733690795352993,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.025098302055406613,
      "loss": 0.7869,
      "num_input_tokens_seen": 6514112,
      "step": 11235
    },
    {
      "epoch": 1.6741137920762585,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.025109472743521002,
      "loss": 0.807,
      "num_input_tokens_seen": 6516928,
      "step": 11240
    },
    {
      "epoch": 1.6748585046172177,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.02512064343163539,
      "loss": 0.8038,
      "num_input_tokens_seen": 6519616,
      "step": 11245
    },
    {
      "epoch": 1.675603217158177,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02513181411974978,
      "loss": 0.7903,
      "num_input_tokens_seen": 6522624,
      "step": 11250
    },
    {
      "epoch": 1.6763479296991362,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.025142984807864165,
      "loss": 0.7936,
      "num_input_tokens_seen": 6525312,
      "step": 11255
    },
    {
      "epoch": 1.6770926422400954,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02515415549597855,
      "loss": 0.8044,
      "num_input_tokens_seen": 6528320,
      "step": 11260
    },
    {
      "epoch": 1.6778373547810546,
      "grad_norm": 0.03515625,
      "learning_rate": 0.025165326184092937,
      "loss": 0.8296,
      "num_input_tokens_seen": 6531168,
      "step": 11265
    },
    {
      "epoch": 1.6785820673220138,
      "grad_norm": 0.017578125,
      "learning_rate": 0.025176496872207327,
      "loss": 0.8041,
      "num_input_tokens_seen": 6534208,
      "step": 11270
    },
    {
      "epoch": 1.679326779862973,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.025187667560321713,
      "loss": 0.7864,
      "num_input_tokens_seen": 6536896,
      "step": 11275
    },
    {
      "epoch": 1.6800714924039322,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.025198838248436103,
      "loss": 0.7896,
      "num_input_tokens_seen": 6539744,
      "step": 11280
    },
    {
      "epoch": 1.6808162049448914,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02521000893655049,
      "loss": 0.8109,
      "num_input_tokens_seen": 6542336,
      "step": 11285
    },
    {
      "epoch": 1.6815609174858506,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02522117962466488,
      "loss": 0.8033,
      "num_input_tokens_seen": 6545632,
      "step": 11290
    },
    {
      "epoch": 1.6823056300268098,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.025232350312779265,
      "loss": 0.8156,
      "num_input_tokens_seen": 6548352,
      "step": 11295
    },
    {
      "epoch": 1.683050342567769,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.025243521000893655,
      "loss": 0.8032,
      "num_input_tokens_seen": 6551488,
      "step": 11300
    },
    {
      "epoch": 1.6837950551087282,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02525469168900804,
      "loss": 0.8146,
      "num_input_tokens_seen": 6554624,
      "step": 11305
    },
    {
      "epoch": 1.6845397676496874,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02526586237712243,
      "loss": 0.8084,
      "num_input_tokens_seen": 6557728,
      "step": 11310
    },
    {
      "epoch": 1.6852844801906464,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.025277033065236817,
      "loss": 0.8448,
      "num_input_tokens_seen": 6560800,
      "step": 11315
    },
    {
      "epoch": 1.6860291927316056,
      "grad_norm": 0.0390625,
      "learning_rate": 0.025288203753351206,
      "loss": 0.8077,
      "num_input_tokens_seen": 6563776,
      "step": 11320
    },
    {
      "epoch": 1.6867739052725648,
      "grad_norm": 0.033203125,
      "learning_rate": 0.025299374441465593,
      "loss": 0.8052,
      "num_input_tokens_seen": 6566624,
      "step": 11325
    },
    {
      "epoch": 1.687518617813524,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.025310545129579982,
      "loss": 0.7815,
      "num_input_tokens_seen": 6570048,
      "step": 11330
    },
    {
      "epoch": 1.6882633303544832,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02532171581769437,
      "loss": 0.8,
      "num_input_tokens_seen": 6572800,
      "step": 11335
    },
    {
      "epoch": 1.6890080428954424,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02533288650580876,
      "loss": 0.8043,
      "num_input_tokens_seen": 6575712,
      "step": 11340
    },
    {
      "epoch": 1.6897527554364016,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.025344057193923145,
      "loss": 0.8299,
      "num_input_tokens_seen": 6578688,
      "step": 11345
    },
    {
      "epoch": 1.6904974679773608,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02535522788203753,
      "loss": 0.8143,
      "num_input_tokens_seen": 6581568,
      "step": 11350
    },
    {
      "epoch": 1.69124218051832,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02536639857015192,
      "loss": 0.8034,
      "num_input_tokens_seen": 6584672,
      "step": 11355
    },
    {
      "epoch": 1.691986893059279,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.025377569258266307,
      "loss": 0.7935,
      "num_input_tokens_seen": 6587552,
      "step": 11360
    },
    {
      "epoch": 1.6927316056002382,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.025388739946380696,
      "loss": 0.8024,
      "num_input_tokens_seen": 6590624,
      "step": 11365
    },
    {
      "epoch": 1.6934763181411974,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.025399910634495083,
      "loss": 0.8132,
      "num_input_tokens_seen": 6593728,
      "step": 11370
    },
    {
      "epoch": 1.6942210306821566,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.025411081322609472,
      "loss": 0.7905,
      "num_input_tokens_seen": 6596480,
      "step": 11375
    },
    {
      "epoch": 1.6949657432231158,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02542225201072386,
      "loss": 0.7911,
      "num_input_tokens_seen": 6599360,
      "step": 11380
    },
    {
      "epoch": 1.695710455764075,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02543342269883825,
      "loss": 0.7887,
      "num_input_tokens_seen": 6602176,
      "step": 11385
    },
    {
      "epoch": 1.6964551683050342,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.025444593386952635,
      "loss": 0.7875,
      "num_input_tokens_seen": 6605024,
      "step": 11390
    },
    {
      "epoch": 1.6971998808459934,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.025455764075067024,
      "loss": 0.8047,
      "num_input_tokens_seen": 6607744,
      "step": 11395
    },
    {
      "epoch": 1.6979445933869526,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02546693476318141,
      "loss": 0.7841,
      "num_input_tokens_seen": 6610848,
      "step": 11400
    },
    {
      "epoch": 1.6986893059279118,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0254781054512958,
      "loss": 0.8048,
      "num_input_tokens_seen": 6613760,
      "step": 11405
    },
    {
      "epoch": 1.699434018468871,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.025489276139410186,
      "loss": 0.8255,
      "num_input_tokens_seen": 6617248,
      "step": 11410
    },
    {
      "epoch": 1.7001787310098302,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.025500446827524576,
      "loss": 0.8169,
      "num_input_tokens_seen": 6619904,
      "step": 11415
    },
    {
      "epoch": 1.7009234435507894,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.025511617515638962,
      "loss": 0.794,
      "num_input_tokens_seen": 6622528,
      "step": 11420
    },
    {
      "epoch": 1.7016681560917486,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.025522788203753352,
      "loss": 0.8136,
      "num_input_tokens_seen": 6625504,
      "step": 11425
    },
    {
      "epoch": 1.7024128686327078,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02553395889186774,
      "loss": 0.7905,
      "num_input_tokens_seen": 6628320,
      "step": 11430
    },
    {
      "epoch": 1.703157581173667,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.025545129579982125,
      "loss": 0.8037,
      "num_input_tokens_seen": 6631328,
      "step": 11435
    },
    {
      "epoch": 1.7039022937146262,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02555630026809651,
      "loss": 0.8093,
      "num_input_tokens_seen": 6634208,
      "step": 11440
    },
    {
      "epoch": 1.7046470062555854,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.0255674709562109,
      "loss": 0.8003,
      "num_input_tokens_seen": 6637216,
      "step": 11445
    },
    {
      "epoch": 1.7053917187965446,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02557864164432529,
      "loss": 0.8164,
      "num_input_tokens_seen": 6639936,
      "step": 11450
    },
    {
      "epoch": 1.7061364313375038,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.025589812332439676,
      "loss": 0.8368,
      "num_input_tokens_seen": 6643712,
      "step": 11455
    },
    {
      "epoch": 1.706881143878463,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.025600983020554066,
      "loss": 0.7764,
      "num_input_tokens_seen": 6646688,
      "step": 11460
    },
    {
      "epoch": 1.7076258564194222,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.025612153708668452,
      "loss": 0.7898,
      "num_input_tokens_seen": 6649728,
      "step": 11465
    },
    {
      "epoch": 1.7083705689603814,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.025623324396782842,
      "loss": 0.7987,
      "num_input_tokens_seen": 6652640,
      "step": 11470
    },
    {
      "epoch": 1.7091152815013406,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02563449508489723,
      "loss": 0.7962,
      "num_input_tokens_seen": 6655872,
      "step": 11475
    },
    {
      "epoch": 1.7098599940422998,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.025645665773011618,
      "loss": 0.796,
      "num_input_tokens_seen": 6658848,
      "step": 11480
    },
    {
      "epoch": 1.710604706583259,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.025656836461126004,
      "loss": 0.8027,
      "num_input_tokens_seen": 6661760,
      "step": 11485
    },
    {
      "epoch": 1.711349419124218,
      "grad_norm": 0.033203125,
      "learning_rate": 0.025668007149240394,
      "loss": 0.7857,
      "num_input_tokens_seen": 6664608,
      "step": 11490
    },
    {
      "epoch": 1.7120941316651772,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02567917783735478,
      "loss": 0.8198,
      "num_input_tokens_seen": 6667744,
      "step": 11495
    },
    {
      "epoch": 1.7128388442061364,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02569034852546917,
      "loss": 0.8151,
      "num_input_tokens_seen": 6670816,
      "step": 11500
    },
    {
      "epoch": 1.7135835567470956,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.025701519213583556,
      "loss": 0.7766,
      "num_input_tokens_seen": 6673664,
      "step": 11505
    },
    {
      "epoch": 1.7143282692880548,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.025712689901697946,
      "loss": 0.8152,
      "num_input_tokens_seen": 6676640,
      "step": 11510
    },
    {
      "epoch": 1.715072981829014,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.025723860589812332,
      "loss": 0.7937,
      "num_input_tokens_seen": 6679424,
      "step": 11515
    },
    {
      "epoch": 1.7158176943699732,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02573503127792672,
      "loss": 0.8008,
      "num_input_tokens_seen": 6682464,
      "step": 11520
    },
    {
      "epoch": 1.7165624069109324,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.025746201966041105,
      "loss": 0.8215,
      "num_input_tokens_seen": 6685280,
      "step": 11525
    },
    {
      "epoch": 1.7173071194518914,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.025757372654155494,
      "loss": 0.8138,
      "num_input_tokens_seen": 6688160,
      "step": 11530
    },
    {
      "epoch": 1.7180518319928506,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02576854334226988,
      "loss": 0.8171,
      "num_input_tokens_seen": 6691200,
      "step": 11535
    },
    {
      "epoch": 1.7187965445338098,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.02577971403038427,
      "loss": 0.8028,
      "num_input_tokens_seen": 6693888,
      "step": 11540
    },
    {
      "epoch": 1.719541257074769,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.025790884718498656,
      "loss": 0.8075,
      "num_input_tokens_seen": 6696768,
      "step": 11545
    },
    {
      "epoch": 1.7202859696157282,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.025802055406613046,
      "loss": 0.812,
      "num_input_tokens_seen": 6699808,
      "step": 11550
    },
    {
      "epoch": 1.7210306821566874,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.025813226094727432,
      "loss": 0.8182,
      "num_input_tokens_seen": 6702784,
      "step": 11555
    },
    {
      "epoch": 1.7217753946976466,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.025824396782841822,
      "loss": 0.8115,
      "num_input_tokens_seen": 6705760,
      "step": 11560
    },
    {
      "epoch": 1.7225201072386058,
      "grad_norm": 0.02734375,
      "learning_rate": 0.025835567470956212,
      "loss": 0.8103,
      "num_input_tokens_seen": 6708736,
      "step": 11565
    },
    {
      "epoch": 1.723264819779565,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.025846738159070598,
      "loss": 0.7983,
      "num_input_tokens_seen": 6711392,
      "step": 11570
    },
    {
      "epoch": 1.7240095323205242,
      "grad_norm": 0.044921875,
      "learning_rate": 0.025857908847184988,
      "loss": 0.8038,
      "num_input_tokens_seen": 6714496,
      "step": 11575
    },
    {
      "epoch": 1.7247542448614834,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.025869079535299374,
      "loss": 0.8061,
      "num_input_tokens_seen": 6717280,
      "step": 11580
    },
    {
      "epoch": 1.7254989574024426,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.025880250223413764,
      "loss": 0.7975,
      "num_input_tokens_seen": 6719904,
      "step": 11585
    },
    {
      "epoch": 1.7262436699434018,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02589142091152815,
      "loss": 0.8122,
      "num_input_tokens_seen": 6722560,
      "step": 11590
    },
    {
      "epoch": 1.726988382484361,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02590259159964254,
      "loss": 0.8012,
      "num_input_tokens_seen": 6725344,
      "step": 11595
    },
    {
      "epoch": 1.7277330950253202,
      "grad_norm": 0.03125,
      "learning_rate": 0.025913762287756926,
      "loss": 0.7941,
      "num_input_tokens_seen": 6728384,
      "step": 11600
    },
    {
      "epoch": 1.7284778075662794,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.025924932975871312,
      "loss": 0.8019,
      "num_input_tokens_seen": 6731360,
      "step": 11605
    },
    {
      "epoch": 1.7292225201072386,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0259361036639857,
      "loss": 0.8117,
      "num_input_tokens_seen": 6734016,
      "step": 11610
    },
    {
      "epoch": 1.7299672326481979,
      "grad_norm": 0.029296875,
      "learning_rate": 0.025947274352100088,
      "loss": 0.7859,
      "num_input_tokens_seen": 6736896,
      "step": 11615
    },
    {
      "epoch": 1.730711945189157,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.025958445040214474,
      "loss": 0.8271,
      "num_input_tokens_seen": 6739744,
      "step": 11620
    },
    {
      "epoch": 1.7314566577301163,
      "grad_norm": 0.04296875,
      "learning_rate": 0.025969615728328864,
      "loss": 0.8055,
      "num_input_tokens_seen": 6742528,
      "step": 11625
    },
    {
      "epoch": 1.7322013702710755,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02598078641644325,
      "loss": 0.8102,
      "num_input_tokens_seen": 6745472,
      "step": 11630
    },
    {
      "epoch": 1.7329460828120347,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02599195710455764,
      "loss": 0.7941,
      "num_input_tokens_seen": 6748736,
      "step": 11635
    },
    {
      "epoch": 1.7336907953529939,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.026003127792672026,
      "loss": 0.8012,
      "num_input_tokens_seen": 6751424,
      "step": 11640
    },
    {
      "epoch": 1.734435507893953,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.026014298480786416,
      "loss": 0.8055,
      "num_input_tokens_seen": 6754176,
      "step": 11645
    },
    {
      "epoch": 1.7351802204349123,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.026025469168900802,
      "loss": 0.8138,
      "num_input_tokens_seen": 6757024,
      "step": 11650
    },
    {
      "epoch": 1.7359249329758715,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.026036639857015192,
      "loss": 0.8029,
      "num_input_tokens_seen": 6759872,
      "step": 11655
    },
    {
      "epoch": 1.7366696455168305,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.026047810545129578,
      "loss": 0.8245,
      "num_input_tokens_seen": 6763104,
      "step": 11660
    },
    {
      "epoch": 1.7374143580577897,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.026058981233243968,
      "loss": 0.7938,
      "num_input_tokens_seen": 6765824,
      "step": 11665
    },
    {
      "epoch": 1.7381590705987489,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.026070151921358354,
      "loss": 0.8036,
      "num_input_tokens_seen": 6768672,
      "step": 11670
    },
    {
      "epoch": 1.738903783139708,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.026081322609472744,
      "loss": 0.7864,
      "num_input_tokens_seen": 6771616,
      "step": 11675
    },
    {
      "epoch": 1.7396484956806673,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.026092493297587133,
      "loss": 0.8252,
      "num_input_tokens_seen": 6774272,
      "step": 11680
    },
    {
      "epoch": 1.7403932082216265,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02610366398570152,
      "loss": 0.8147,
      "num_input_tokens_seen": 6777280,
      "step": 11685
    },
    {
      "epoch": 1.7411379207625857,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.026114834673815906,
      "loss": 0.8218,
      "num_input_tokens_seen": 6780096,
      "step": 11690
    },
    {
      "epoch": 1.7418826333035449,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.026126005361930292,
      "loss": 0.7954,
      "num_input_tokens_seen": 6783072,
      "step": 11695
    },
    {
      "epoch": 1.742627345844504,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.026137176050044682,
      "loss": 0.7967,
      "num_input_tokens_seen": 6785792,
      "step": 11700
    },
    {
      "epoch": 1.743372058385463,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.026148346738159068,
      "loss": 0.7927,
      "num_input_tokens_seen": 6788544,
      "step": 11705
    },
    {
      "epoch": 1.7441167709264223,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.026159517426273458,
      "loss": 0.8031,
      "num_input_tokens_seen": 6791680,
      "step": 11710
    },
    {
      "epoch": 1.7448614834673815,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.026170688114387844,
      "loss": 0.8142,
      "num_input_tokens_seen": 6794496,
      "step": 11715
    },
    {
      "epoch": 1.7456061960083407,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.026181858802502234,
      "loss": 0.8062,
      "num_input_tokens_seen": 6797408,
      "step": 11720
    },
    {
      "epoch": 1.7463509085492999,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.02619302949061662,
      "loss": 0.8231,
      "num_input_tokens_seen": 6800320,
      "step": 11725
    },
    {
      "epoch": 1.747095621090259,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02620420017873101,
      "loss": 0.8017,
      "num_input_tokens_seen": 6803360,
      "step": 11730
    },
    {
      "epoch": 1.7478403336312183,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.026215370866845396,
      "loss": 0.7974,
      "num_input_tokens_seen": 6806592,
      "step": 11735
    },
    {
      "epoch": 1.7485850461721775,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.026226541554959786,
      "loss": 0.787,
      "num_input_tokens_seen": 6809440,
      "step": 11740
    },
    {
      "epoch": 1.7493297587131367,
      "grad_norm": 0.03125,
      "learning_rate": 0.026237712243074172,
      "loss": 0.8171,
      "num_input_tokens_seen": 6812320,
      "step": 11745
    },
    {
      "epoch": 1.7500744712540959,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02624888293118856,
      "loss": 0.8086,
      "num_input_tokens_seen": 6814976,
      "step": 11750
    },
    {
      "epoch": 1.750819183795055,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.026260053619302948,
      "loss": 0.8181,
      "num_input_tokens_seen": 6817728,
      "step": 11755
    },
    {
      "epoch": 1.7515638963360143,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.026271224307417337,
      "loss": 0.8046,
      "num_input_tokens_seen": 6820672,
      "step": 11760
    },
    {
      "epoch": 1.7523086088769735,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.026282394995531724,
      "loss": 0.7965,
      "num_input_tokens_seen": 6823552,
      "step": 11765
    },
    {
      "epoch": 1.7530533214179327,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.026293565683646113,
      "loss": 0.8012,
      "num_input_tokens_seen": 6826976,
      "step": 11770
    },
    {
      "epoch": 1.7537980339588919,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.0263047363717605,
      "loss": 0.7925,
      "num_input_tokens_seen": 6830048,
      "step": 11775
    },
    {
      "epoch": 1.754542746499851,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.026315907059874886,
      "loss": 0.8143,
      "num_input_tokens_seen": 6832896,
      "step": 11780
    },
    {
      "epoch": 1.7552874590408103,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.026327077747989275,
      "loss": 0.7961,
      "num_input_tokens_seen": 6835712,
      "step": 11785
    },
    {
      "epoch": 1.7560321715817695,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02633824843610366,
      "loss": 0.8003,
      "num_input_tokens_seen": 6838368,
      "step": 11790
    },
    {
      "epoch": 1.7567768841227287,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02634941912421805,
      "loss": 0.8096,
      "num_input_tokens_seen": 6841216,
      "step": 11795
    },
    {
      "epoch": 1.757521596663688,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.026360589812332438,
      "loss": 0.8052,
      "num_input_tokens_seen": 6844512,
      "step": 11800
    },
    {
      "epoch": 1.758266309204647,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.026371760500446827,
      "loss": 0.7939,
      "num_input_tokens_seen": 6847584,
      "step": 11805
    },
    {
      "epoch": 1.7590110217456063,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.026382931188561214,
      "loss": 0.8115,
      "num_input_tokens_seen": 6850528,
      "step": 11810
    },
    {
      "epoch": 1.7597557342865655,
      "grad_norm": 0.025390625,
      "learning_rate": 0.026394101876675603,
      "loss": 0.7719,
      "num_input_tokens_seen": 6853504,
      "step": 11815
    },
    {
      "epoch": 1.7605004468275247,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02640527256478999,
      "loss": 0.8071,
      "num_input_tokens_seen": 6856064,
      "step": 11820
    },
    {
      "epoch": 1.761245159368484,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.02641644325290438,
      "loss": 0.7973,
      "num_input_tokens_seen": 6859040,
      "step": 11825
    },
    {
      "epoch": 1.7619898719094431,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.026427613941018765,
      "loss": 0.8343,
      "num_input_tokens_seen": 6861888,
      "step": 11830
    },
    {
      "epoch": 1.762734584450402,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.026438784629133155,
      "loss": 0.8018,
      "num_input_tokens_seen": 6864640,
      "step": 11835
    },
    {
      "epoch": 1.7634792969913613,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02644995531724754,
      "loss": 0.8152,
      "num_input_tokens_seen": 6867680,
      "step": 11840
    },
    {
      "epoch": 1.7642240095323205,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.02646112600536193,
      "loss": 0.8019,
      "num_input_tokens_seen": 6870592,
      "step": 11845
    },
    {
      "epoch": 1.7649687220732797,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.026472296693476317,
      "loss": 0.8024,
      "num_input_tokens_seen": 6873344,
      "step": 11850
    },
    {
      "epoch": 1.765713434614239,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.026483467381590707,
      "loss": 0.7941,
      "num_input_tokens_seen": 6876576,
      "step": 11855
    },
    {
      "epoch": 1.766458147155198,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.026494638069705093,
      "loss": 0.84,
      "num_input_tokens_seen": 6879232,
      "step": 11860
    },
    {
      "epoch": 1.7672028596961573,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02650580875781948,
      "loss": 0.7777,
      "num_input_tokens_seen": 6882048,
      "step": 11865
    },
    {
      "epoch": 1.7679475722371165,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.026516979445933866,
      "loss": 0.8098,
      "num_input_tokens_seen": 6884832,
      "step": 11870
    },
    {
      "epoch": 1.7686922847780755,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.026528150134048255,
      "loss": 0.7858,
      "num_input_tokens_seen": 6887904,
      "step": 11875
    },
    {
      "epoch": 1.7694369973190347,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02653932082216264,
      "loss": 0.7987,
      "num_input_tokens_seen": 6890720,
      "step": 11880
    },
    {
      "epoch": 1.770181709859994,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02655049151027703,
      "loss": 0.8109,
      "num_input_tokens_seen": 6893440,
      "step": 11885
    },
    {
      "epoch": 1.770926422400953,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.026561662198391418,
      "loss": 0.7947,
      "num_input_tokens_seen": 6896352,
      "step": 11890
    },
    {
      "epoch": 1.7716711349419123,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.026572832886505807,
      "loss": 0.8006,
      "num_input_tokens_seen": 6898816,
      "step": 11895
    },
    {
      "epoch": 1.7724158474828715,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.026584003574620197,
      "loss": 0.8038,
      "num_input_tokens_seen": 6901888,
      "step": 11900
    },
    {
      "epoch": 1.7731605600238307,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.026595174262734583,
      "loss": 0.804,
      "num_input_tokens_seen": 6904704,
      "step": 11905
    },
    {
      "epoch": 1.77390527256479,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.026606344950848973,
      "loss": 0.7797,
      "num_input_tokens_seen": 6907648,
      "step": 11910
    },
    {
      "epoch": 1.7746499851057491,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.02661751563896336,
      "loss": 0.8183,
      "num_input_tokens_seen": 6910272,
      "step": 11915
    },
    {
      "epoch": 1.7753946976467083,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02662868632707775,
      "loss": 0.8004,
      "num_input_tokens_seen": 6913344,
      "step": 11920
    },
    {
      "epoch": 1.7761394101876675,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.026639857015192135,
      "loss": 0.7971,
      "num_input_tokens_seen": 6916224,
      "step": 11925
    },
    {
      "epoch": 1.7768841227286267,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.026651027703306525,
      "loss": 0.8066,
      "num_input_tokens_seen": 6919456,
      "step": 11930
    },
    {
      "epoch": 1.777628835269586,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.02666219839142091,
      "loss": 0.829,
      "num_input_tokens_seen": 6922144,
      "step": 11935
    },
    {
      "epoch": 1.7783735478105451,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0266733690795353,
      "loss": 0.7981,
      "num_input_tokens_seen": 6925344,
      "step": 11940
    },
    {
      "epoch": 1.7791182603515043,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.026684539767649687,
      "loss": 0.7888,
      "num_input_tokens_seen": 6928224,
      "step": 11945
    },
    {
      "epoch": 1.7798629728924635,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.026695710455764073,
      "loss": 0.8157,
      "num_input_tokens_seen": 6931328,
      "step": 11950
    },
    {
      "epoch": 1.7806076854334227,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02670688114387846,
      "loss": 0.8177,
      "num_input_tokens_seen": 6934432,
      "step": 11955
    },
    {
      "epoch": 1.781352397974382,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02671805183199285,
      "loss": 0.8016,
      "num_input_tokens_seen": 6937216,
      "step": 11960
    },
    {
      "epoch": 1.7820971105153411,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.026729222520107235,
      "loss": 0.7729,
      "num_input_tokens_seen": 6940032,
      "step": 11965
    },
    {
      "epoch": 1.7828418230563003,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.026740393208221625,
      "loss": 0.8054,
      "num_input_tokens_seen": 6942816,
      "step": 11970
    },
    {
      "epoch": 1.7835865355972595,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02675156389633601,
      "loss": 0.7881,
      "num_input_tokens_seen": 6945760,
      "step": 11975
    },
    {
      "epoch": 1.7843312481382188,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0267627345844504,
      "loss": 0.8104,
      "num_input_tokens_seen": 6948896,
      "step": 11980
    },
    {
      "epoch": 1.785075960679178,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.026773905272564787,
      "loss": 0.7859,
      "num_input_tokens_seen": 6951904,
      "step": 11985
    },
    {
      "epoch": 1.7858206732201372,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.026785075960679177,
      "loss": 0.7984,
      "num_input_tokens_seen": 6955104,
      "step": 11990
    },
    {
      "epoch": 1.7865653857610964,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.026796246648793563,
      "loss": 0.8074,
      "num_input_tokens_seen": 6958048,
      "step": 11995
    },
    {
      "epoch": 1.7873100983020556,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.026807417336907953,
      "loss": 0.7724,
      "num_input_tokens_seen": 6960864,
      "step": 12000
    },
    {
      "epoch": 1.7880548108430145,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02681858802502234,
      "loss": 0.7916,
      "num_input_tokens_seen": 6963648,
      "step": 12005
    },
    {
      "epoch": 1.7887995233839737,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.02682975871313673,
      "loss": 0.7977,
      "num_input_tokens_seen": 6966528,
      "step": 12010
    },
    {
      "epoch": 1.789544235924933,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.02684092940125112,
      "loss": 0.7746,
      "num_input_tokens_seen": 6969280,
      "step": 12015
    },
    {
      "epoch": 1.7902889484658921,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.026852100089365505,
      "loss": 0.8211,
      "num_input_tokens_seen": 6972320,
      "step": 12020
    },
    {
      "epoch": 1.7910336610068514,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.026863270777479895,
      "loss": 0.7893,
      "num_input_tokens_seen": 6975552,
      "step": 12025
    },
    {
      "epoch": 1.7917783735478106,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.02687444146559428,
      "loss": 0.8298,
      "num_input_tokens_seen": 6978560,
      "step": 12030
    },
    {
      "epoch": 1.7925230860887698,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02688561215370867,
      "loss": 0.7829,
      "num_input_tokens_seen": 6981408,
      "step": 12035
    },
    {
      "epoch": 1.793267798629729,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.026896782841823053,
      "loss": 0.8483,
      "num_input_tokens_seen": 6984384,
      "step": 12040
    },
    {
      "epoch": 1.7940125111706882,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.026907953529937443,
      "loss": 0.8288,
      "num_input_tokens_seen": 6987104,
      "step": 12045
    },
    {
      "epoch": 1.7947572237116471,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02691912421805183,
      "loss": 0.816,
      "num_input_tokens_seen": 6990112,
      "step": 12050
    },
    {
      "epoch": 1.7955019362526063,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02693029490616622,
      "loss": 0.7952,
      "num_input_tokens_seen": 6992928,
      "step": 12055
    },
    {
      "epoch": 1.7962466487935655,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.026941465594280605,
      "loss": 0.8235,
      "num_input_tokens_seen": 6995872,
      "step": 12060
    },
    {
      "epoch": 1.7969913613345248,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.026952636282394995,
      "loss": 0.8129,
      "num_input_tokens_seen": 6998784,
      "step": 12065
    },
    {
      "epoch": 1.797736073875484,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02696380697050938,
      "loss": 0.8034,
      "num_input_tokens_seen": 7001728,
      "step": 12070
    },
    {
      "epoch": 1.7984807864164432,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02697497765862377,
      "loss": 0.7925,
      "num_input_tokens_seen": 7004672,
      "step": 12075
    },
    {
      "epoch": 1.7992254989574024,
      "grad_norm": 0.025390625,
      "learning_rate": 0.026986148346738157,
      "loss": 0.8014,
      "num_input_tokens_seen": 7007712,
      "step": 12080
    },
    {
      "epoch": 1.7999702114983616,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.026997319034852547,
      "loss": 0.8019,
      "num_input_tokens_seen": 7010656,
      "step": 12085
    },
    {
      "epoch": 1.8007149240393208,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.027008489722966933,
      "loss": 0.7922,
      "num_input_tokens_seen": 7013472,
      "step": 12090
    },
    {
      "epoch": 1.80145963658028,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.027019660411081323,
      "loss": 0.7822,
      "num_input_tokens_seen": 7016352,
      "step": 12095
    },
    {
      "epoch": 1.8022043491212392,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02703083109919571,
      "loss": 0.8175,
      "num_input_tokens_seen": 7019168,
      "step": 12100
    },
    {
      "epoch": 1.8029490616621984,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.0270420017873101,
      "loss": 0.7958,
      "num_input_tokens_seen": 7022400,
      "step": 12105
    },
    {
      "epoch": 1.8036937742031576,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.027053172475424485,
      "loss": 0.82,
      "num_input_tokens_seen": 7025440,
      "step": 12110
    },
    {
      "epoch": 1.8044384867441168,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.027064343163538875,
      "loss": 0.8099,
      "num_input_tokens_seen": 7028320,
      "step": 12115
    },
    {
      "epoch": 1.805183199285076,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02707551385165326,
      "loss": 0.8102,
      "num_input_tokens_seen": 7031168,
      "step": 12120
    },
    {
      "epoch": 1.8059279118260352,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.027086684539767647,
      "loss": 0.8068,
      "num_input_tokens_seen": 7033888,
      "step": 12125
    },
    {
      "epoch": 1.8066726243669944,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.027097855227882037,
      "loss": 0.7879,
      "num_input_tokens_seen": 7036544,
      "step": 12130
    },
    {
      "epoch": 1.8074173369079536,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.027109025915996423,
      "loss": 0.8094,
      "num_input_tokens_seen": 7039264,
      "step": 12135
    },
    {
      "epoch": 1.8081620494489128,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.027120196604110813,
      "loss": 0.7859,
      "num_input_tokens_seen": 7041920,
      "step": 12140
    },
    {
      "epoch": 1.808906761989872,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.0271313672922252,
      "loss": 0.8206,
      "num_input_tokens_seen": 7044800,
      "step": 12145
    },
    {
      "epoch": 1.8096514745308312,
      "grad_norm": 0.013671875,
      "learning_rate": 0.02714253798033959,
      "loss": 0.8075,
      "num_input_tokens_seen": 7047968,
      "step": 12150
    },
    {
      "epoch": 1.8103961870717904,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.027153708668453975,
      "loss": 0.8136,
      "num_input_tokens_seen": 7050784,
      "step": 12155
    },
    {
      "epoch": 1.8111408996127496,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.027164879356568365,
      "loss": 0.7887,
      "num_input_tokens_seen": 7053632,
      "step": 12160
    },
    {
      "epoch": 1.8118856121537088,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02717605004468275,
      "loss": 0.8071,
      "num_input_tokens_seen": 7056736,
      "step": 12165
    },
    {
      "epoch": 1.812630324694668,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02718722073279714,
      "loss": 0.8002,
      "num_input_tokens_seen": 7059488,
      "step": 12170
    },
    {
      "epoch": 1.8133750372356272,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.027198391420911527,
      "loss": 0.7941,
      "num_input_tokens_seen": 7063616,
      "step": 12175
    },
    {
      "epoch": 1.8141197497765862,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.027209562109025916,
      "loss": 0.8149,
      "num_input_tokens_seen": 7066720,
      "step": 12180
    },
    {
      "epoch": 1.8148644623175454,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.027220732797140303,
      "loss": 0.8184,
      "num_input_tokens_seen": 7069568,
      "step": 12185
    },
    {
      "epoch": 1.8156091748585046,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.027231903485254692,
      "loss": 0.8065,
      "num_input_tokens_seen": 7072320,
      "step": 12190
    },
    {
      "epoch": 1.8163538873994638,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.02724307417336908,
      "loss": 0.8163,
      "num_input_tokens_seen": 7075264,
      "step": 12195
    },
    {
      "epoch": 1.817098599940423,
      "grad_norm": 0.03125,
      "learning_rate": 0.02725424486148347,
      "loss": 0.8264,
      "num_input_tokens_seen": 7078272,
      "step": 12200
    },
    {
      "epoch": 1.8178433124813822,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.027265415549597855,
      "loss": 0.7874,
      "num_input_tokens_seen": 7081088,
      "step": 12205
    },
    {
      "epoch": 1.8185880250223414,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02727658623771224,
      "loss": 0.7959,
      "num_input_tokens_seen": 7084032,
      "step": 12210
    },
    {
      "epoch": 1.8193327375633006,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.027287756925826627,
      "loss": 0.8006,
      "num_input_tokens_seen": 7087104,
      "step": 12215
    },
    {
      "epoch": 1.8200774501042598,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.027298927613941017,
      "loss": 0.7907,
      "num_input_tokens_seen": 7089920,
      "step": 12220
    },
    {
      "epoch": 1.8208221626452188,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.027310098302055403,
      "loss": 0.7835,
      "num_input_tokens_seen": 7093056,
      "step": 12225
    },
    {
      "epoch": 1.821566875186178,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.027321268990169793,
      "loss": 0.8025,
      "num_input_tokens_seen": 7095584,
      "step": 12230
    },
    {
      "epoch": 1.8223115877271372,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.027332439678284182,
      "loss": 0.8007,
      "num_input_tokens_seen": 7098464,
      "step": 12235
    },
    {
      "epoch": 1.8230563002680964,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02734361036639857,
      "loss": 0.7971,
      "num_input_tokens_seen": 7101440,
      "step": 12240
    },
    {
      "epoch": 1.8238010128090556,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02735478105451296,
      "loss": 0.8042,
      "num_input_tokens_seen": 7104224,
      "step": 12245
    },
    {
      "epoch": 1.8245457253500148,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.027365951742627345,
      "loss": 0.7772,
      "num_input_tokens_seen": 7107488,
      "step": 12250
    },
    {
      "epoch": 1.825290437890974,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.027377122430741734,
      "loss": 0.8136,
      "num_input_tokens_seen": 7110176,
      "step": 12255
    },
    {
      "epoch": 1.8260351504319332,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02738829311885612,
      "loss": 0.8029,
      "num_input_tokens_seen": 7113056,
      "step": 12260
    },
    {
      "epoch": 1.8267798629728924,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02739946380697051,
      "loss": 0.8374,
      "num_input_tokens_seen": 7115840,
      "step": 12265
    },
    {
      "epoch": 1.8275245755138516,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.027410634495084896,
      "loss": 0.7896,
      "num_input_tokens_seen": 7118880,
      "step": 12270
    },
    {
      "epoch": 1.8282692880548108,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.027421805183199286,
      "loss": 0.7931,
      "num_input_tokens_seen": 7121664,
      "step": 12275
    },
    {
      "epoch": 1.82901400059577,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.027432975871313672,
      "loss": 0.8006,
      "num_input_tokens_seen": 7124480,
      "step": 12280
    },
    {
      "epoch": 1.8297587131367292,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.027444146559428062,
      "loss": 0.7857,
      "num_input_tokens_seen": 7127424,
      "step": 12285
    },
    {
      "epoch": 1.8305034256776884,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.027455317247542448,
      "loss": 0.8023,
      "num_input_tokens_seen": 7130208,
      "step": 12290
    },
    {
      "epoch": 1.8312481382186476,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.027466487935656838,
      "loss": 0.782,
      "num_input_tokens_seen": 7133344,
      "step": 12295
    },
    {
      "epoch": 1.8319928507596068,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02747765862377122,
      "loss": 0.8197,
      "num_input_tokens_seen": 7136416,
      "step": 12300
    },
    {
      "epoch": 1.832737563300566,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02748882931188561,
      "loss": 0.8038,
      "num_input_tokens_seen": 7139168,
      "step": 12305
    },
    {
      "epoch": 1.8334822758415252,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.027499999999999997,
      "loss": 0.8115,
      "num_input_tokens_seen": 7142080,
      "step": 12310
    },
    {
      "epoch": 1.8342269883824844,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.027511170688114386,
      "loss": 0.8199,
      "num_input_tokens_seen": 7144832,
      "step": 12315
    },
    {
      "epoch": 1.8349717009234436,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.027522341376228773,
      "loss": 0.8291,
      "num_input_tokens_seen": 7147840,
      "step": 12320
    },
    {
      "epoch": 1.8357164134644028,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.027533512064343162,
      "loss": 0.8064,
      "num_input_tokens_seen": 7150912,
      "step": 12325
    },
    {
      "epoch": 1.836461126005362,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.02754468275245755,
      "loss": 0.8011,
      "num_input_tokens_seen": 7153952,
      "step": 12330
    },
    {
      "epoch": 1.8372058385463212,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.027555853440571938,
      "loss": 0.8165,
      "num_input_tokens_seen": 7156864,
      "step": 12335
    },
    {
      "epoch": 1.8379505510872804,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.027567024128686324,
      "loss": 0.81,
      "num_input_tokens_seen": 7159872,
      "step": 12340
    },
    {
      "epoch": 1.8386952636282397,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.027578194816800714,
      "loss": 0.8096,
      "num_input_tokens_seen": 7162944,
      "step": 12345
    },
    {
      "epoch": 1.8394399761691989,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.027589365504915104,
      "loss": 0.7984,
      "num_input_tokens_seen": 7165952,
      "step": 12350
    },
    {
      "epoch": 1.8401846887101578,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02760053619302949,
      "loss": 0.8148,
      "num_input_tokens_seen": 7168768,
      "step": 12355
    },
    {
      "epoch": 1.840929401251117,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02761170688114388,
      "loss": 0.8088,
      "num_input_tokens_seen": 7171744,
      "step": 12360
    },
    {
      "epoch": 1.8416741137920762,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.027622877569258266,
      "loss": 0.8026,
      "num_input_tokens_seen": 7174816,
      "step": 12365
    },
    {
      "epoch": 1.8424188263330354,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.027634048257372656,
      "loss": 0.803,
      "num_input_tokens_seen": 7177792,
      "step": 12370
    },
    {
      "epoch": 1.8431635388739946,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.027645218945487042,
      "loss": 0.808,
      "num_input_tokens_seen": 7180704,
      "step": 12375
    },
    {
      "epoch": 1.8439082514149538,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02765638963360143,
      "loss": 0.792,
      "num_input_tokens_seen": 7183648,
      "step": 12380
    },
    {
      "epoch": 1.844652963955913,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.027667560321715814,
      "loss": 0.8017,
      "num_input_tokens_seen": 7186592,
      "step": 12385
    },
    {
      "epoch": 1.8453976764968723,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.027678731009830204,
      "loss": 0.794,
      "num_input_tokens_seen": 7189280,
      "step": 12390
    },
    {
      "epoch": 1.8461423890378312,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02768990169794459,
      "loss": 0.7928,
      "num_input_tokens_seen": 7192160,
      "step": 12395
    },
    {
      "epoch": 1.8468871015787904,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.02770107238605898,
      "loss": 0.814,
      "num_input_tokens_seen": 7195232,
      "step": 12400
    },
    {
      "epoch": 1.8476318141197496,
      "grad_norm": 0.033203125,
      "learning_rate": 0.027712243074173366,
      "loss": 0.7996,
      "num_input_tokens_seen": 7198112,
      "step": 12405
    },
    {
      "epoch": 1.8483765266607088,
      "grad_norm": 0.03515625,
      "learning_rate": 0.027723413762287756,
      "loss": 0.8092,
      "num_input_tokens_seen": 7201056,
      "step": 12410
    },
    {
      "epoch": 1.849121239201668,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.027734584450402142,
      "loss": 0.764,
      "num_input_tokens_seen": 7203968,
      "step": 12415
    },
    {
      "epoch": 1.8498659517426272,
      "grad_norm": 0.033203125,
      "learning_rate": 0.027745755138516532,
      "loss": 0.8104,
      "num_input_tokens_seen": 7206976,
      "step": 12420
    },
    {
      "epoch": 1.8506106642835864,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.027756925826630918,
      "loss": 0.8061,
      "num_input_tokens_seen": 7209920,
      "step": 12425
    },
    {
      "epoch": 1.8513553768245457,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.027768096514745308,
      "loss": 0.8215,
      "num_input_tokens_seen": 7212640,
      "step": 12430
    },
    {
      "epoch": 1.8521000893655049,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.027779267202859694,
      "loss": 0.7963,
      "num_input_tokens_seen": 7215328,
      "step": 12435
    },
    {
      "epoch": 1.852844801906464,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.027790437890974084,
      "loss": 0.8068,
      "num_input_tokens_seen": 7218080,
      "step": 12440
    },
    {
      "epoch": 1.8535895144474233,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02780160857908847,
      "loss": 0.8234,
      "num_input_tokens_seen": 7220832,
      "step": 12445
    },
    {
      "epoch": 1.8543342269883825,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02781277926720286,
      "loss": 0.8175,
      "num_input_tokens_seen": 7223680,
      "step": 12450
    },
    {
      "epoch": 1.8550789395293417,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.027823949955317246,
      "loss": 0.8169,
      "num_input_tokens_seen": 7226368,
      "step": 12455
    },
    {
      "epoch": 1.8558236520703009,
      "grad_norm": 0.041015625,
      "learning_rate": 0.027835120643431636,
      "loss": 0.8053,
      "num_input_tokens_seen": 7229120,
      "step": 12460
    },
    {
      "epoch": 1.85656836461126,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.027846291331546025,
      "loss": 0.8068,
      "num_input_tokens_seen": 7232000,
      "step": 12465
    },
    {
      "epoch": 1.8573130771522193,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.027857462019660408,
      "loss": 0.7882,
      "num_input_tokens_seen": 7234784,
      "step": 12470
    },
    {
      "epoch": 1.8580577896931785,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.027868632707774798,
      "loss": 0.8048,
      "num_input_tokens_seen": 7237632,
      "step": 12475
    },
    {
      "epoch": 1.8588025022341377,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.027879803395889184,
      "loss": 0.796,
      "num_input_tokens_seen": 7240608,
      "step": 12480
    },
    {
      "epoch": 1.8595472147750969,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.027890974084003574,
      "loss": 0.78,
      "num_input_tokens_seen": 7243392,
      "step": 12485
    },
    {
      "epoch": 1.860291927316056,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02790214477211796,
      "loss": 0.8077,
      "num_input_tokens_seen": 7246432,
      "step": 12490
    },
    {
      "epoch": 1.8610366398570153,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02791331546023235,
      "loss": 0.7922,
      "num_input_tokens_seen": 7249120,
      "step": 12495
    },
    {
      "epoch": 1.8617813523979745,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.027924486148346736,
      "loss": 0.8111,
      "num_input_tokens_seen": 7251872,
      "step": 12500
    },
    {
      "epoch": 1.8625260649389337,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.027935656836461126,
      "loss": 0.7889,
      "num_input_tokens_seen": 7254752,
      "step": 12505
    },
    {
      "epoch": 1.863270777479893,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.027946827524575512,
      "loss": 0.8318,
      "num_input_tokens_seen": 7257536,
      "step": 12510
    },
    {
      "epoch": 1.864015490020852,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0279579982126899,
      "loss": 0.8336,
      "num_input_tokens_seen": 7260576,
      "step": 12515
    },
    {
      "epoch": 1.8647602025618113,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.027969168900804288,
      "loss": 0.8098,
      "num_input_tokens_seen": 7263424,
      "step": 12520
    },
    {
      "epoch": 1.8655049151027703,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.027980339588918678,
      "loss": 0.8102,
      "num_input_tokens_seen": 7266208,
      "step": 12525
    },
    {
      "epoch": 1.8662496276437295,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.027991510277033064,
      "loss": 0.7881,
      "num_input_tokens_seen": 7269248,
      "step": 12530
    },
    {
      "epoch": 1.8669943401846887,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.028002680965147454,
      "loss": 0.7939,
      "num_input_tokens_seen": 7272320,
      "step": 12535
    },
    {
      "epoch": 1.8677390527256479,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02801385165326184,
      "loss": 0.8045,
      "num_input_tokens_seen": 7275552,
      "step": 12540
    },
    {
      "epoch": 1.868483765266607,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02802502234137623,
      "loss": 0.8254,
      "num_input_tokens_seen": 7278432,
      "step": 12545
    },
    {
      "epoch": 1.8692284778075663,
      "grad_norm": 0.02734375,
      "learning_rate": 0.028036193029490616,
      "loss": 0.8117,
      "num_input_tokens_seen": 7281024,
      "step": 12550
    },
    {
      "epoch": 1.8699731903485255,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.028047363717605005,
      "loss": 0.8063,
      "num_input_tokens_seen": 7284032,
      "step": 12555
    },
    {
      "epoch": 1.8707179028894847,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.028058534405719388,
      "loss": 0.8176,
      "num_input_tokens_seen": 7287424,
      "step": 12560
    },
    {
      "epoch": 1.871462615430444,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.028069705093833778,
      "loss": 0.8114,
      "num_input_tokens_seen": 7290272,
      "step": 12565
    },
    {
      "epoch": 1.8722073279714029,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.028080875781948168,
      "loss": 0.8237,
      "num_input_tokens_seen": 7292864,
      "step": 12570
    },
    {
      "epoch": 1.872952040512362,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.028092046470062554,
      "loss": 0.8039,
      "num_input_tokens_seen": 7295680,
      "step": 12575
    },
    {
      "epoch": 1.8736967530533213,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.028103217158176944,
      "loss": 0.8056,
      "num_input_tokens_seen": 7298656,
      "step": 12580
    },
    {
      "epoch": 1.8744414655942805,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.02811438784629133,
      "loss": 0.8064,
      "num_input_tokens_seen": 7301408,
      "step": 12585
    },
    {
      "epoch": 1.8751861781352397,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02812555853440572,
      "loss": 0.8046,
      "num_input_tokens_seen": 7304128,
      "step": 12590
    },
    {
      "epoch": 1.875930890676199,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.028136729222520106,
      "loss": 0.7961,
      "num_input_tokens_seen": 7306976,
      "step": 12595
    },
    {
      "epoch": 1.876675603217158,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.028147899910634495,
      "loss": 0.8078,
      "num_input_tokens_seen": 7309504,
      "step": 12600
    },
    {
      "epoch": 1.8774203157581173,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.02815907059874888,
      "loss": 0.787,
      "num_input_tokens_seen": 7312256,
      "step": 12605
    },
    {
      "epoch": 1.8781650282990765,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02817024128686327,
      "loss": 0.7962,
      "num_input_tokens_seen": 7315040,
      "step": 12610
    },
    {
      "epoch": 1.8789097408400357,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.028181411974977658,
      "loss": 0.7964,
      "num_input_tokens_seen": 7318016,
      "step": 12615
    },
    {
      "epoch": 1.879654453380995,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.028192582663092047,
      "loss": 0.7955,
      "num_input_tokens_seen": 7320608,
      "step": 12620
    },
    {
      "epoch": 1.880399165921954,
      "grad_norm": 0.0234375,
      "learning_rate": 0.028203753351206434,
      "loss": 0.8276,
      "num_input_tokens_seen": 7323520,
      "step": 12625
    },
    {
      "epoch": 1.8811438784629133,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.028214924039320823,
      "loss": 0.7942,
      "num_input_tokens_seen": 7326816,
      "step": 12630
    },
    {
      "epoch": 1.8818885910038725,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02822609472743521,
      "loss": 0.8039,
      "num_input_tokens_seen": 7329824,
      "step": 12635
    },
    {
      "epoch": 1.8826333035448317,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.0282372654155496,
      "loss": 0.7946,
      "num_input_tokens_seen": 7332736,
      "step": 12640
    },
    {
      "epoch": 1.883378016085791,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.028248436103663982,
      "loss": 0.7996,
      "num_input_tokens_seen": 7335584,
      "step": 12645
    },
    {
      "epoch": 1.8841227286267501,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.02825960679177837,
      "loss": 0.8362,
      "num_input_tokens_seen": 7338432,
      "step": 12650
    },
    {
      "epoch": 1.8848674411677093,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.028270777479892758,
      "loss": 0.7965,
      "num_input_tokens_seen": 7341344,
      "step": 12655
    },
    {
      "epoch": 1.8856121537086685,
      "grad_norm": 0.029296875,
      "learning_rate": 0.028281948168007148,
      "loss": 0.7904,
      "num_input_tokens_seen": 7344512,
      "step": 12660
    },
    {
      "epoch": 1.8863568662496277,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.028293118856121534,
      "loss": 0.799,
      "num_input_tokens_seen": 7347264,
      "step": 12665
    },
    {
      "epoch": 1.887101578790587,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.028304289544235924,
      "loss": 0.8062,
      "num_input_tokens_seen": 7350336,
      "step": 12670
    },
    {
      "epoch": 1.8878462913315461,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02831546023235031,
      "loss": 0.8049,
      "num_input_tokens_seen": 7353216,
      "step": 12675
    },
    {
      "epoch": 1.8885910038725053,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.0283266309204647,
      "loss": 0.8082,
      "num_input_tokens_seen": 7355968,
      "step": 12680
    },
    {
      "epoch": 1.8893357164134645,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.028337801608579086,
      "loss": 0.7899,
      "num_input_tokens_seen": 7358912,
      "step": 12685
    },
    {
      "epoch": 1.8900804289544237,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.028348972296693475,
      "loss": 0.7958,
      "num_input_tokens_seen": 7361664,
      "step": 12690
    },
    {
      "epoch": 1.890825141495383,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.028360142984807865,
      "loss": 0.7841,
      "num_input_tokens_seen": 7364704,
      "step": 12695
    },
    {
      "epoch": 1.891569854036342,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02837131367292225,
      "loss": 0.8098,
      "num_input_tokens_seen": 7367424,
      "step": 12700
    },
    {
      "epoch": 1.8923145665773011,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02838248436103664,
      "loss": 0.8033,
      "num_input_tokens_seen": 7370432,
      "step": 12705
    },
    {
      "epoch": 1.8930592791182603,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.028393655049151027,
      "loss": 0.8082,
      "num_input_tokens_seen": 7373152,
      "step": 12710
    },
    {
      "epoch": 1.8938039916592195,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.028404825737265417,
      "loss": 0.8114,
      "num_input_tokens_seen": 7375872,
      "step": 12715
    },
    {
      "epoch": 1.8945487042001787,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.028415996425379803,
      "loss": 0.7903,
      "num_input_tokens_seen": 7378624,
      "step": 12720
    },
    {
      "epoch": 1.895293416741138,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.028427167113494193,
      "loss": 0.8099,
      "num_input_tokens_seen": 7381632,
      "step": 12725
    },
    {
      "epoch": 1.8960381292820971,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.028438337801608576,
      "loss": 0.7975,
      "num_input_tokens_seen": 7384832,
      "step": 12730
    },
    {
      "epoch": 1.8967828418230563,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.028449508489722965,
      "loss": 0.7992,
      "num_input_tokens_seen": 7387808,
      "step": 12735
    },
    {
      "epoch": 1.8975275543640155,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02846067917783735,
      "loss": 0.7869,
      "num_input_tokens_seen": 7390752,
      "step": 12740
    },
    {
      "epoch": 1.8982722669049745,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.02847184986595174,
      "loss": 0.781,
      "num_input_tokens_seen": 7393408,
      "step": 12745
    },
    {
      "epoch": 1.8990169794459337,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.028483020554066128,
      "loss": 0.7914,
      "num_input_tokens_seen": 7396448,
      "step": 12750
    },
    {
      "epoch": 1.899761691986893,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.028494191242180517,
      "loss": 0.79,
      "num_input_tokens_seen": 7399584,
      "step": 12755
    },
    {
      "epoch": 1.9005064045278521,
      "grad_norm": 0.037109375,
      "learning_rate": 0.028505361930294904,
      "loss": 0.835,
      "num_input_tokens_seen": 7402400,
      "step": 12760
    },
    {
      "epoch": 1.9012511170688113,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.028516532618409293,
      "loss": 0.8232,
      "num_input_tokens_seen": 7405280,
      "step": 12765
    },
    {
      "epoch": 1.9019958296097705,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.02852770330652368,
      "loss": 0.8009,
      "num_input_tokens_seen": 7408160,
      "step": 12770
    },
    {
      "epoch": 1.9027405421507297,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02853887399463807,
      "loss": 0.8381,
      "num_input_tokens_seen": 7411136,
      "step": 12775
    },
    {
      "epoch": 1.903485254691689,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.028550044682752455,
      "loss": 0.7699,
      "num_input_tokens_seen": 7413984,
      "step": 12780
    },
    {
      "epoch": 1.9042299672326481,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.028561215370866845,
      "loss": 0.8001,
      "num_input_tokens_seen": 7416960,
      "step": 12785
    },
    {
      "epoch": 1.9049746797736073,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02857238605898123,
      "loss": 0.7944,
      "num_input_tokens_seen": 7419776,
      "step": 12790
    },
    {
      "epoch": 1.9057193923145666,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02858355674709562,
      "loss": 0.7767,
      "num_input_tokens_seen": 7422752,
      "step": 12795
    },
    {
      "epoch": 1.9064641048555258,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.028594727435210007,
      "loss": 0.792,
      "num_input_tokens_seen": 7425664,
      "step": 12800
    },
    {
      "epoch": 1.907208817396485,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.028605898123324397,
      "loss": 0.806,
      "num_input_tokens_seen": 7428448,
      "step": 12805
    },
    {
      "epoch": 1.9079535299374442,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.028617068811438787,
      "loss": 0.8235,
      "num_input_tokens_seen": 7431200,
      "step": 12810
    },
    {
      "epoch": 1.9086982424784034,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02862823949955317,
      "loss": 0.7969,
      "num_input_tokens_seen": 7434144,
      "step": 12815
    },
    {
      "epoch": 1.9094429550193626,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02863941018766756,
      "loss": 0.8067,
      "num_input_tokens_seen": 7437056,
      "step": 12820
    },
    {
      "epoch": 1.9101876675603218,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.028650580875781945,
      "loss": 0.7801,
      "num_input_tokens_seen": 7439648,
      "step": 12825
    },
    {
      "epoch": 1.910932380101281,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.028661751563896335,
      "loss": 0.7886,
      "num_input_tokens_seen": 7442272,
      "step": 12830
    },
    {
      "epoch": 1.9116770926422402,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02867292225201072,
      "loss": 0.8017,
      "num_input_tokens_seen": 7445152,
      "step": 12835
    },
    {
      "epoch": 1.9124218051831994,
      "grad_norm": 0.015625,
      "learning_rate": 0.02868409294012511,
      "loss": 0.7989,
      "num_input_tokens_seen": 7448192,
      "step": 12840
    },
    {
      "epoch": 1.9131665177241586,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.028695263628239497,
      "loss": 0.8532,
      "num_input_tokens_seen": 7450976,
      "step": 12845
    },
    {
      "epoch": 1.9139112302651178,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.028706434316353887,
      "loss": 0.8053,
      "num_input_tokens_seen": 7454048,
      "step": 12850
    },
    {
      "epoch": 1.914655942806077,
      "grad_norm": 0.015625,
      "learning_rate": 0.028717605004468273,
      "loss": 0.7914,
      "num_input_tokens_seen": 7457248,
      "step": 12855
    },
    {
      "epoch": 1.9154006553470362,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028728775692582663,
      "loss": 0.7771,
      "num_input_tokens_seen": 7460192,
      "step": 12860
    },
    {
      "epoch": 1.9161453678879954,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02873994638069705,
      "loss": 0.8124,
      "num_input_tokens_seen": 7463008,
      "step": 12865
    },
    {
      "epoch": 1.9168900804289544,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02875111706881144,
      "loss": 0.8264,
      "num_input_tokens_seen": 7465952,
      "step": 12870
    },
    {
      "epoch": 1.9176347929699136,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.028762287756925825,
      "loss": 0.7962,
      "num_input_tokens_seen": 7468864,
      "step": 12875
    },
    {
      "epoch": 1.9183795055108728,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.028773458445040215,
      "loss": 0.8089,
      "num_input_tokens_seen": 7472032,
      "step": 12880
    },
    {
      "epoch": 1.919124218051832,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0287846291331546,
      "loss": 0.8057,
      "num_input_tokens_seen": 7474912,
      "step": 12885
    },
    {
      "epoch": 1.9198689305927912,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.02879579982126899,
      "loss": 0.81,
      "num_input_tokens_seen": 7477568,
      "step": 12890
    },
    {
      "epoch": 1.9206136431337504,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.028806970509383377,
      "loss": 0.8438,
      "num_input_tokens_seen": 7480256,
      "step": 12895
    },
    {
      "epoch": 1.9213583556747096,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.028818141197497767,
      "loss": 0.8238,
      "num_input_tokens_seen": 7483072,
      "step": 12900
    },
    {
      "epoch": 1.9221030682156688,
      "grad_norm": 0.015625,
      "learning_rate": 0.02882931188561215,
      "loss": 0.7848,
      "num_input_tokens_seen": 7486016,
      "step": 12905
    },
    {
      "epoch": 1.922847780756628,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02884048257372654,
      "loss": 0.8141,
      "num_input_tokens_seen": 7488864,
      "step": 12910
    },
    {
      "epoch": 1.923592493297587,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02885165326184093,
      "loss": 0.7981,
      "num_input_tokens_seen": 7491808,
      "step": 12915
    },
    {
      "epoch": 1.9243372058385462,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.028862823949955315,
      "loss": 0.8048,
      "num_input_tokens_seen": 7495072,
      "step": 12920
    },
    {
      "epoch": 1.9250819183795054,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.028873994638069705,
      "loss": 0.7991,
      "num_input_tokens_seen": 7497984,
      "step": 12925
    },
    {
      "epoch": 1.9258266309204646,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02888516532618409,
      "loss": 0.8037,
      "num_input_tokens_seen": 7500736,
      "step": 12930
    },
    {
      "epoch": 1.9265713434614238,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02889633601429848,
      "loss": 0.8217,
      "num_input_tokens_seen": 7503520,
      "step": 12935
    },
    {
      "epoch": 1.927316056002383,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028907506702412867,
      "loss": 0.7863,
      "num_input_tokens_seen": 7506304,
      "step": 12940
    },
    {
      "epoch": 1.9280607685433422,
      "grad_norm": 0.0234375,
      "learning_rate": 0.028918677390527257,
      "loss": 0.8002,
      "num_input_tokens_seen": 7509248,
      "step": 12945
    },
    {
      "epoch": 1.9288054810843014,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028929848078641643,
      "loss": 0.8146,
      "num_input_tokens_seen": 7512224,
      "step": 12950
    },
    {
      "epoch": 1.9295501936252606,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.028941018766756033,
      "loss": 0.7993,
      "num_input_tokens_seen": 7515328,
      "step": 12955
    },
    {
      "epoch": 1.9302949061662198,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02895218945487042,
      "loss": 0.7948,
      "num_input_tokens_seen": 7517984,
      "step": 12960
    },
    {
      "epoch": 1.931039618707179,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.02896336014298481,
      "loss": 0.8247,
      "num_input_tokens_seen": 7520864,
      "step": 12965
    },
    {
      "epoch": 1.9317843312481382,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.028974530831099195,
      "loss": 0.8083,
      "num_input_tokens_seen": 7524000,
      "step": 12970
    },
    {
      "epoch": 1.9325290437890974,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.028985701519213584,
      "loss": 0.8174,
      "num_input_tokens_seen": 7526688,
      "step": 12975
    },
    {
      "epoch": 1.9332737563300566,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02899687220732797,
      "loss": 0.804,
      "num_input_tokens_seen": 7529440,
      "step": 12980
    },
    {
      "epoch": 1.9340184688710158,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02900804289544236,
      "loss": 0.7953,
      "num_input_tokens_seen": 7532448,
      "step": 12985
    },
    {
      "epoch": 1.934763181411975,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.029019213583556743,
      "loss": 0.813,
      "num_input_tokens_seen": 7535328,
      "step": 12990
    },
    {
      "epoch": 1.9355078939529342,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029030384271671133,
      "loss": 0.8121,
      "num_input_tokens_seen": 7537984,
      "step": 12995
    },
    {
      "epoch": 1.9362526064938934,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02904155495978552,
      "loss": 0.8,
      "num_input_tokens_seen": 7540928,
      "step": 13000
    },
    {
      "epoch": 1.9369973190348526,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02905272564789991,
      "loss": 0.808,
      "num_input_tokens_seen": 7543680,
      "step": 13005
    },
    {
      "epoch": 1.9377420315758118,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.029063896336014295,
      "loss": 0.812,
      "num_input_tokens_seen": 7546720,
      "step": 13010
    },
    {
      "epoch": 1.938486744116771,
      "grad_norm": 0.013671875,
      "learning_rate": 0.029075067024128685,
      "loss": 0.8153,
      "num_input_tokens_seen": 7549920,
      "step": 13015
    },
    {
      "epoch": 1.9392314566577302,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02908623771224307,
      "loss": 0.817,
      "num_input_tokens_seen": 7552960,
      "step": 13020
    },
    {
      "epoch": 1.9399761691986894,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02909740840035746,
      "loss": 0.8004,
      "num_input_tokens_seen": 7555584,
      "step": 13025
    },
    {
      "epoch": 1.9407208817396486,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02910857908847185,
      "loss": 0.8,
      "num_input_tokens_seen": 7558688,
      "step": 13030
    },
    {
      "epoch": 1.9414655942806078,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029119749776586237,
      "loss": 0.8025,
      "num_input_tokens_seen": 7561344,
      "step": 13035
    },
    {
      "epoch": 1.942210306821567,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029130920464700626,
      "loss": 0.8105,
      "num_input_tokens_seen": 7564544,
      "step": 13040
    },
    {
      "epoch": 1.942955019362526,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.029142091152815013,
      "loss": 0.8021,
      "num_input_tokens_seen": 7567488,
      "step": 13045
    },
    {
      "epoch": 1.9436997319034852,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029153261840929402,
      "loss": 0.8008,
      "num_input_tokens_seen": 7570048,
      "step": 13050
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02916443252904379,
      "loss": 0.8,
      "num_input_tokens_seen": 7572704,
      "step": 13055
    },
    {
      "epoch": 1.9451891569854036,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.029175603217158178,
      "loss": 0.7983,
      "num_input_tokens_seen": 7575712,
      "step": 13060
    },
    {
      "epoch": 1.9459338695263628,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.029186773905272564,
      "loss": 0.7983,
      "num_input_tokens_seen": 7578496,
      "step": 13065
    },
    {
      "epoch": 1.946678582067322,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029197944593386954,
      "loss": 0.7881,
      "num_input_tokens_seen": 7581248,
      "step": 13070
    },
    {
      "epoch": 1.9474232946082812,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029209115281501337,
      "loss": 0.7989,
      "num_input_tokens_seen": 7583712,
      "step": 13075
    },
    {
      "epoch": 1.9481680071492404,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029220285969615727,
      "loss": 0.796,
      "num_input_tokens_seen": 7586336,
      "step": 13080
    },
    {
      "epoch": 1.9489127196901996,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.029231456657730113,
      "loss": 0.804,
      "num_input_tokens_seen": 7589216,
      "step": 13085
    },
    {
      "epoch": 1.9496574322311586,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029242627345844503,
      "loss": 0.7961,
      "num_input_tokens_seen": 7592128,
      "step": 13090
    },
    {
      "epoch": 1.9504021447721178,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.02925379803395889,
      "loss": 0.7939,
      "num_input_tokens_seen": 7594880,
      "step": 13095
    },
    {
      "epoch": 1.951146857313077,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.02926496872207328,
      "loss": 0.7988,
      "num_input_tokens_seen": 7597632,
      "step": 13100
    },
    {
      "epoch": 1.9518915698540362,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029276139410187665,
      "loss": 0.8022,
      "num_input_tokens_seen": 7600672,
      "step": 13105
    },
    {
      "epoch": 1.9526362823949954,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029287310098302054,
      "loss": 0.8199,
      "num_input_tokens_seen": 7603680,
      "step": 13110
    },
    {
      "epoch": 1.9533809949359546,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02929848078641644,
      "loss": 0.7994,
      "num_input_tokens_seen": 7606464,
      "step": 13115
    },
    {
      "epoch": 1.9541257074769138,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.02930965147453083,
      "loss": 0.8039,
      "num_input_tokens_seen": 7609408,
      "step": 13120
    },
    {
      "epoch": 1.954870420017873,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.029320822162645217,
      "loss": 0.8302,
      "num_input_tokens_seen": 7612960,
      "step": 13125
    },
    {
      "epoch": 1.9556151325588322,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029331992850759606,
      "loss": 0.7856,
      "num_input_tokens_seen": 7615680,
      "step": 13130
    },
    {
      "epoch": 1.9563598450997914,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029343163538873993,
      "loss": 0.7912,
      "num_input_tokens_seen": 7618752,
      "step": 13135
    },
    {
      "epoch": 1.9571045576407506,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.029354334226988382,
      "loss": 0.7992,
      "num_input_tokens_seen": 7621472,
      "step": 13140
    },
    {
      "epoch": 1.9578492701817098,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.029365504915102772,
      "loss": 0.7951,
      "num_input_tokens_seen": 7624480,
      "step": 13145
    },
    {
      "epoch": 1.958593982722669,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.029376675603217158,
      "loss": 0.8131,
      "num_input_tokens_seen": 7627360,
      "step": 13150
    },
    {
      "epoch": 1.9593386952636282,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.029387846291331548,
      "loss": 0.8131,
      "num_input_tokens_seen": 7630336,
      "step": 13155
    },
    {
      "epoch": 1.9600834078045875,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029399016979445934,
      "loss": 0.8123,
      "num_input_tokens_seen": 7632992,
      "step": 13160
    },
    {
      "epoch": 1.9608281203455467,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.02941018766756032,
      "loss": 0.8161,
      "num_input_tokens_seen": 7635648,
      "step": 13165
    },
    {
      "epoch": 1.9615728328865059,
      "grad_norm": 0.037109375,
      "learning_rate": 0.029421358355674707,
      "loss": 0.7919,
      "num_input_tokens_seen": 7638656,
      "step": 13170
    },
    {
      "epoch": 1.962317545427465,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029432529043789096,
      "loss": 0.7925,
      "num_input_tokens_seen": 7641760,
      "step": 13175
    },
    {
      "epoch": 1.9630622579684243,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.029443699731903483,
      "loss": 0.7941,
      "num_input_tokens_seen": 7644448,
      "step": 13180
    },
    {
      "epoch": 1.9638069705093835,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029454870420017872,
      "loss": 0.793,
      "num_input_tokens_seen": 7647392,
      "step": 13185
    },
    {
      "epoch": 1.9645516830503427,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02946604110813226,
      "loss": 0.7878,
      "num_input_tokens_seen": 7650304,
      "step": 13190
    },
    {
      "epoch": 1.9652963955913019,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.029477211796246648,
      "loss": 0.7764,
      "num_input_tokens_seen": 7652960,
      "step": 13195
    },
    {
      "epoch": 1.966041108132261,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029488382484361034,
      "loss": 0.7838,
      "num_input_tokens_seen": 7655808,
      "step": 13200
    },
    {
      "epoch": 1.9667858206732203,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.029499553172475424,
      "loss": 0.7829,
      "num_input_tokens_seen": 7658560,
      "step": 13205
    },
    {
      "epoch": 1.9675305332141795,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02951072386058981,
      "loss": 0.7372,
      "num_input_tokens_seen": 7661568,
      "step": 13210
    },
    {
      "epoch": 1.9682752457551387,
      "grad_norm": 0.029296875,
      "learning_rate": 0.0295218945487042,
      "loss": 0.7826,
      "num_input_tokens_seen": 7664352,
      "step": 13215
    },
    {
      "epoch": 1.9690199582960977,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.029533065236818586,
      "loss": 0.8313,
      "num_input_tokens_seen": 7667392,
      "step": 13220
    },
    {
      "epoch": 1.9697646708370569,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029544235924932976,
      "loss": 0.8264,
      "num_input_tokens_seen": 7670240,
      "step": 13225
    },
    {
      "epoch": 1.970509383378016,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029555406613047362,
      "loss": 0.779,
      "num_input_tokens_seen": 7673312,
      "step": 13230
    },
    {
      "epoch": 1.9712540959189753,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.029566577301161752,
      "loss": 0.7962,
      "num_input_tokens_seen": 7676448,
      "step": 13235
    },
    {
      "epoch": 1.9719988084599345,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.029577747989276138,
      "loss": 0.8006,
      "num_input_tokens_seen": 7679360,
      "step": 13240
    },
    {
      "epoch": 1.9727435210008937,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.029588918677390528,
      "loss": 0.7831,
      "num_input_tokens_seen": 7682272,
      "step": 13245
    },
    {
      "epoch": 1.9734882335418529,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029600089365504914,
      "loss": 0.8102,
      "num_input_tokens_seen": 7685280,
      "step": 13250
    },
    {
      "epoch": 1.974232946082812,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0296112600536193,
      "loss": 0.8037,
      "num_input_tokens_seen": 7687968,
      "step": 13255
    },
    {
      "epoch": 1.974977658623771,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.02962243074173369,
      "loss": 0.7791,
      "num_input_tokens_seen": 7690976,
      "step": 13260
    },
    {
      "epoch": 1.9757223711647303,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.029633601429848076,
      "loss": 0.8314,
      "num_input_tokens_seen": 7693856,
      "step": 13265
    },
    {
      "epoch": 1.9764670837056895,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029644772117962466,
      "loss": 0.7831,
      "num_input_tokens_seen": 7696672,
      "step": 13270
    },
    {
      "epoch": 1.9772117962466487,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029655942806076852,
      "loss": 0.828,
      "num_input_tokens_seen": 7699808,
      "step": 13275
    },
    {
      "epoch": 1.9779565087876079,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.029667113494191242,
      "loss": 0.7942,
      "num_input_tokens_seen": 7702720,
      "step": 13280
    },
    {
      "epoch": 1.978701221328567,
      "grad_norm": 0.01031494140625,
      "learning_rate": 0.029678284182305628,
      "loss": 0.8058,
      "num_input_tokens_seen": 7705504,
      "step": 13285
    },
    {
      "epoch": 1.9794459338695263,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029689454870420018,
      "loss": 0.7945,
      "num_input_tokens_seen": 7708672,
      "step": 13290
    },
    {
      "epoch": 1.9801906464104855,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029700625558534404,
      "loss": 0.8177,
      "num_input_tokens_seen": 7711744,
      "step": 13295
    },
    {
      "epoch": 1.9809353589514447,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.029711796246648794,
      "loss": 0.79,
      "num_input_tokens_seen": 7714656,
      "step": 13300
    },
    {
      "epoch": 1.9816800714924039,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02972296693476318,
      "loss": 0.8201,
      "num_input_tokens_seen": 7717344,
      "step": 13305
    },
    {
      "epoch": 1.982424784033363,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02973413762287757,
      "loss": 0.7921,
      "num_input_tokens_seen": 7720384,
      "step": 13310
    },
    {
      "epoch": 1.9831694965743223,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.029745308310991956,
      "loss": 0.7989,
      "num_input_tokens_seen": 7723040,
      "step": 13315
    },
    {
      "epoch": 1.9839142091152815,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029756478999106346,
      "loss": 0.8046,
      "num_input_tokens_seen": 7725920,
      "step": 13320
    },
    {
      "epoch": 1.9846589216562407,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029767649687220732,
      "loss": 0.7941,
      "num_input_tokens_seen": 7728960,
      "step": 13325
    },
    {
      "epoch": 1.9854036341972,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02977882037533512,
      "loss": 0.7976,
      "num_input_tokens_seen": 7731616,
      "step": 13330
    },
    {
      "epoch": 1.986148346738159,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029789991063449504,
      "loss": 0.805,
      "num_input_tokens_seen": 7734688,
      "step": 13335
    },
    {
      "epoch": 1.9868930592791183,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029801161751563894,
      "loss": 0.791,
      "num_input_tokens_seen": 7737376,
      "step": 13340
    },
    {
      "epoch": 1.9876377718200775,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02981233243967828,
      "loss": 0.8079,
      "num_input_tokens_seen": 7740352,
      "step": 13345
    },
    {
      "epoch": 1.9883824843610367,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02982350312779267,
      "loss": 0.7846,
      "num_input_tokens_seen": 7743200,
      "step": 13350
    },
    {
      "epoch": 1.989127196901996,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029834673815907056,
      "loss": 0.7621,
      "num_input_tokens_seen": 7746144,
      "step": 13355
    },
    {
      "epoch": 1.9898719094429551,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.029845844504021446,
      "loss": 0.8097,
      "num_input_tokens_seen": 7749024,
      "step": 13360
    },
    {
      "epoch": 1.9906166219839143,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.029857015192135836,
      "loss": 0.8146,
      "num_input_tokens_seen": 7752224,
      "step": 13365
    },
    {
      "epoch": 1.9913613345248735,
      "grad_norm": 0.03125,
      "learning_rate": 0.029868185880250222,
      "loss": 0.7938,
      "num_input_tokens_seen": 7755008,
      "step": 13370
    },
    {
      "epoch": 1.9921060470658327,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02987935656836461,
      "loss": 0.8186,
      "num_input_tokens_seen": 7757728,
      "step": 13375
    },
    {
      "epoch": 1.992850759606792,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029890527256478998,
      "loss": 0.7959,
      "num_input_tokens_seen": 7760640,
      "step": 13380
    },
    {
      "epoch": 1.9935954721477511,
      "grad_norm": 0.01019287109375,
      "learning_rate": 0.029901697944593388,
      "loss": 0.7815,
      "num_input_tokens_seen": 7763552,
      "step": 13385
    },
    {
      "epoch": 1.99434018468871,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029912868632707774,
      "loss": 0.794,
      "num_input_tokens_seen": 7766368,
      "step": 13390
    },
    {
      "epoch": 1.9950848972296693,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029924039320822163,
      "loss": 0.8233,
      "num_input_tokens_seen": 7769664,
      "step": 13395
    },
    {
      "epoch": 1.9958296097706285,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.02993521000893655,
      "loss": 0.7885,
      "num_input_tokens_seen": 7772704,
      "step": 13400
    },
    {
      "epoch": 1.9965743223115877,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02994638069705094,
      "loss": 0.8068,
      "num_input_tokens_seen": 7776032,
      "step": 13405
    },
    {
      "epoch": 1.997319034852547,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029957551385165326,
      "loss": 0.7839,
      "num_input_tokens_seen": 7779008,
      "step": 13410
    },
    {
      "epoch": 1.9980637473935061,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029968722073279715,
      "loss": 0.8018,
      "num_input_tokens_seen": 7781888,
      "step": 13415
    },
    {
      "epoch": 1.9988084599344653,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.0299798927613941,
      "loss": 0.8059,
      "num_input_tokens_seen": 7785120,
      "step": 13420
    },
    {
      "epoch": 1.9995531724754245,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029991063449508488,
      "loss": 0.8422,
      "num_input_tokens_seen": 7788096,
      "step": 13425
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.8020058870315552,
      "eval_runtime": 70.6448,
      "eval_samples_per_second": 42.24,
      "eval_steps_per_second": 10.56,
      "num_input_tokens_seen": 7789256,
      "step": 13428
    },
    {
      "epoch": 2.0002978850163835,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029999999994931803,
      "loss": 0.8008,
      "num_input_tokens_seen": 7790536,
      "step": 13430
    },
    {
      "epoch": 2.0010425975573427,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029999999817544976,
      "loss": 0.8042,
      "num_input_tokens_seen": 7793416,
      "step": 13435
    },
    {
      "epoch": 2.001787310098302,
      "grad_norm": 0.03125,
      "learning_rate": 0.029999999386748397,
      "loss": 0.8011,
      "num_input_tokens_seen": 7796296,
      "step": 13440
    },
    {
      "epoch": 2.002532022639261,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02999999870254207,
      "loss": 0.7854,
      "num_input_tokens_seen": 7799144,
      "step": 13445
    },
    {
      "epoch": 2.0032767351802203,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029999997764926012,
      "loss": 0.8005,
      "num_input_tokens_seen": 7802088,
      "step": 13450
    },
    {
      "epoch": 2.0040214477211795,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02999999657390024,
      "loss": 0.7948,
      "num_input_tokens_seen": 7805128,
      "step": 13455
    },
    {
      "epoch": 2.0047661602621387,
      "grad_norm": 0.01055908203125,
      "learning_rate": 0.02999999512946477,
      "loss": 0.8036,
      "num_input_tokens_seen": 7807880,
      "step": 13460
    },
    {
      "epoch": 2.005510872803098,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.029999993431619625,
      "loss": 0.8612,
      "num_input_tokens_seen": 7811080,
      "step": 13465
    },
    {
      "epoch": 2.006255585344057,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029999991480364842,
      "loss": 0.7916,
      "num_input_tokens_seen": 7814120,
      "step": 13470
    },
    {
      "epoch": 2.0070002978850163,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.029999989275700443,
      "loss": 0.7857,
      "num_input_tokens_seen": 7816872,
      "step": 13475
    },
    {
      "epoch": 2.0077450104259755,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029999986817626473,
      "loss": 0.8227,
      "num_input_tokens_seen": 7819464,
      "step": 13480
    },
    {
      "epoch": 2.0084897229669347,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02999998410614297,
      "loss": 0.7951,
      "num_input_tokens_seen": 7822216,
      "step": 13485
    },
    {
      "epoch": 2.009234435507894,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.02999998114124998,
      "loss": 0.787,
      "num_input_tokens_seen": 7825128,
      "step": 13490
    },
    {
      "epoch": 2.009979148048853,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.029999977922947556,
      "loss": 0.8191,
      "num_input_tokens_seen": 7828264,
      "step": 13495
    },
    {
      "epoch": 2.0107238605898123,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02999997445123575,
      "loss": 0.7929,
      "num_input_tokens_seen": 7831016,
      "step": 13500
    },
    {
      "epoch": 2.0114685731307715,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029999970726114617,
      "loss": 0.8079,
      "num_input_tokens_seen": 7834056,
      "step": 13505
    },
    {
      "epoch": 2.0122132856717307,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029999966747584227,
      "loss": 0.7975,
      "num_input_tokens_seen": 7836648,
      "step": 13510
    },
    {
      "epoch": 2.01295799821269,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.029999962515644647,
      "loss": 0.8147,
      "num_input_tokens_seen": 7839720,
      "step": 13515
    },
    {
      "epoch": 2.013702710753649,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02999995803029594,
      "loss": 0.8096,
      "num_input_tokens_seen": 7842696,
      "step": 13520
    },
    {
      "epoch": 2.0144474232946084,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.029999953291538192,
      "loss": 0.7975,
      "num_input_tokens_seen": 7845576,
      "step": 13525
    },
    {
      "epoch": 2.0151921358355676,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029999948299371474,
      "loss": 0.8253,
      "num_input_tokens_seen": 7848200,
      "step": 13530
    },
    {
      "epoch": 2.0159368483765268,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02999994305379588,
      "loss": 0.7833,
      "num_input_tokens_seen": 7850856,
      "step": 13535
    },
    {
      "epoch": 2.016681560917486,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02999993755481149,
      "loss": 0.7802,
      "num_input_tokens_seen": 7853608,
      "step": 13540
    },
    {
      "epoch": 2.017426273458445,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.0299999318024184,
      "loss": 0.7778,
      "num_input_tokens_seen": 7856552,
      "step": 13545
    },
    {
      "epoch": 2.0181709859994044,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029999925796616707,
      "loss": 0.8001,
      "num_input_tokens_seen": 7859496,
      "step": 13550
    },
    {
      "epoch": 2.0189156985403636,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02999991953740652,
      "loss": 0.8089,
      "num_input_tokens_seen": 7862632,
      "step": 13555
    },
    {
      "epoch": 2.0196604110813228,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029999913024787925,
      "loss": 0.7918,
      "num_input_tokens_seen": 7865480,
      "step": 13560
    },
    {
      "epoch": 2.020405123622282,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029999906258761054,
      "loss": 0.8245,
      "num_input_tokens_seen": 7868392,
      "step": 13565
    },
    {
      "epoch": 2.021149836163241,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029999899239326007,
      "loss": 0.7723,
      "num_input_tokens_seen": 7871208,
      "step": 13570
    },
    {
      "epoch": 2.0218945487042004,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029999891966482912,
      "loss": 0.8241,
      "num_input_tokens_seen": 7873896,
      "step": 13575
    },
    {
      "epoch": 2.0226392612451596,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029999884440231883,
      "loss": 0.7913,
      "num_input_tokens_seen": 7876616,
      "step": 13580
    },
    {
      "epoch": 2.0233839737861183,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029999876660573056,
      "loss": 0.8459,
      "num_input_tokens_seen": 7879592,
      "step": 13585
    },
    {
      "epoch": 2.0241286863270775,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.029999868627506556,
      "loss": 0.7886,
      "num_input_tokens_seen": 7882120,
      "step": 13590
    },
    {
      "epoch": 2.0248733988680367,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02999986034103252,
      "loss": 0.7896,
      "num_input_tokens_seen": 7884808,
      "step": 13595
    },
    {
      "epoch": 2.025618111408996,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029999851801151094,
      "loss": 0.8087,
      "num_input_tokens_seen": 7887560,
      "step": 13600
    },
    {
      "epoch": 2.026362823949955,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02999984300786241,
      "loss": 0.8023,
      "num_input_tokens_seen": 7890248,
      "step": 13605
    },
    {
      "epoch": 2.0271075364909144,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.029999833961166622,
      "loss": 0.7887,
      "num_input_tokens_seen": 7893352,
      "step": 13610
    },
    {
      "epoch": 2.0278522490318736,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029999824661063892,
      "loss": 0.7812,
      "num_input_tokens_seen": 7896136,
      "step": 13615
    },
    {
      "epoch": 2.0285969615728328,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.029999815107554364,
      "loss": 0.8429,
      "num_input_tokens_seen": 7899112,
      "step": 13620
    },
    {
      "epoch": 2.029341674113792,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.029999805300638206,
      "loss": 0.8085,
      "num_input_tokens_seen": 7902120,
      "step": 13625
    },
    {
      "epoch": 2.030086386654751,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.029999795240315583,
      "loss": 0.8142,
      "num_input_tokens_seen": 7905384,
      "step": 13630
    },
    {
      "epoch": 2.0308310991957104,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029999784926586667,
      "loss": 0.7991,
      "num_input_tokens_seen": 7908616,
      "step": 13635
    },
    {
      "epoch": 2.0315758117366696,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029999774359451628,
      "loss": 0.806,
      "num_input_tokens_seen": 7911688,
      "step": 13640
    },
    {
      "epoch": 2.0323205242776288,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.029999763538910644,
      "loss": 0.8023,
      "num_input_tokens_seen": 7914664,
      "step": 13645
    },
    {
      "epoch": 2.033065236818588,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029999752464963905,
      "loss": 0.8108,
      "num_input_tokens_seen": 7917800,
      "step": 13650
    },
    {
      "epoch": 2.033809949359547,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02999974113761159,
      "loss": 0.8,
      "num_input_tokens_seen": 7920776,
      "step": 13655
    },
    {
      "epoch": 2.0345546619005064,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.029999729556853896,
      "loss": 0.8178,
      "num_input_tokens_seen": 7923592,
      "step": 13660
    },
    {
      "epoch": 2.0352993744414656,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029999717722691013,
      "loss": 0.7963,
      "num_input_tokens_seen": 7926632,
      "step": 13665
    },
    {
      "epoch": 2.036044086982425,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029999705635123143,
      "loss": 0.7933,
      "num_input_tokens_seen": 7929384,
      "step": 13670
    },
    {
      "epoch": 2.036788799523384,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.029999693294150493,
      "loss": 0.7814,
      "num_input_tokens_seen": 7932200,
      "step": 13675
    },
    {
      "epoch": 2.037533512064343,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02999968069977327,
      "loss": 0.8255,
      "num_input_tokens_seen": 7935080,
      "step": 13680
    },
    {
      "epoch": 2.0382782246053024,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02999966785199169,
      "loss": 0.8191,
      "num_input_tokens_seen": 7937960,
      "step": 13685
    },
    {
      "epoch": 2.0390229371462616,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02999965475080596,
      "loss": 0.815,
      "num_input_tokens_seen": 7941032,
      "step": 13690
    },
    {
      "epoch": 2.039767649687221,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02999964139621631,
      "loss": 0.8103,
      "num_input_tokens_seen": 7943976,
      "step": 13695
    },
    {
      "epoch": 2.04051236222818,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029999627788222967,
      "loss": 0.801,
      "num_input_tokens_seen": 7946952,
      "step": 13700
    },
    {
      "epoch": 2.041257074769139,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.029999613926826155,
      "loss": 0.8152,
      "num_input_tokens_seen": 7949736,
      "step": 13705
    },
    {
      "epoch": 2.0420017873100984,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.02999959981202611,
      "loss": 0.801,
      "num_input_tokens_seen": 7952648,
      "step": 13710
    },
    {
      "epoch": 2.0427464998510576,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02999958544382307,
      "loss": 0.7953,
      "num_input_tokens_seen": 7955912,
      "step": 13715
    },
    {
      "epoch": 2.043491212392017,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.029999570822217284,
      "loss": 0.8124,
      "num_input_tokens_seen": 7958856,
      "step": 13720
    },
    {
      "epoch": 2.044235924932976,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.02999955594720899,
      "loss": 0.806,
      "num_input_tokens_seen": 7961992,
      "step": 13725
    },
    {
      "epoch": 2.044980637473935,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.029999540818798444,
      "loss": 0.8116,
      "num_input_tokens_seen": 7964712,
      "step": 13730
    },
    {
      "epoch": 2.0457253500148944,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.0299995254369859,
      "loss": 0.8014,
      "num_input_tokens_seen": 7967656,
      "step": 13735
    },
    {
      "epoch": 2.0464700625558536,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.029999509801771615,
      "loss": 0.7826,
      "num_input_tokens_seen": 7970696,
      "step": 13740
    },
    {
      "epoch": 2.047214775096813,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029999493913155865,
      "loss": 0.793,
      "num_input_tokens_seen": 7973192,
      "step": 13745
    },
    {
      "epoch": 2.047959487637772,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029999477771138903,
      "loss": 0.8078,
      "num_input_tokens_seen": 7976168,
      "step": 13750
    },
    {
      "epoch": 2.0487042001787312,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.029999461375721015,
      "loss": 0.8064,
      "num_input_tokens_seen": 7979240,
      "step": 13755
    },
    {
      "epoch": 2.04944891271969,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02999944472690247,
      "loss": 0.8208,
      "num_input_tokens_seen": 7982088,
      "step": 13760
    },
    {
      "epoch": 2.050193625260649,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029999427824683546,
      "loss": 0.8046,
      "num_input_tokens_seen": 7984904,
      "step": 13765
    },
    {
      "epoch": 2.0509383378016084,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02999941066906454,
      "loss": 0.803,
      "num_input_tokens_seen": 7987976,
      "step": 13770
    },
    {
      "epoch": 2.0516830503425676,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029999393260045734,
      "loss": 0.8049,
      "num_input_tokens_seen": 7990664,
      "step": 13775
    },
    {
      "epoch": 2.052427762883527,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029999375597627418,
      "loss": 0.8068,
      "num_input_tokens_seen": 7994856,
      "step": 13780
    },
    {
      "epoch": 2.053172475424486,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029999357681809904,
      "loss": 0.8054,
      "num_input_tokens_seen": 7997608,
      "step": 13785
    },
    {
      "epoch": 2.053917187965445,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02999933951259348,
      "loss": 0.8118,
      "num_input_tokens_seen": 8000488,
      "step": 13790
    },
    {
      "epoch": 2.0546619005064044,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.029999321089978465,
      "loss": 0.8059,
      "num_input_tokens_seen": 8003368,
      "step": 13795
    },
    {
      "epoch": 2.0554066130473636,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029999302413965163,
      "loss": 0.7983,
      "num_input_tokens_seen": 8006280,
      "step": 13800
    },
    {
      "epoch": 2.056151325588323,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02999928348455389,
      "loss": 0.7901,
      "num_input_tokens_seen": 8009160,
      "step": 13805
    },
    {
      "epoch": 2.056896038129282,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02999926430174497,
      "loss": 0.7801,
      "num_input_tokens_seen": 8012168,
      "step": 13810
    },
    {
      "epoch": 2.057640750670241,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02999924486553872,
      "loss": 0.8073,
      "num_input_tokens_seen": 8015016,
      "step": 13815
    },
    {
      "epoch": 2.0583854632112004,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.029999225175935472,
      "loss": 0.8287,
      "num_input_tokens_seen": 8017864,
      "step": 13820
    },
    {
      "epoch": 2.0591301757521596,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029999205232935564,
      "loss": 0.8077,
      "num_input_tokens_seen": 8020936,
      "step": 13825
    },
    {
      "epoch": 2.059874888293119,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029999185036539325,
      "loss": 0.7836,
      "num_input_tokens_seen": 8023592,
      "step": 13830
    },
    {
      "epoch": 2.060619600834078,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.0299991645867471,
      "loss": 0.8135,
      "num_input_tokens_seen": 8026792,
      "step": 13835
    },
    {
      "epoch": 2.0613643133750372,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02999914388355923,
      "loss": 0.8265,
      "num_input_tokens_seen": 8029448,
      "step": 13840
    },
    {
      "epoch": 2.0621090259159964,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.02999912292697608,
      "loss": 0.8245,
      "num_input_tokens_seen": 8032424,
      "step": 13845
    },
    {
      "epoch": 2.0628537384569556,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029999101716997982,
      "loss": 0.8041,
      "num_input_tokens_seen": 8035080,
      "step": 13850
    },
    {
      "epoch": 2.063598450997915,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02999908025362531,
      "loss": 0.8018,
      "num_input_tokens_seen": 8038056,
      "step": 13855
    },
    {
      "epoch": 2.064343163538874,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02999905853685842,
      "loss": 0.7937,
      "num_input_tokens_seen": 8040840,
      "step": 13860
    },
    {
      "epoch": 2.0650878760798332,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029999036566697675,
      "loss": 0.7913,
      "num_input_tokens_seen": 8043912,
      "step": 13865
    },
    {
      "epoch": 2.0658325886207924,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02999901434314346,
      "loss": 0.8102,
      "num_input_tokens_seen": 8046824,
      "step": 13870
    },
    {
      "epoch": 2.0665773011617516,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.029998991866196135,
      "loss": 0.7817,
      "num_input_tokens_seen": 8049704,
      "step": 13875
    },
    {
      "epoch": 2.067322013702711,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02999896913585609,
      "loss": 0.7774,
      "num_input_tokens_seen": 8052680,
      "step": 13880
    },
    {
      "epoch": 2.06806672624367,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029998946152123708,
      "loss": 0.7859,
      "num_input_tokens_seen": 8055400,
      "step": 13885
    },
    {
      "epoch": 2.0688114387846293,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029998922914999372,
      "loss": 0.8237,
      "num_input_tokens_seen": 8058312,
      "step": 13890
    },
    {
      "epoch": 2.0695561513255885,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.02999889942448348,
      "loss": 0.7853,
      "num_input_tokens_seen": 8061160,
      "step": 13895
    },
    {
      "epoch": 2.0703008638665477,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029998875680576423,
      "loss": 0.763,
      "num_input_tokens_seen": 8064424,
      "step": 13900
    },
    {
      "epoch": 2.071045576407507,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029998851683278607,
      "loss": 0.7773,
      "num_input_tokens_seen": 8067336,
      "step": 13905
    },
    {
      "epoch": 2.071790288948466,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029998827432590438,
      "loss": 0.811,
      "num_input_tokens_seen": 8070600,
      "step": 13910
    },
    {
      "epoch": 2.0725350014894253,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.029998802928512324,
      "loss": 0.7897,
      "num_input_tokens_seen": 8073352,
      "step": 13915
    },
    {
      "epoch": 2.0732797140303845,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.029998778171044675,
      "loss": 0.8355,
      "num_input_tokens_seen": 8076456,
      "step": 13920
    },
    {
      "epoch": 2.0740244265713437,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029998753160187915,
      "loss": 0.7819,
      "num_input_tokens_seen": 8079208,
      "step": 13925
    },
    {
      "epoch": 2.074769139112303,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.029998727895942462,
      "loss": 0.7878,
      "num_input_tokens_seen": 8082280,
      "step": 13930
    },
    {
      "epoch": 2.0755138516532616,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.029998702378308748,
      "loss": 0.8167,
      "num_input_tokens_seen": 8084968,
      "step": 13935
    },
    {
      "epoch": 2.076258564194221,
      "grad_norm": 0.02734375,
      "learning_rate": 0.0299986766072872,
      "loss": 0.8235,
      "num_input_tokens_seen": 8088008,
      "step": 13940
    },
    {
      "epoch": 2.07700327673518,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029998650582878254,
      "loss": 0.801,
      "num_input_tokens_seen": 8091048,
      "step": 13945
    },
    {
      "epoch": 2.0777479892761392,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029998624305082353,
      "loss": 0.8475,
      "num_input_tokens_seen": 8093992,
      "step": 13950
    },
    {
      "epoch": 2.0784927018170984,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02999859777389994,
      "loss": 0.8027,
      "num_input_tokens_seen": 8096648,
      "step": 13955
    },
    {
      "epoch": 2.0792374143580576,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.029998570989331456,
      "loss": 0.7849,
      "num_input_tokens_seen": 8099624,
      "step": 13960
    },
    {
      "epoch": 2.079982126899017,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02999854395137736,
      "loss": 0.7933,
      "num_input_tokens_seen": 8102504,
      "step": 13965
    },
    {
      "epoch": 2.080726839439976,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.029998516660038107,
      "loss": 0.7819,
      "num_input_tokens_seen": 8105224,
      "step": 13970
    },
    {
      "epoch": 2.0814715519809353,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029998489115314164,
      "loss": 0.8155,
      "num_input_tokens_seen": 8108040,
      "step": 13975
    },
    {
      "epoch": 2.0822162645218945,
      "grad_norm": 0.01123046875,
      "learning_rate": 0.029998461317205988,
      "loss": 0.7998,
      "num_input_tokens_seen": 8111144,
      "step": 13980
    },
    {
      "epoch": 2.0829609770628537,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029998433265714053,
      "loss": 0.8274,
      "num_input_tokens_seen": 8114056,
      "step": 13985
    },
    {
      "epoch": 2.083705689603813,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029998404960838828,
      "loss": 0.8033,
      "num_input_tokens_seen": 8116744,
      "step": 13990
    },
    {
      "epoch": 2.084450402144772,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.029998376402580795,
      "loss": 0.7938,
      "num_input_tokens_seen": 8120008,
      "step": 13995
    },
    {
      "epoch": 2.0851951146857313,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.029998347590940444,
      "loss": 0.7749,
      "num_input_tokens_seen": 8122664,
      "step": 14000
    },
    {
      "epoch": 2.0859398272266905,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029998318525918246,
      "loss": 0.7979,
      "num_input_tokens_seen": 8125832,
      "step": 14005
    },
    {
      "epoch": 2.0866845397676497,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.0299982892075147,
      "loss": 0.7916,
      "num_input_tokens_seen": 8129032,
      "step": 14010
    },
    {
      "epoch": 2.087429252308609,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029998259635730308,
      "loss": 0.816,
      "num_input_tokens_seen": 8131880,
      "step": 14015
    },
    {
      "epoch": 2.088173964849568,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02999822981056556,
      "loss": 0.8454,
      "num_input_tokens_seen": 8134632,
      "step": 14020
    },
    {
      "epoch": 2.0889186773905273,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02999819973202096,
      "loss": 0.8102,
      "num_input_tokens_seen": 8137512,
      "step": 14025
    },
    {
      "epoch": 2.0896633899314865,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029998169400097023,
      "loss": 0.8103,
      "num_input_tokens_seen": 8140424,
      "step": 14030
    },
    {
      "epoch": 2.0904081024724457,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029998138814794253,
      "loss": 0.8051,
      "num_input_tokens_seen": 8143272,
      "step": 14035
    },
    {
      "epoch": 2.091152815013405,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029998107976113176,
      "loss": 0.8001,
      "num_input_tokens_seen": 8146376,
      "step": 14040
    },
    {
      "epoch": 2.091897527554364,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029998076884054307,
      "loss": 0.8042,
      "num_input_tokens_seen": 8149224,
      "step": 14045
    },
    {
      "epoch": 2.0926422400953233,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02999804553861817,
      "loss": 0.8002,
      "num_input_tokens_seen": 8152040,
      "step": 14050
    },
    {
      "epoch": 2.0933869526362825,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.0299980139398053,
      "loss": 0.8133,
      "num_input_tokens_seen": 8154856,
      "step": 14055
    },
    {
      "epoch": 2.0941316651772417,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029997982087616226,
      "loss": 0.8041,
      "num_input_tokens_seen": 8157512,
      "step": 14060
    },
    {
      "epoch": 2.094876377718201,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029997949982051488,
      "loss": 0.8132,
      "num_input_tokens_seen": 8160584,
      "step": 14065
    },
    {
      "epoch": 2.09562109025916,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.029997917623111627,
      "loss": 0.8009,
      "num_input_tokens_seen": 8163624,
      "step": 14070
    },
    {
      "epoch": 2.0963658028001193,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02999788501079719,
      "loss": 0.8188,
      "num_input_tokens_seen": 8166280,
      "step": 14075
    },
    {
      "epoch": 2.0971105153410785,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029997852145108736,
      "loss": 0.799,
      "num_input_tokens_seen": 8169096,
      "step": 14080
    },
    {
      "epoch": 2.0978552278820377,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02999781902604681,
      "loss": 0.8106,
      "num_input_tokens_seen": 8171944,
      "step": 14085
    },
    {
      "epoch": 2.098599940422997,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02999778565361197,
      "loss": 0.8047,
      "num_input_tokens_seen": 8175240,
      "step": 14090
    },
    {
      "epoch": 2.099344652963956,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02999775202780479,
      "loss": 0.8035,
      "num_input_tokens_seen": 8177896,
      "step": 14095
    },
    {
      "epoch": 2.1000893655049153,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.029997718148625827,
      "loss": 0.8014,
      "num_input_tokens_seen": 8180456,
      "step": 14100
    },
    {
      "epoch": 2.1008340780458745,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.029997684016075662,
      "loss": 0.7978,
      "num_input_tokens_seen": 8183272,
      "step": 14105
    },
    {
      "epoch": 2.1015787905868333,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02999764963015487,
      "loss": 0.8039,
      "num_input_tokens_seen": 8186280,
      "step": 14110
    },
    {
      "epoch": 2.1023235031277925,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029997614990864025,
      "loss": 0.8012,
      "num_input_tokens_seen": 8189288,
      "step": 14115
    },
    {
      "epoch": 2.1030682156687517,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.029997580098203726,
      "loss": 0.796,
      "num_input_tokens_seen": 8192136,
      "step": 14120
    },
    {
      "epoch": 2.103812928209711,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02999754495217455,
      "loss": 0.7822,
      "num_input_tokens_seen": 8194952,
      "step": 14125
    },
    {
      "epoch": 2.10455764075067,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.029997509552777094,
      "loss": 0.8283,
      "num_input_tokens_seen": 8197736,
      "step": 14130
    },
    {
      "epoch": 2.1053023532916293,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029997473900011958,
      "loss": 0.8176,
      "num_input_tokens_seen": 8200584,
      "step": 14135
    },
    {
      "epoch": 2.1060470658325885,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02999743799387974,
      "loss": 0.8226,
      "num_input_tokens_seen": 8203560,
      "step": 14140
    },
    {
      "epoch": 2.1067917783735477,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029997401834381054,
      "loss": 0.8106,
      "num_input_tokens_seen": 8206344,
      "step": 14145
    },
    {
      "epoch": 2.107536490914507,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029997365421516504,
      "loss": 0.7908,
      "num_input_tokens_seen": 8209128,
      "step": 14150
    },
    {
      "epoch": 2.108281203455466,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02999732875528671,
      "loss": 0.8042,
      "num_input_tokens_seen": 8211976,
      "step": 14155
    },
    {
      "epoch": 2.1090259159964253,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.029997291835692285,
      "loss": 0.8012,
      "num_input_tokens_seen": 8214920,
      "step": 14160
    },
    {
      "epoch": 2.1097706285373845,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02999725466273386,
      "loss": 0.7968,
      "num_input_tokens_seen": 8217992,
      "step": 14165
    },
    {
      "epoch": 2.1105153410783437,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.029997217236412063,
      "loss": 0.7859,
      "num_input_tokens_seen": 8221000,
      "step": 14170
    },
    {
      "epoch": 2.111260053619303,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02999717955672752,
      "loss": 0.7779,
      "num_input_tokens_seen": 8224008,
      "step": 14175
    },
    {
      "epoch": 2.112004766160262,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.029997141623680872,
      "loss": 0.7939,
      "num_input_tokens_seen": 8226760,
      "step": 14180
    },
    {
      "epoch": 2.1127494787012213,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02999710343727276,
      "loss": 0.8169,
      "num_input_tokens_seen": 8229768,
      "step": 14185
    },
    {
      "epoch": 2.1134941912421805,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.029997064997503826,
      "loss": 0.8045,
      "num_input_tokens_seen": 8232744,
      "step": 14190
    },
    {
      "epoch": 2.1142389037831397,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.029997026304374716,
      "loss": 0.8103,
      "num_input_tokens_seen": 8235880,
      "step": 14195
    },
    {
      "epoch": 2.114983616324099,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.0299969873578861,
      "loss": 0.8067,
      "num_input_tokens_seen": 8238920,
      "step": 14200
    },
    {
      "epoch": 2.115728328865058,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029996948158038616,
      "loss": 0.8039,
      "num_input_tokens_seen": 8241896,
      "step": 14205
    },
    {
      "epoch": 2.1164730414060173,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.029996908704832938,
      "loss": 0.8155,
      "num_input_tokens_seen": 8244744,
      "step": 14210
    },
    {
      "epoch": 2.1172177539469765,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.02999686899826973,
      "loss": 0.7796,
      "num_input_tokens_seen": 8247752,
      "step": 14215
    },
    {
      "epoch": 2.1179624664879357,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02999682903834966,
      "loss": 0.8232,
      "num_input_tokens_seen": 8250536,
      "step": 14220
    },
    {
      "epoch": 2.118707179028895,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02999678882507341,
      "loss": 0.81,
      "num_input_tokens_seen": 8253480,
      "step": 14225
    },
    {
      "epoch": 2.119451891569854,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029996748358441648,
      "loss": 0.7979,
      "num_input_tokens_seen": 8256328,
      "step": 14230
    },
    {
      "epoch": 2.1201966041108133,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029996707638455068,
      "loss": 0.7942,
      "num_input_tokens_seen": 8259176,
      "step": 14235
    },
    {
      "epoch": 2.1209413166517725,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.029996666665114358,
      "loss": 0.8091,
      "num_input_tokens_seen": 8261992,
      "step": 14240
    },
    {
      "epoch": 2.1216860291927317,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029996625438420202,
      "loss": 0.7916,
      "num_input_tokens_seen": 8265032,
      "step": 14245
    },
    {
      "epoch": 2.122430741733691,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029996583958373305,
      "loss": 0.819,
      "num_input_tokens_seen": 8268168,
      "step": 14250
    },
    {
      "epoch": 2.12317545427465,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02999654222497436,
      "loss": 0.8019,
      "num_input_tokens_seen": 8271208,
      "step": 14255
    },
    {
      "epoch": 2.1239201668156094,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029996500238224075,
      "loss": 0.8086,
      "num_input_tokens_seen": 8274184,
      "step": 14260
    },
    {
      "epoch": 2.1246648793565686,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.029996457998123165,
      "loss": 0.8015,
      "num_input_tokens_seen": 8277064,
      "step": 14265
    },
    {
      "epoch": 2.1254095918975278,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029996415504672337,
      "loss": 0.8082,
      "num_input_tokens_seen": 8280200,
      "step": 14270
    },
    {
      "epoch": 2.1261543044384865,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029996372757872307,
      "loss": 0.7967,
      "num_input_tokens_seen": 8282888,
      "step": 14275
    },
    {
      "epoch": 2.126899016979446,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029996329757723806,
      "loss": 0.8034,
      "num_input_tokens_seen": 8285704,
      "step": 14280
    },
    {
      "epoch": 2.127643729520405,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029996286504227556,
      "loss": 0.8002,
      "num_input_tokens_seen": 8288520,
      "step": 14285
    },
    {
      "epoch": 2.128388442061364,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.029996242997384282,
      "loss": 0.8086,
      "num_input_tokens_seen": 8291304,
      "step": 14290
    },
    {
      "epoch": 2.1291331546023233,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029996199237194727,
      "loss": 0.8054,
      "num_input_tokens_seen": 8294120,
      "step": 14295
    },
    {
      "epoch": 2.1298778671432825,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02999615522365963,
      "loss": 0.8054,
      "num_input_tokens_seen": 8297160,
      "step": 14300
    },
    {
      "epoch": 2.1306225796842417,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02999611095677973,
      "loss": 0.8041,
      "num_input_tokens_seen": 8299944,
      "step": 14305
    },
    {
      "epoch": 2.131367292225201,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.029996066436555776,
      "loss": 0.8047,
      "num_input_tokens_seen": 8302920,
      "step": 14310
    },
    {
      "epoch": 2.13211200476616,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02999602166298852,
      "loss": 0.8015,
      "num_input_tokens_seen": 8306024,
      "step": 14315
    },
    {
      "epoch": 2.1328567173071193,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.029995976636078718,
      "loss": 0.8144,
      "num_input_tokens_seen": 8308840,
      "step": 14320
    },
    {
      "epoch": 2.1336014298480785,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.029995931355827135,
      "loss": 0.7978,
      "num_input_tokens_seen": 8312008,
      "step": 14325
    },
    {
      "epoch": 2.1343461423890377,
      "grad_norm": 0.01123046875,
      "learning_rate": 0.02999588582223453,
      "loss": 0.7956,
      "num_input_tokens_seen": 8314856,
      "step": 14330
    },
    {
      "epoch": 2.135090854929997,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029995840035301678,
      "loss": 0.8089,
      "num_input_tokens_seen": 8317608,
      "step": 14335
    },
    {
      "epoch": 2.135835567470956,
      "grad_norm": 0.01544189453125,
      "learning_rate": 0.02999579399502935,
      "loss": 0.8102,
      "num_input_tokens_seen": 8320520,
      "step": 14340
    },
    {
      "epoch": 2.1365802800119154,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.02999574770141832,
      "loss": 0.8273,
      "num_input_tokens_seen": 8323272,
      "step": 14345
    },
    {
      "epoch": 2.1373249925528746,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.029995701154469378,
      "loss": 0.8137,
      "num_input_tokens_seen": 8326184,
      "step": 14350
    },
    {
      "epoch": 2.1380697050938338,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029995654354183304,
      "loss": 0.8006,
      "num_input_tokens_seen": 8329192,
      "step": 14355
    },
    {
      "epoch": 2.138814417634793,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.02999560730056089,
      "loss": 0.8051,
      "num_input_tokens_seen": 8332392,
      "step": 14360
    },
    {
      "epoch": 2.139559130175752,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029995559993602927,
      "loss": 0.8021,
      "num_input_tokens_seen": 8335432,
      "step": 14365
    },
    {
      "epoch": 2.1403038427167114,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029995512433310226,
      "loss": 0.814,
      "num_input_tokens_seen": 8338216,
      "step": 14370
    },
    {
      "epoch": 2.1410485552576706,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029995464619683578,
      "loss": 0.8026,
      "num_input_tokens_seen": 8341224,
      "step": 14375
    },
    {
      "epoch": 2.1417932677986298,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.0299954165527238,
      "loss": 0.8044,
      "num_input_tokens_seen": 8344104,
      "step": 14380
    },
    {
      "epoch": 2.142537980339589,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.0299953682324317,
      "loss": 0.8069,
      "num_input_tokens_seen": 8346792,
      "step": 14385
    },
    {
      "epoch": 2.143282692880548,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.029995319658808098,
      "loss": 0.7966,
      "num_input_tokens_seen": 8349864,
      "step": 14390
    },
    {
      "epoch": 2.1440274054215074,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029995270831853802,
      "loss": 0.8168,
      "num_input_tokens_seen": 8352904,
      "step": 14395
    },
    {
      "epoch": 2.1447721179624666,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02999522175156965,
      "loss": 0.813,
      "num_input_tokens_seen": 8355848,
      "step": 14400
    },
    {
      "epoch": 2.145516830503426,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.029995172417956465,
      "loss": 0.7967,
      "num_input_tokens_seen": 8358792,
      "step": 14405
    },
    {
      "epoch": 2.146261543044385,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029995122831015086,
      "loss": 0.7977,
      "num_input_tokens_seen": 8361224,
      "step": 14410
    },
    {
      "epoch": 2.147006255585344,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.029995072990746347,
      "loss": 0.8086,
      "num_input_tokens_seen": 8364360,
      "step": 14415
    },
    {
      "epoch": 2.1477509681263034,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02999502289715109,
      "loss": 0.7939,
      "num_input_tokens_seen": 8367016,
      "step": 14420
    },
    {
      "epoch": 2.1484956806672626,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.029994972550230155,
      "loss": 0.8085,
      "num_input_tokens_seen": 8369928,
      "step": 14425
    },
    {
      "epoch": 2.149240393208222,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.0299949219499844,
      "loss": 0.8051,
      "num_input_tokens_seen": 8372712,
      "step": 14430
    },
    {
      "epoch": 2.149985105749181,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.029994871096414688,
      "loss": 0.8002,
      "num_input_tokens_seen": 8375560,
      "step": 14435
    },
    {
      "epoch": 2.15072981829014,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029994819989521863,
      "loss": 0.8003,
      "num_input_tokens_seen": 8378536,
      "step": 14440
    },
    {
      "epoch": 2.1514745308310994,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029994768629306796,
      "loss": 0.7941,
      "num_input_tokens_seen": 8381288,
      "step": 14445
    },
    {
      "epoch": 2.152219243372058,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02999471701577035,
      "loss": 0.8137,
      "num_input_tokens_seen": 8384200,
      "step": 14450
    },
    {
      "epoch": 2.1529639559130174,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.0299946651489134,
      "loss": 0.8017,
      "num_input_tokens_seen": 8387208,
      "step": 14455
    },
    {
      "epoch": 2.1537086684539766,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.029994613028736825,
      "loss": 0.8036,
      "num_input_tokens_seen": 8390280,
      "step": 14460
    },
    {
      "epoch": 2.1544533809949358,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.029994560655241503,
      "loss": 0.8059,
      "num_input_tokens_seen": 8393128,
      "step": 14465
    },
    {
      "epoch": 2.155198093535895,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02999450802842832,
      "loss": 0.8132,
      "num_input_tokens_seen": 8396136,
      "step": 14470
    },
    {
      "epoch": 2.155942806076854,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.029994455148298158,
      "loss": 0.7956,
      "num_input_tokens_seen": 8399208,
      "step": 14475
    },
    {
      "epoch": 2.1566875186178134,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02999440201485192,
      "loss": 0.7949,
      "num_input_tokens_seen": 8402152,
      "step": 14480
    },
    {
      "epoch": 2.1574322311587726,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.029994348628090497,
      "loss": 0.7961,
      "num_input_tokens_seen": 8405064,
      "step": 14485
    },
    {
      "epoch": 2.158176943699732,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029994294988014796,
      "loss": 0.7812,
      "num_input_tokens_seen": 8408808,
      "step": 14490
    },
    {
      "epoch": 2.158921656240691,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029994241094625715,
      "loss": 0.8055,
      "num_input_tokens_seen": 8411880,
      "step": 14495
    },
    {
      "epoch": 2.15966636878165,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02999418694792418,
      "loss": 0.8014,
      "num_input_tokens_seen": 8414568,
      "step": 14500
    },
    {
      "epoch": 2.1604110813226094,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029994132547911088,
      "loss": 0.7824,
      "num_input_tokens_seen": 8417192,
      "step": 14505
    },
    {
      "epoch": 2.1611557938635686,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02999407789458737,
      "loss": 0.7915,
      "num_input_tokens_seen": 8420200,
      "step": 14510
    },
    {
      "epoch": 2.161900506404528,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029994022987953943,
      "loss": 0.78,
      "num_input_tokens_seen": 8422952,
      "step": 14515
    },
    {
      "epoch": 2.162645218945487,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02999396782801174,
      "loss": 0.8278,
      "num_input_tokens_seen": 8425544,
      "step": 14520
    },
    {
      "epoch": 2.163389931486446,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029993912414761684,
      "loss": 0.8038,
      "num_input_tokens_seen": 8428616,
      "step": 14525
    },
    {
      "epoch": 2.1641346440274054,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.02999385674820472,
      "loss": 0.8085,
      "num_input_tokens_seen": 8431592,
      "step": 14530
    },
    {
      "epoch": 2.1648793565683646,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.029993800828341785,
      "loss": 0.7958,
      "num_input_tokens_seen": 8434568,
      "step": 14535
    },
    {
      "epoch": 2.165624069109324,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02999374465517383,
      "loss": 0.7964,
      "num_input_tokens_seen": 8437288,
      "step": 14540
    },
    {
      "epoch": 2.166368781650283,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029993688228701788,
      "loss": 0.82,
      "num_input_tokens_seen": 8439976,
      "step": 14545
    },
    {
      "epoch": 2.167113494191242,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02999363154892663,
      "loss": 0.822,
      "num_input_tokens_seen": 8442728,
      "step": 14550
    },
    {
      "epoch": 2.1678582067322014,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.0299935746158493,
      "loss": 0.8073,
      "num_input_tokens_seen": 8445608,
      "step": 14555
    },
    {
      "epoch": 2.1686029192731606,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02999351742947077,
      "loss": 0.7936,
      "num_input_tokens_seen": 8448360,
      "step": 14560
    },
    {
      "epoch": 2.16934763181412,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029993459989792002,
      "loss": 0.8128,
      "num_input_tokens_seen": 8451368,
      "step": 14565
    },
    {
      "epoch": 2.170092344355079,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02999340229681396,
      "loss": 0.8041,
      "num_input_tokens_seen": 8453960,
      "step": 14570
    },
    {
      "epoch": 2.1708370568960382,
      "grad_norm": 0.01544189453125,
      "learning_rate": 0.029993344350537632,
      "loss": 0.7979,
      "num_input_tokens_seen": 8456872,
      "step": 14575
    },
    {
      "epoch": 2.1715817694369974,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029993286150963985,
      "loss": 0.8056,
      "num_input_tokens_seen": 8459656,
      "step": 14580
    },
    {
      "epoch": 2.1723264819779566,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029993227698094014,
      "loss": 0.7876,
      "num_input_tokens_seen": 8462536,
      "step": 14585
    },
    {
      "epoch": 2.173071194518916,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02999316899192869,
      "loss": 0.7832,
      "num_input_tokens_seen": 8465544,
      "step": 14590
    },
    {
      "epoch": 2.173815907059875,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029993110032469024,
      "loss": 0.792,
      "num_input_tokens_seen": 8468424,
      "step": 14595
    },
    {
      "epoch": 2.1745606196008342,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.029993050819715997,
      "loss": 0.8075,
      "num_input_tokens_seen": 8471496,
      "step": 14600
    },
    {
      "epoch": 2.1753053321417934,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029992991353670614,
      "loss": 0.8226,
      "num_input_tokens_seen": 8474728,
      "step": 14605
    },
    {
      "epoch": 2.1760500446827526,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029992931634333878,
      "loss": 0.8403,
      "num_input_tokens_seen": 8477320,
      "step": 14610
    },
    {
      "epoch": 2.176794757223712,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029992871661706805,
      "loss": 0.7986,
      "num_input_tokens_seen": 8480136,
      "step": 14615
    },
    {
      "epoch": 2.177539469764671,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029992811435790398,
      "loss": 0.8088,
      "num_input_tokens_seen": 8482952,
      "step": 14620
    },
    {
      "epoch": 2.17828418230563,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029992750956585684,
      "loss": 0.8334,
      "num_input_tokens_seen": 8485736,
      "step": 14625
    },
    {
      "epoch": 2.179028894846589,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.02999269022409368,
      "loss": 0.8134,
      "num_input_tokens_seen": 8488360,
      "step": 14630
    },
    {
      "epoch": 2.179773607387548,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.029992629238315412,
      "loss": 0.7946,
      "num_input_tokens_seen": 8491368,
      "step": 14635
    },
    {
      "epoch": 2.1805183199285074,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02999256799925191,
      "loss": 0.7886,
      "num_input_tokens_seen": 8494120,
      "step": 14640
    },
    {
      "epoch": 2.1812630324694666,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02999250650690421,
      "loss": 0.7989,
      "num_input_tokens_seen": 8497000,
      "step": 14645
    },
    {
      "epoch": 2.182007745010426,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.029992444761273353,
      "loss": 0.7921,
      "num_input_tokens_seen": 8499784,
      "step": 14650
    },
    {
      "epoch": 2.182752457551385,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02999238276236038,
      "loss": 0.7884,
      "num_input_tokens_seen": 8502632,
      "step": 14655
    },
    {
      "epoch": 2.1834971700923442,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029992320510166332,
      "loss": 0.7844,
      "num_input_tokens_seen": 8505352,
      "step": 14660
    },
    {
      "epoch": 2.1842418826333034,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02999225800469227,
      "loss": 0.8152,
      "num_input_tokens_seen": 8508264,
      "step": 14665
    },
    {
      "epoch": 2.1849865951742626,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02999219524593925,
      "loss": 0.7646,
      "num_input_tokens_seen": 8510952,
      "step": 14670
    },
    {
      "epoch": 2.185731307715222,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029992132233908324,
      "loss": 0.7855,
      "num_input_tokens_seen": 8514312,
      "step": 14675
    },
    {
      "epoch": 2.186476020256181,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.02999206896860056,
      "loss": 0.8245,
      "num_input_tokens_seen": 8517224,
      "step": 14680
    },
    {
      "epoch": 2.1872207327971402,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02999200545001703,
      "loss": 0.8052,
      "num_input_tokens_seen": 8520008,
      "step": 14685
    },
    {
      "epoch": 2.1879654453380994,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.029991941678158805,
      "loss": 0.8445,
      "num_input_tokens_seen": 8522952,
      "step": 14690
    },
    {
      "epoch": 2.1887101578790586,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02999187765302696,
      "loss": 0.7854,
      "num_input_tokens_seen": 8526088,
      "step": 14695
    },
    {
      "epoch": 2.189454870420018,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029991813374622586,
      "loss": 0.8308,
      "num_input_tokens_seen": 8529128,
      "step": 14700
    },
    {
      "epoch": 2.190199582960977,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02999174884294676,
      "loss": 0.7855,
      "num_input_tokens_seen": 8531880,
      "step": 14705
    },
    {
      "epoch": 2.1909442955019363,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02999168405800057,
      "loss": 0.7891,
      "num_input_tokens_seen": 8534888,
      "step": 14710
    },
    {
      "epoch": 2.1916890080428955,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.02999161901978512,
      "loss": 0.7993,
      "num_input_tokens_seen": 8537992,
      "step": 14715
    },
    {
      "epoch": 2.1924337205838547,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.029991553728301496,
      "loss": 0.8222,
      "num_input_tokens_seen": 8541032,
      "step": 14720
    },
    {
      "epoch": 2.193178433124814,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.02999148818355081,
      "loss": 0.7913,
      "num_input_tokens_seen": 8543944,
      "step": 14725
    },
    {
      "epoch": 2.193923145665773,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.02999142238553417,
      "loss": 0.8071,
      "num_input_tokens_seen": 8546824,
      "step": 14730
    },
    {
      "epoch": 2.1946678582067323,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02999135633425269,
      "loss": 0.8433,
      "num_input_tokens_seen": 8549704,
      "step": 14735
    },
    {
      "epoch": 2.1954125707476915,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029991290029707476,
      "loss": 0.7842,
      "num_input_tokens_seen": 8552488,
      "step": 14740
    },
    {
      "epoch": 2.1961572832886507,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029991223471899652,
      "loss": 0.8182,
      "num_input_tokens_seen": 8555560,
      "step": 14745
    },
    {
      "epoch": 2.19690199582961,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.029991156660830347,
      "loss": 0.8071,
      "num_input_tokens_seen": 8558760,
      "step": 14750
    },
    {
      "epoch": 2.197646708370569,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.029991089596500684,
      "loss": 0.7868,
      "num_input_tokens_seen": 8561544,
      "step": 14755
    },
    {
      "epoch": 2.1983914209115283,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029991022278911803,
      "loss": 0.7933,
      "num_input_tokens_seen": 8564456,
      "step": 14760
    },
    {
      "epoch": 2.1991361334524875,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029990954708064833,
      "loss": 0.8026,
      "num_input_tokens_seen": 8567016,
      "step": 14765
    },
    {
      "epoch": 2.1998808459934467,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029990886883960923,
      "loss": 0.8145,
      "num_input_tokens_seen": 8569736,
      "step": 14770
    },
    {
      "epoch": 2.200625558534406,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.029990818806601212,
      "loss": 0.8002,
      "num_input_tokens_seen": 8572680,
      "step": 14775
    },
    {
      "epoch": 2.201370271075365,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.029990750475986854,
      "loss": 0.8248,
      "num_input_tokens_seen": 8575656,
      "step": 14780
    },
    {
      "epoch": 2.2021149836163243,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.029990681892119002,
      "loss": 0.8048,
      "num_input_tokens_seen": 8578632,
      "step": 14785
    },
    {
      "epoch": 2.202859696157283,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02999061305499882,
      "loss": 0.815,
      "num_input_tokens_seen": 8581512,
      "step": 14790
    },
    {
      "epoch": 2.2036044086982427,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02999054396462746,
      "loss": 0.8262,
      "num_input_tokens_seen": 8584584,
      "step": 14795
    },
    {
      "epoch": 2.2043491212392015,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029990474621006097,
      "loss": 0.8018,
      "num_input_tokens_seen": 8587688,
      "step": 14800
    },
    {
      "epoch": 2.2050938337801607,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029990405024135903,
      "loss": 0.8002,
      "num_input_tokens_seen": 8590664,
      "step": 14805
    },
    {
      "epoch": 2.20583854632112,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.029990335174018046,
      "loss": 0.8064,
      "num_input_tokens_seen": 8593480,
      "step": 14810
    },
    {
      "epoch": 2.206583258862079,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02999026507065372,
      "loss": 0.8034,
      "num_input_tokens_seen": 8596168,
      "step": 14815
    },
    {
      "epoch": 2.2073279714030383,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.029990194714044096,
      "loss": 0.7989,
      "num_input_tokens_seen": 8599048,
      "step": 14820
    },
    {
      "epoch": 2.2080726839439975,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029990124104190368,
      "loss": 0.8045,
      "num_input_tokens_seen": 8601864,
      "step": 14825
    },
    {
      "epoch": 2.2088173964849567,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029990053241093727,
      "loss": 0.8079,
      "num_input_tokens_seen": 8604776,
      "step": 14830
    },
    {
      "epoch": 2.209562109025916,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029989982124755375,
      "loss": 0.808,
      "num_input_tokens_seen": 8607816,
      "step": 14835
    },
    {
      "epoch": 2.210306821566875,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02998991075517651,
      "loss": 0.8121,
      "num_input_tokens_seen": 8610536,
      "step": 14840
    },
    {
      "epoch": 2.2110515341078343,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029989839132358337,
      "loss": 0.7963,
      "num_input_tokens_seen": 8613480,
      "step": 14845
    },
    {
      "epoch": 2.2117962466487935,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029989767256302067,
      "loss": 0.804,
      "num_input_tokens_seen": 8616296,
      "step": 14850
    },
    {
      "epoch": 2.2125409591897527,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.029989695127008914,
      "loss": 0.8021,
      "num_input_tokens_seen": 8619464,
      "step": 14855
    },
    {
      "epoch": 2.213285671730712,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.029989622744480096,
      "loss": 0.8188,
      "num_input_tokens_seen": 8622312,
      "step": 14860
    },
    {
      "epoch": 2.214030384271671,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029989550108716834,
      "loss": 0.8116,
      "num_input_tokens_seen": 8625000,
      "step": 14865
    },
    {
      "epoch": 2.2147750968126303,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02998947721972036,
      "loss": 0.8235,
      "num_input_tokens_seen": 8628040,
      "step": 14870
    },
    {
      "epoch": 2.2155198093535895,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029989404077491906,
      "loss": 0.8082,
      "num_input_tokens_seen": 8630792,
      "step": 14875
    },
    {
      "epoch": 2.2162645218945487,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.029989330682032703,
      "loss": 0.8029,
      "num_input_tokens_seen": 8633736,
      "step": 14880
    },
    {
      "epoch": 2.217009234435508,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.029989257033343993,
      "loss": 0.809,
      "num_input_tokens_seen": 8636840,
      "step": 14885
    },
    {
      "epoch": 2.217753946976467,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.029989183131427018,
      "loss": 0.8032,
      "num_input_tokens_seen": 8639816,
      "step": 14890
    },
    {
      "epoch": 2.2184986595174263,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.029989108976283027,
      "loss": 0.8014,
      "num_input_tokens_seen": 8642536,
      "step": 14895
    },
    {
      "epoch": 2.2192433720583855,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02998903456791328,
      "loss": 0.8014,
      "num_input_tokens_seen": 8645576,
      "step": 14900
    },
    {
      "epoch": 2.2199880845993447,
      "grad_norm": 0.02734375,
      "learning_rate": 0.029988959906319022,
      "loss": 0.8062,
      "num_input_tokens_seen": 8648712,
      "step": 14905
    },
    {
      "epoch": 2.220732797140304,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.029988884991501524,
      "loss": 0.7973,
      "num_input_tokens_seen": 8651976,
      "step": 14910
    },
    {
      "epoch": 2.221477509681263,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.029988809823462043,
      "loss": 0.8041,
      "num_input_tokens_seen": 8654952,
      "step": 14915
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029988734402201858,
      "loss": 0.8047,
      "num_input_tokens_seen": 8658056,
      "step": 14920
    },
    {
      "epoch": 2.2229669347631815,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02998865872772224,
      "loss": 0.8123,
      "num_input_tokens_seen": 8661160,
      "step": 14925
    },
    {
      "epoch": 2.2237116473041407,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029988582800024467,
      "loss": 0.806,
      "num_input_tokens_seen": 8664040,
      "step": 14930
    },
    {
      "epoch": 2.2244563598451,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.029988506619109817,
      "loss": 0.7965,
      "num_input_tokens_seen": 8667016,
      "step": 14935
    },
    {
      "epoch": 2.225201072386059,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029988430184979584,
      "loss": 0.8079,
      "num_input_tokens_seen": 8669928,
      "step": 14940
    },
    {
      "epoch": 2.2259457849270183,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029988353497635057,
      "loss": 0.8034,
      "num_input_tokens_seen": 8673032,
      "step": 14945
    },
    {
      "epoch": 2.2266904974679775,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02998827655707753,
      "loss": 0.8037,
      "num_input_tokens_seen": 8675976,
      "step": 14950
    },
    {
      "epoch": 2.2274352100089367,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029988199363308307,
      "loss": 0.7947,
      "num_input_tokens_seen": 8678856,
      "step": 14955
    },
    {
      "epoch": 2.228179922549896,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029988121916328686,
      "loss": 0.8147,
      "num_input_tokens_seen": 8681896,
      "step": 14960
    },
    {
      "epoch": 2.2289246350908547,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.029988044216139983,
      "loss": 0.7926,
      "num_input_tokens_seen": 8684968,
      "step": 14965
    },
    {
      "epoch": 2.2296693476318143,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.0299879662627435,
      "loss": 0.8117,
      "num_input_tokens_seen": 8687912,
      "step": 14970
    },
    {
      "epoch": 2.230414060172773,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.029987888056140565,
      "loss": 0.793,
      "num_input_tokens_seen": 8691176,
      "step": 14975
    },
    {
      "epoch": 2.2311587727137323,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.02998780959633249,
      "loss": 0.8088,
      "num_input_tokens_seen": 8694248,
      "step": 14980
    },
    {
      "epoch": 2.2319034852546915,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.029987730883320605,
      "loss": 0.7969,
      "num_input_tokens_seen": 8697096,
      "step": 14985
    },
    {
      "epoch": 2.2326481977956507,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029987651917106246,
      "loss": 0.8031,
      "num_input_tokens_seen": 8699880,
      "step": 14990
    },
    {
      "epoch": 2.23339291033661,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029987572697690732,
      "loss": 0.8099,
      "num_input_tokens_seen": 8702728,
      "step": 14995
    },
    {
      "epoch": 2.234137622877569,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.029987493225075417,
      "loss": 0.7809,
      "num_input_tokens_seen": 8705608,
      "step": 15000
    },
    {
      "epoch": 2.2348823354185283,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029987413499261633,
      "loss": 0.7979,
      "num_input_tokens_seen": 8708296,
      "step": 15005
    },
    {
      "epoch": 2.2356270479594875,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02998733352025073,
      "loss": 0.7893,
      "num_input_tokens_seen": 8711080,
      "step": 15010
    },
    {
      "epoch": 2.2363717605004467,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02998725328804406,
      "loss": 0.7744,
      "num_input_tokens_seen": 8713896,
      "step": 15015
    },
    {
      "epoch": 2.237116473041406,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02998717280264298,
      "loss": 0.8147,
      "num_input_tokens_seen": 8717000,
      "step": 15020
    },
    {
      "epoch": 2.237861185582365,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029987092064048846,
      "loss": 0.8107,
      "num_input_tokens_seen": 8720264,
      "step": 15025
    },
    {
      "epoch": 2.2386058981233243,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029987011072263023,
      "loss": 0.8262,
      "num_input_tokens_seen": 8723272,
      "step": 15030
    },
    {
      "epoch": 2.2393506106642835,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02998692982728688,
      "loss": 0.7972,
      "num_input_tokens_seen": 8725960,
      "step": 15035
    },
    {
      "epoch": 2.2400953232052427,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02998684832912179,
      "loss": 0.8108,
      "num_input_tokens_seen": 8729064,
      "step": 15040
    },
    {
      "epoch": 2.240840035746202,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029986766577769133,
      "loss": 0.7994,
      "num_input_tokens_seen": 8732040,
      "step": 15045
    },
    {
      "epoch": 2.241584748287161,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.029986684573230282,
      "loss": 0.8237,
      "num_input_tokens_seen": 8735080,
      "step": 15050
    },
    {
      "epoch": 2.2423294608281203,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029986602315506632,
      "loss": 0.8013,
      "num_input_tokens_seen": 8737800,
      "step": 15055
    },
    {
      "epoch": 2.2430741733690795,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029986519804599564,
      "loss": 0.7889,
      "num_input_tokens_seen": 8740648,
      "step": 15060
    },
    {
      "epoch": 2.2438188859100388,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.029986437040510475,
      "loss": 0.8082,
      "num_input_tokens_seen": 8743560,
      "step": 15065
    },
    {
      "epoch": 2.244563598450998,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029986354023240763,
      "loss": 0.8038,
      "num_input_tokens_seen": 8746824,
      "step": 15070
    },
    {
      "epoch": 2.245308310991957,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029986270752791835,
      "loss": 0.8213,
      "num_input_tokens_seen": 8749672,
      "step": 15075
    },
    {
      "epoch": 2.2460530235329164,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029986187229165087,
      "loss": 0.8019,
      "num_input_tokens_seen": 8752392,
      "step": 15080
    },
    {
      "epoch": 2.2467977360738756,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029986103452361943,
      "loss": 0.784,
      "num_input_tokens_seen": 8755080,
      "step": 15085
    },
    {
      "epoch": 2.2475424486148348,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02998601942238381,
      "loss": 0.8162,
      "num_input_tokens_seen": 8758152,
      "step": 15090
    },
    {
      "epoch": 2.248287161155794,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029985935139232114,
      "loss": 0.7955,
      "num_input_tokens_seen": 8761160,
      "step": 15095
    },
    {
      "epoch": 2.249031873696753,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02998585060290827,
      "loss": 0.7869,
      "num_input_tokens_seen": 8764104,
      "step": 15100
    },
    {
      "epoch": 2.2497765862377124,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.029985765813413712,
      "loss": 0.8175,
      "num_input_tokens_seen": 8766952,
      "step": 15105
    },
    {
      "epoch": 2.2505212987786716,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029985680770749874,
      "loss": 0.8162,
      "num_input_tokens_seen": 8769800,
      "step": 15110
    },
    {
      "epoch": 2.2512660113196308,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02998559547491819,
      "loss": 0.7814,
      "num_input_tokens_seen": 8773224,
      "step": 15115
    },
    {
      "epoch": 2.25201072386059,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029985509925920098,
      "loss": 0.813,
      "num_input_tokens_seen": 8775848,
      "step": 15120
    },
    {
      "epoch": 2.252755436401549,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02998542412375705,
      "loss": 0.8005,
      "num_input_tokens_seen": 8778952,
      "step": 15125
    },
    {
      "epoch": 2.2535001489425084,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02998533806843049,
      "loss": 0.8161,
      "num_input_tokens_seen": 8782280,
      "step": 15130
    },
    {
      "epoch": 2.2542448614834676,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029985251759941874,
      "loss": 0.7925,
      "num_input_tokens_seen": 8785416,
      "step": 15135
    },
    {
      "epoch": 2.2549895740244263,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029985165198292658,
      "loss": 0.7955,
      "num_input_tokens_seen": 8788328,
      "step": 15140
    },
    {
      "epoch": 2.255734286565386,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029985078383484312,
      "loss": 0.8046,
      "num_input_tokens_seen": 8791240,
      "step": 15145
    },
    {
      "epoch": 2.2564789991063448,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029984991315518288,
      "loss": 0.8018,
      "num_input_tokens_seen": 8794344,
      "step": 15150
    },
    {
      "epoch": 2.257223711647304,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.029984903994396073,
      "loss": 0.8017,
      "num_input_tokens_seen": 8797544,
      "step": 15155
    },
    {
      "epoch": 2.257968424188263,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02998481642011913,
      "loss": 0.8123,
      "num_input_tokens_seen": 8800680,
      "step": 15160
    },
    {
      "epoch": 2.2587131367292224,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02998472859268895,
      "loss": 0.793,
      "num_input_tokens_seen": 8803592,
      "step": 15165
    },
    {
      "epoch": 2.2594578492701816,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.029984640512107003,
      "loss": 0.7852,
      "num_input_tokens_seen": 8806472,
      "step": 15170
    },
    {
      "epoch": 2.2602025618111408,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02998455217837479,
      "loss": 0.8221,
      "num_input_tokens_seen": 8809224,
      "step": 15175
    },
    {
      "epoch": 2.2609472743521,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.029984463591493794,
      "loss": 0.7951,
      "num_input_tokens_seen": 8811976,
      "step": 15180
    },
    {
      "epoch": 2.261691986893059,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029984374751465512,
      "loss": 0.7923,
      "num_input_tokens_seen": 8814824,
      "step": 15185
    },
    {
      "epoch": 2.2624366994340184,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.029984285658291455,
      "loss": 0.8072,
      "num_input_tokens_seen": 8817832,
      "step": 15190
    },
    {
      "epoch": 2.2631814119749776,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02998419631197312,
      "loss": 0.815,
      "num_input_tokens_seen": 8820968,
      "step": 15195
    },
    {
      "epoch": 2.2639261245159368,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.029984106712512015,
      "loss": 0.8062,
      "num_input_tokens_seen": 8823656,
      "step": 15200
    },
    {
      "epoch": 2.264670837056896,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029984016859909656,
      "loss": 0.8009,
      "num_input_tokens_seen": 8826408,
      "step": 15205
    },
    {
      "epoch": 2.265415549597855,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029983926754167563,
      "loss": 0.8023,
      "num_input_tokens_seen": 8829224,
      "step": 15210
    },
    {
      "epoch": 2.2661602621388144,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.029983836395287258,
      "loss": 0.8084,
      "num_input_tokens_seen": 8831976,
      "step": 15215
    },
    {
      "epoch": 2.2669049746797736,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02998374578327026,
      "loss": 0.8124,
      "num_input_tokens_seen": 8835048,
      "step": 15220
    },
    {
      "epoch": 2.267649687220733,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02998365491811811,
      "loss": 0.8114,
      "num_input_tokens_seen": 8838024,
      "step": 15225
    },
    {
      "epoch": 2.268394399761692,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02998356379983234,
      "loss": 0.8131,
      "num_input_tokens_seen": 8841032,
      "step": 15230
    },
    {
      "epoch": 2.269139112302651,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029983472428414485,
      "loss": 0.8178,
      "num_input_tokens_seen": 8844008,
      "step": 15235
    },
    {
      "epoch": 2.2698838248436104,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029983380803866096,
      "loss": 0.8031,
      "num_input_tokens_seen": 8846856,
      "step": 15240
    },
    {
      "epoch": 2.2706285373845696,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029983288926188716,
      "loss": 0.7956,
      "num_input_tokens_seen": 8849512,
      "step": 15245
    },
    {
      "epoch": 2.271373249925529,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029983196795383894,
      "loss": 0.7844,
      "num_input_tokens_seen": 8852552,
      "step": 15250
    },
    {
      "epoch": 2.272117962466488,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.02998310441145319,
      "loss": 0.808,
      "num_input_tokens_seen": 8855720,
      "step": 15255
    },
    {
      "epoch": 2.272862675007447,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.02998301177439817,
      "loss": 0.7908,
      "num_input_tokens_seen": 8858536,
      "step": 15260
    },
    {
      "epoch": 2.2736073875484064,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029982918884220392,
      "loss": 0.7984,
      "num_input_tokens_seen": 8861352,
      "step": 15265
    },
    {
      "epoch": 2.2743521000893656,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.029982825740921426,
      "loss": 0.8309,
      "num_input_tokens_seen": 8864392,
      "step": 15270
    },
    {
      "epoch": 2.275096812630325,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.029982732344502848,
      "loss": 0.8007,
      "num_input_tokens_seen": 8867368,
      "step": 15275
    },
    {
      "epoch": 2.275841525171284,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029982638694966236,
      "loss": 0.7891,
      "num_input_tokens_seen": 8870184,
      "step": 15280
    },
    {
      "epoch": 2.276586237712243,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02998254479231317,
      "loss": 0.8041,
      "num_input_tokens_seen": 8873160,
      "step": 15285
    },
    {
      "epoch": 2.2773309502532024,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029982450636545237,
      "loss": 0.8124,
      "num_input_tokens_seen": 8875880,
      "step": 15290
    },
    {
      "epoch": 2.2780756627941616,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.029982356227664028,
      "loss": 0.7997,
      "num_input_tokens_seen": 8878952,
      "step": 15295
    },
    {
      "epoch": 2.278820375335121,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.029982261565671138,
      "loss": 0.8738,
      "num_input_tokens_seen": 8881672,
      "step": 15300
    },
    {
      "epoch": 2.2795650878760796,
      "grad_norm": 0.029296875,
      "learning_rate": 0.029982166650568166,
      "loss": 0.7914,
      "num_input_tokens_seen": 8884936,
      "step": 15305
    },
    {
      "epoch": 2.2803098004170392,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029982071482356716,
      "loss": 0.813,
      "num_input_tokens_seen": 8887720,
      "step": 15310
    },
    {
      "epoch": 2.281054512957998,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029981976061038394,
      "loss": 0.7876,
      "num_input_tokens_seen": 8891240,
      "step": 15315
    },
    {
      "epoch": 2.2817992254989576,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02998188038661482,
      "loss": 0.7998,
      "num_input_tokens_seen": 8894120,
      "step": 15320
    },
    {
      "epoch": 2.2825439380399164,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029981784459087595,
      "loss": 0.8186,
      "num_input_tokens_seen": 8897032,
      "step": 15325
    },
    {
      "epoch": 2.2832886505808756,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029981688278458353,
      "loss": 0.8039,
      "num_input_tokens_seen": 8899688,
      "step": 15330
    },
    {
      "epoch": 2.284033363121835,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029981591844728712,
      "loss": 0.7997,
      "num_input_tokens_seen": 8902504,
      "step": 15335
    },
    {
      "epoch": 2.284778075662794,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.029981495157900304,
      "loss": 0.8022,
      "num_input_tokens_seen": 8905224,
      "step": 15340
    },
    {
      "epoch": 2.285522788203753,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029981398217974763,
      "loss": 0.7909,
      "num_input_tokens_seen": 8908040,
      "step": 15345
    },
    {
      "epoch": 2.2862675007447124,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.029981301024953725,
      "loss": 0.8139,
      "num_input_tokens_seen": 8911208,
      "step": 15350
    },
    {
      "epoch": 2.2870122132856716,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029981203578838836,
      "loss": 0.791,
      "num_input_tokens_seen": 8914472,
      "step": 15355
    },
    {
      "epoch": 2.287756925826631,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02998110587963173,
      "loss": 0.7975,
      "num_input_tokens_seen": 8917160,
      "step": 15360
    },
    {
      "epoch": 2.28850163836759,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02998100792733407,
      "loss": 0.79,
      "num_input_tokens_seen": 8920008,
      "step": 15365
    },
    {
      "epoch": 2.289246350908549,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029980909721947512,
      "loss": 0.8059,
      "num_input_tokens_seen": 8922888,
      "step": 15370
    },
    {
      "epoch": 2.2899910634495084,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029980811263473708,
      "loss": 0.795,
      "num_input_tokens_seen": 8925576,
      "step": 15375
    },
    {
      "epoch": 2.2907357759904676,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.02998071255191432,
      "loss": 0.8138,
      "num_input_tokens_seen": 8928328,
      "step": 15380
    },
    {
      "epoch": 2.291480488531427,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.02998061358727102,
      "loss": 0.8408,
      "num_input_tokens_seen": 8931208,
      "step": 15385
    },
    {
      "epoch": 2.292225201072386,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02998051436954548,
      "loss": 0.7943,
      "num_input_tokens_seen": 8933896,
      "step": 15390
    },
    {
      "epoch": 2.2929699136133452,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02998041489873938,
      "loss": 0.8062,
      "num_input_tokens_seen": 8936712,
      "step": 15395
    },
    {
      "epoch": 2.2937146261543044,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02998031517485439,
      "loss": 0.7924,
      "num_input_tokens_seen": 8939304,
      "step": 15400
    },
    {
      "epoch": 2.2944593386952636,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.0299802151978922,
      "loss": 0.7938,
      "num_input_tokens_seen": 8942120,
      "step": 15405
    },
    {
      "epoch": 2.295204051236223,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0299801149678545,
      "loss": 0.7994,
      "num_input_tokens_seen": 8945192,
      "step": 15410
    },
    {
      "epoch": 2.295948763777182,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.029980014484742985,
      "loss": 0.7895,
      "num_input_tokens_seen": 8948392,
      "step": 15415
    },
    {
      "epoch": 2.2966934763181412,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.02997991374855935,
      "loss": 0.7897,
      "num_input_tokens_seen": 8951016,
      "step": 15420
    },
    {
      "epoch": 2.2974381888591004,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029979812759305294,
      "loss": 0.7943,
      "num_input_tokens_seen": 8953800,
      "step": 15425
    },
    {
      "epoch": 2.2981829014000597,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02997971151698253,
      "loss": 0.8188,
      "num_input_tokens_seen": 8956808,
      "step": 15430
    },
    {
      "epoch": 2.298927613941019,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.029979610021592763,
      "loss": 0.8042,
      "num_input_tokens_seen": 8959592,
      "step": 15435
    },
    {
      "epoch": 2.299672326481978,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02997950827313771,
      "loss": 0.7745,
      "num_input_tokens_seen": 8962504,
      "step": 15440
    },
    {
      "epoch": 2.3004170390229373,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02997940627161909,
      "loss": 0.7954,
      "num_input_tokens_seen": 8965416,
      "step": 15445
    },
    {
      "epoch": 2.3011617515638965,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029979304017038626,
      "loss": 0.8053,
      "num_input_tokens_seen": 8968264,
      "step": 15450
    },
    {
      "epoch": 2.3019064641048557,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02997920150939804,
      "loss": 0.8173,
      "num_input_tokens_seen": 8970888,
      "step": 15455
    },
    {
      "epoch": 2.302651176645815,
      "grad_norm": 0.01190185546875,
      "learning_rate": 0.02997909874869907,
      "loss": 0.8039,
      "num_input_tokens_seen": 8973896,
      "step": 15460
    },
    {
      "epoch": 2.303395889186774,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029978995734943455,
      "loss": 0.7893,
      "num_input_tokens_seen": 8976712,
      "step": 15465
    },
    {
      "epoch": 2.3041406017277333,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02997889246813293,
      "loss": 0.8037,
      "num_input_tokens_seen": 8979784,
      "step": 15470
    },
    {
      "epoch": 2.3048853142686925,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029978788948269237,
      "loss": 0.8078,
      "num_input_tokens_seen": 8982760,
      "step": 15475
    },
    {
      "epoch": 2.3056300268096512,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02997868517535413,
      "loss": 0.8155,
      "num_input_tokens_seen": 8985448,
      "step": 15480
    },
    {
      "epoch": 2.306374739350611,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02997858114938936,
      "loss": 0.7943,
      "num_input_tokens_seen": 8988904,
      "step": 15485
    },
    {
      "epoch": 2.3071194518915696,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.029978476870376688,
      "loss": 0.8295,
      "num_input_tokens_seen": 8991368,
      "step": 15490
    },
    {
      "epoch": 2.3078641644325293,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029978372338317873,
      "loss": 0.7978,
      "num_input_tokens_seen": 8994152,
      "step": 15495
    },
    {
      "epoch": 2.308608876973488,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02997826755321468,
      "loss": 0.7956,
      "num_input_tokens_seen": 8997000,
      "step": 15500
    },
    {
      "epoch": 2.3093535895144472,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029978162515068876,
      "loss": 0.7963,
      "num_input_tokens_seen": 8999912,
      "step": 15505
    },
    {
      "epoch": 2.3100983020554064,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.02997805722388224,
      "loss": 0.807,
      "num_input_tokens_seen": 9002760,
      "step": 15510
    },
    {
      "epoch": 2.3108430145963657,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029977951679656556,
      "loss": 0.806,
      "num_input_tokens_seen": 9005672,
      "step": 15515
    },
    {
      "epoch": 2.311587727137325,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029977845882393595,
      "loss": 0.7861,
      "num_input_tokens_seen": 9008648,
      "step": 15520
    },
    {
      "epoch": 2.312332439678284,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029977739832095153,
      "loss": 0.7888,
      "num_input_tokens_seen": 9011464,
      "step": 15525
    },
    {
      "epoch": 2.3130771522192433,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.02997763352876302,
      "loss": 0.807,
      "num_input_tokens_seen": 9014280,
      "step": 15530
    },
    {
      "epoch": 2.3138218647602025,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02997752697239899,
      "loss": 0.7897,
      "num_input_tokens_seen": 9017032,
      "step": 15535
    },
    {
      "epoch": 2.3145665773011617,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029977420163004864,
      "loss": 0.7949,
      "num_input_tokens_seen": 9020072,
      "step": 15540
    },
    {
      "epoch": 2.315311289842121,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02997731310058245,
      "loss": 0.8038,
      "num_input_tokens_seen": 9022856,
      "step": 15545
    },
    {
      "epoch": 2.31605600238308,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02997720578513355,
      "loss": 0.814,
      "num_input_tokens_seen": 9025736,
      "step": 15550
    },
    {
      "epoch": 2.3168007149240393,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02997709821665998,
      "loss": 0.7938,
      "num_input_tokens_seen": 9029000,
      "step": 15555
    },
    {
      "epoch": 2.3175454274649985,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02997699039516356,
      "loss": 0.7751,
      "num_input_tokens_seen": 9031816,
      "step": 15560
    },
    {
      "epoch": 2.3182901400059577,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029976882320646107,
      "loss": 0.7904,
      "num_input_tokens_seen": 9034792,
      "step": 15565
    },
    {
      "epoch": 2.319034852546917,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02997677399310945,
      "loss": 0.7792,
      "num_input_tokens_seen": 9037608,
      "step": 15570
    },
    {
      "epoch": 2.319779565087876,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029976665412555416,
      "loss": 0.7795,
      "num_input_tokens_seen": 9040392,
      "step": 15575
    },
    {
      "epoch": 2.3205242776288353,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029976556578985847,
      "loss": 0.814,
      "num_input_tokens_seen": 9043272,
      "step": 15580
    },
    {
      "epoch": 2.3212689901697945,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02997644749240257,
      "loss": 0.8141,
      "num_input_tokens_seen": 9046376,
      "step": 15585
    },
    {
      "epoch": 2.3220137027107537,
      "grad_norm": 0.044921875,
      "learning_rate": 0.02997633815280744,
      "loss": 0.8095,
      "num_input_tokens_seen": 9049128,
      "step": 15590
    },
    {
      "epoch": 2.322758415251713,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02997622856020229,
      "loss": 0.7988,
      "num_input_tokens_seen": 9051944,
      "step": 15595
    },
    {
      "epoch": 2.323503127792672,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029976118714588988,
      "loss": 0.8362,
      "num_input_tokens_seen": 9054600,
      "step": 15600
    },
    {
      "epoch": 2.3242478403336313,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029976008615969378,
      "loss": 0.7863,
      "num_input_tokens_seen": 9057352,
      "step": 15605
    },
    {
      "epoch": 2.3249925528745905,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02997589826434532,
      "loss": 0.791,
      "num_input_tokens_seen": 9060040,
      "step": 15610
    },
    {
      "epoch": 2.3257372654155497,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02997578765971869,
      "loss": 0.7943,
      "num_input_tokens_seen": 9062984,
      "step": 15615
    },
    {
      "epoch": 2.326481977956509,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.029975676802091338,
      "loss": 0.8169,
      "num_input_tokens_seen": 9065928,
      "step": 15620
    },
    {
      "epoch": 2.327226690497468,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029975565691465155,
      "loss": 0.7992,
      "num_input_tokens_seen": 9068872,
      "step": 15625
    },
    {
      "epoch": 2.3279714030384273,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02997545432784201,
      "loss": 0.8184,
      "num_input_tokens_seen": 9071432,
      "step": 15630
    },
    {
      "epoch": 2.3287161155793865,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02997534271122378,
      "loss": 0.7791,
      "num_input_tokens_seen": 9074312,
      "step": 15635
    },
    {
      "epoch": 2.3294608281203457,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029975230841612358,
      "loss": 0.805,
      "num_input_tokens_seen": 9077128,
      "step": 15640
    },
    {
      "epoch": 2.330205540661305,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029975118719009628,
      "loss": 0.7913,
      "num_input_tokens_seen": 9080008,
      "step": 15645
    },
    {
      "epoch": 2.330950253202264,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02997500634341749,
      "loss": 0.8157,
      "num_input_tokens_seen": 9083016,
      "step": 15650
    },
    {
      "epoch": 2.331694965743223,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029974893714837837,
      "loss": 0.8097,
      "num_input_tokens_seen": 9085672,
      "step": 15655
    },
    {
      "epoch": 2.3324396782841825,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.02997478083327258,
      "loss": 0.8182,
      "num_input_tokens_seen": 9088392,
      "step": 15660
    },
    {
      "epoch": 2.3331843908251413,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.029974667698723617,
      "loss": 0.7827,
      "num_input_tokens_seen": 9091368,
      "step": 15665
    },
    {
      "epoch": 2.333929103366101,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02997455431119286,
      "loss": 0.8026,
      "num_input_tokens_seen": 9094216,
      "step": 15670
    },
    {
      "epoch": 2.3346738159070597,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.029974440670682233,
      "loss": 0.837,
      "num_input_tokens_seen": 9096968,
      "step": 15675
    },
    {
      "epoch": 2.335418528448019,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029974326777193645,
      "loss": 0.8036,
      "num_input_tokens_seen": 9099976,
      "step": 15680
    },
    {
      "epoch": 2.336163240988978,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.029974212630729028,
      "loss": 0.7864,
      "num_input_tokens_seen": 9102856,
      "step": 15685
    },
    {
      "epoch": 2.3369079535299373,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02997409823129031,
      "loss": 0.8024,
      "num_input_tokens_seen": 9106088,
      "step": 15690
    },
    {
      "epoch": 2.3376526660708965,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02997398357887942,
      "loss": 0.8104,
      "num_input_tokens_seen": 9109096,
      "step": 15695
    },
    {
      "epoch": 2.3383973786118557,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029973868673498296,
      "loss": 0.8084,
      "num_input_tokens_seen": 9111848,
      "step": 15700
    },
    {
      "epoch": 2.339142091152815,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029973753515148874,
      "loss": 0.8073,
      "num_input_tokens_seen": 9114504,
      "step": 15705
    },
    {
      "epoch": 2.339886803693774,
      "grad_norm": 0.01422119140625,
      "learning_rate": 0.029973638103833113,
      "loss": 0.7975,
      "num_input_tokens_seen": 9117256,
      "step": 15710
    },
    {
      "epoch": 2.3406315162347333,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.02997352243955295,
      "loss": 0.8011,
      "num_input_tokens_seen": 9120552,
      "step": 15715
    },
    {
      "epoch": 2.3413762287756925,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029973406522310342,
      "loss": 0.8182,
      "num_input_tokens_seen": 9123144,
      "step": 15720
    },
    {
      "epoch": 2.3421209413166517,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029973290352107256,
      "loss": 0.804,
      "num_input_tokens_seen": 9126120,
      "step": 15725
    },
    {
      "epoch": 2.342865653857611,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.029973173928945645,
      "loss": 0.7936,
      "num_input_tokens_seen": 9129064,
      "step": 15730
    },
    {
      "epoch": 2.34361036639857,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029973057252827475,
      "loss": 0.8118,
      "num_input_tokens_seen": 9131880,
      "step": 15735
    },
    {
      "epoch": 2.3443550789395293,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029972940323754724,
      "loss": 0.7791,
      "num_input_tokens_seen": 9134600,
      "step": 15740
    },
    {
      "epoch": 2.3450997914804885,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.029972823141729363,
      "loss": 0.8124,
      "num_input_tokens_seen": 9137448,
      "step": 15745
    },
    {
      "epoch": 2.3458445040214477,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029972705706753373,
      "loss": 0.797,
      "num_input_tokens_seen": 9140072,
      "step": 15750
    },
    {
      "epoch": 2.346589216562407,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02997258801882874,
      "loss": 0.7949,
      "num_input_tokens_seen": 9142728,
      "step": 15755
    },
    {
      "epoch": 2.347333929103366,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029972470077957447,
      "loss": 0.8038,
      "num_input_tokens_seen": 9145448,
      "step": 15760
    },
    {
      "epoch": 2.3480786416443253,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029972351884141487,
      "loss": 0.7996,
      "num_input_tokens_seen": 9148232,
      "step": 15765
    },
    {
      "epoch": 2.3488233541852845,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029972233437382867,
      "loss": 0.7735,
      "num_input_tokens_seen": 9151400,
      "step": 15770
    },
    {
      "epoch": 2.3495680667262437,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02997211473768357,
      "loss": 0.802,
      "num_input_tokens_seen": 9154440,
      "step": 15775
    },
    {
      "epoch": 2.350312779267203,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029971995785045617,
      "loss": 0.8192,
      "num_input_tokens_seen": 9157352,
      "step": 15780
    },
    {
      "epoch": 2.351057491808162,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.02997187657947101,
      "loss": 0.8064,
      "num_input_tokens_seen": 9160200,
      "step": 15785
    },
    {
      "epoch": 2.3518022043491214,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02997175712096177,
      "loss": 0.8129,
      "num_input_tokens_seen": 9163144,
      "step": 15790
    },
    {
      "epoch": 2.3525469168900806,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029971637409519906,
      "loss": 0.8197,
      "num_input_tokens_seen": 9166344,
      "step": 15795
    },
    {
      "epoch": 2.3532916294310398,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.029971517445147446,
      "loss": 0.8124,
      "num_input_tokens_seen": 9169032,
      "step": 15800
    },
    {
      "epoch": 2.354036341971999,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029971397227846416,
      "loss": 0.7713,
      "num_input_tokens_seen": 9172104,
      "step": 15805
    },
    {
      "epoch": 2.354781054512958,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029971276757618844,
      "loss": 0.7907,
      "num_input_tokens_seen": 9174920,
      "step": 15810
    },
    {
      "epoch": 2.3555257670539174,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.029971156034466765,
      "loss": 0.8126,
      "num_input_tokens_seen": 9177864,
      "step": 15815
    },
    {
      "epoch": 2.3562704795948766,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029971035058392225,
      "loss": 0.7967,
      "num_input_tokens_seen": 9180552,
      "step": 15820
    },
    {
      "epoch": 2.3570151921358358,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.02997091382939726,
      "loss": 0.8037,
      "num_input_tokens_seen": 9183304,
      "step": 15825
    },
    {
      "epoch": 2.3577599046767945,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.029970792347483927,
      "loss": 0.7863,
      "num_input_tokens_seen": 9186088,
      "step": 15830
    },
    {
      "epoch": 2.358504617217754,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02997067061265427,
      "loss": 0.8176,
      "num_input_tokens_seen": 9188968,
      "step": 15835
    },
    {
      "epoch": 2.359249329758713,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029970548624910345,
      "loss": 0.8024,
      "num_input_tokens_seen": 9191592,
      "step": 15840
    },
    {
      "epoch": 2.359994042299672,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02997042638425422,
      "loss": 0.7885,
      "num_input_tokens_seen": 9194504,
      "step": 15845
    },
    {
      "epoch": 2.3607387548406313,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029970303890687958,
      "loss": 0.8295,
      "num_input_tokens_seen": 9197224,
      "step": 15850
    },
    {
      "epoch": 2.3614834673815905,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.029970181144213626,
      "loss": 0.7795,
      "num_input_tokens_seen": 9200264,
      "step": 15855
    },
    {
      "epoch": 2.3622281799225497,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029970058144833295,
      "loss": 0.8081,
      "num_input_tokens_seen": 9203112,
      "step": 15860
    },
    {
      "epoch": 2.362972892463509,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.029969934892549052,
      "loss": 0.8026,
      "num_input_tokens_seen": 9205928,
      "step": 15865
    },
    {
      "epoch": 2.363717605004468,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.029969811387362967,
      "loss": 0.8055,
      "num_input_tokens_seen": 9208584,
      "step": 15870
    },
    {
      "epoch": 2.3644623175454274,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02996968762927714,
      "loss": 0.7905,
      "num_input_tokens_seen": 9211240,
      "step": 15875
    },
    {
      "epoch": 2.3652070300863866,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029969563618293654,
      "loss": 0.7961,
      "num_input_tokens_seen": 9214568,
      "step": 15880
    },
    {
      "epoch": 2.3659517426273458,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.0299694393544146,
      "loss": 0.8024,
      "num_input_tokens_seen": 9217384,
      "step": 15885
    },
    {
      "epoch": 2.366696455168305,
      "grad_norm": 0.03125,
      "learning_rate": 0.029969314837642085,
      "loss": 0.8032,
      "num_input_tokens_seen": 9220136,
      "step": 15890
    },
    {
      "epoch": 2.367441167709264,
      "grad_norm": 0.013671875,
      "learning_rate": 0.029969190067978213,
      "loss": 0.8001,
      "num_input_tokens_seen": 9222984,
      "step": 15895
    },
    {
      "epoch": 2.3681858802502234,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029969065045425083,
      "loss": 0.7927,
      "num_input_tokens_seen": 9225896,
      "step": 15900
    },
    {
      "epoch": 2.3689305927911826,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029968939769984815,
      "loss": 0.8017,
      "num_input_tokens_seen": 9228776,
      "step": 15905
    },
    {
      "epoch": 2.3696753053321418,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029968814241659526,
      "loss": 0.7734,
      "num_input_tokens_seen": 9231720,
      "step": 15910
    },
    {
      "epoch": 2.370420017873101,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029968688460451334,
      "loss": 0.7981,
      "num_input_tokens_seen": 9234600,
      "step": 15915
    },
    {
      "epoch": 2.37116473041406,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029968562426362363,
      "loss": 0.8094,
      "num_input_tokens_seen": 9237288,
      "step": 15920
    },
    {
      "epoch": 2.3719094429550194,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.029968436139394743,
      "loss": 0.7745,
      "num_input_tokens_seen": 9240104,
      "step": 15925
    },
    {
      "epoch": 2.3726541554959786,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.029968309599550606,
      "loss": 0.8383,
      "num_input_tokens_seen": 9242888,
      "step": 15930
    },
    {
      "epoch": 2.373398868036938,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029968182806832094,
      "loss": 0.7948,
      "num_input_tokens_seen": 9246024,
      "step": 15935
    },
    {
      "epoch": 2.374143580577897,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.029968055761241345,
      "loss": 0.8159,
      "num_input_tokens_seen": 9248936,
      "step": 15940
    },
    {
      "epoch": 2.374888293118856,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029967928462780506,
      "loss": 0.7799,
      "num_input_tokens_seen": 9251912,
      "step": 15945
    },
    {
      "epoch": 2.3756330056598154,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02996780091145173,
      "loss": 0.8146,
      "num_input_tokens_seen": 9254600,
      "step": 15950
    },
    {
      "epoch": 2.3763777182007746,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.02996767310725717,
      "loss": 0.7904,
      "num_input_tokens_seen": 9257864,
      "step": 15955
    },
    {
      "epoch": 2.377122430741734,
      "grad_norm": 0.03125,
      "learning_rate": 0.029967545050198982,
      "loss": 0.7856,
      "num_input_tokens_seen": 9260680,
      "step": 15960
    },
    {
      "epoch": 2.377867143282693,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029967416740279335,
      "loss": 0.801,
      "num_input_tokens_seen": 9263432,
      "step": 15965
    },
    {
      "epoch": 2.378611855823652,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.029967288177500398,
      "loss": 0.816,
      "num_input_tokens_seen": 9266056,
      "step": 15970
    },
    {
      "epoch": 2.3793565683646114,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029967159361864334,
      "loss": 0.8293,
      "num_input_tokens_seen": 9269224,
      "step": 15975
    },
    {
      "epoch": 2.3801012809055706,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.029967030293373324,
      "loss": 0.8205,
      "num_input_tokens_seen": 9272136,
      "step": 15980
    },
    {
      "epoch": 2.38084599344653,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029966900972029555,
      "loss": 0.7811,
      "num_input_tokens_seen": 9275048,
      "step": 15985
    },
    {
      "epoch": 2.381590705987489,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.029966771397835205,
      "loss": 0.8169,
      "num_input_tokens_seen": 9277768,
      "step": 15990
    },
    {
      "epoch": 2.382335418528448,
      "grad_norm": 0.03125,
      "learning_rate": 0.029966641570792457,
      "loss": 0.8008,
      "num_input_tokens_seen": 9280584,
      "step": 15995
    },
    {
      "epoch": 2.3830801310694074,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029966511490903517,
      "loss": 0.8137,
      "num_input_tokens_seen": 9283624,
      "step": 16000
    },
    {
      "epoch": 2.383824843610366,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029966381158170574,
      "loss": 0.8003,
      "num_input_tokens_seen": 9286472,
      "step": 16005
    },
    {
      "epoch": 2.384569556151326,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029966250572595835,
      "loss": 0.8027,
      "num_input_tokens_seen": 9289416,
      "step": 16010
    },
    {
      "epoch": 2.3853142686922846,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.029966119734181503,
      "loss": 0.8253,
      "num_input_tokens_seen": 9292488,
      "step": 16015
    },
    {
      "epoch": 2.386058981233244,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.029965988642929785,
      "loss": 0.8047,
      "num_input_tokens_seen": 9295528,
      "step": 16020
    },
    {
      "epoch": 2.386803693774203,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029965857298842904,
      "loss": 0.8049,
      "num_input_tokens_seen": 9298664,
      "step": 16025
    },
    {
      "epoch": 2.387548406315162,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029965725701923074,
      "loss": 0.7893,
      "num_input_tokens_seen": 9301352,
      "step": 16030
    },
    {
      "epoch": 2.3882931188561214,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02996559385217252,
      "loss": 0.8006,
      "num_input_tokens_seen": 9304232,
      "step": 16035
    },
    {
      "epoch": 2.3890378313970806,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.029965461749593462,
      "loss": 0.7887,
      "num_input_tokens_seen": 9307176,
      "step": 16040
    },
    {
      "epoch": 2.38978254393804,
      "grad_norm": 0.037109375,
      "learning_rate": 0.029965329394188146,
      "loss": 0.8017,
      "num_input_tokens_seen": 9309960,
      "step": 16045
    },
    {
      "epoch": 2.390527256478999,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029965196785958794,
      "loss": 0.7937,
      "num_input_tokens_seen": 9312968,
      "step": 16050
    },
    {
      "epoch": 2.391271969019958,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.029965063924907658,
      "loss": 0.8209,
      "num_input_tokens_seen": 9315912,
      "step": 16055
    },
    {
      "epoch": 2.3920166815609174,
      "grad_norm": 0.01422119140625,
      "learning_rate": 0.02996493081103697,
      "loss": 0.7917,
      "num_input_tokens_seen": 9318984,
      "step": 16060
    },
    {
      "epoch": 2.3927613941018766,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02996479744434899,
      "loss": 0.7942,
      "num_input_tokens_seen": 9321832,
      "step": 16065
    },
    {
      "epoch": 2.393506106642836,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02996466382484596,
      "loss": 0.807,
      "num_input_tokens_seen": 9324968,
      "step": 16070
    },
    {
      "epoch": 2.394250819183795,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.029964529952530156,
      "loss": 0.8129,
      "num_input_tokens_seen": 9327816,
      "step": 16075
    },
    {
      "epoch": 2.394995531724754,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.029964395827403817,
      "loss": 0.8087,
      "num_input_tokens_seen": 9330472,
      "step": 16080
    },
    {
      "epoch": 2.3957402442657134,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029964261449469225,
      "loss": 0.8,
      "num_input_tokens_seen": 9333544,
      "step": 16085
    },
    {
      "epoch": 2.3964849568066726,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.029964126818728645,
      "loss": 0.7858,
      "num_input_tokens_seen": 9336424,
      "step": 16090
    },
    {
      "epoch": 2.397229669347632,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02996399193518435,
      "loss": 0.8202,
      "num_input_tokens_seen": 9339464,
      "step": 16095
    },
    {
      "epoch": 2.397974381888591,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02996385679883862,
      "loss": 0.8134,
      "num_input_tokens_seen": 9342344,
      "step": 16100
    },
    {
      "epoch": 2.3987190944295502,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02996372140969374,
      "loss": 0.7902,
      "num_input_tokens_seen": 9345160,
      "step": 16105
    },
    {
      "epoch": 2.3994638069705094,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02996358576775199,
      "loss": 0.8157,
      "num_input_tokens_seen": 9348200,
      "step": 16110
    },
    {
      "epoch": 2.4002085195114686,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02996344987301567,
      "loss": 0.8097,
      "num_input_tokens_seen": 9350824,
      "step": 16115
    },
    {
      "epoch": 2.400953232052428,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02996331372548707,
      "loss": 0.811,
      "num_input_tokens_seen": 9353832,
      "step": 16120
    },
    {
      "epoch": 2.401697944593387,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.029963177325168497,
      "loss": 0.8196,
      "num_input_tokens_seen": 9356456,
      "step": 16125
    },
    {
      "epoch": 2.4024426571343462,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.02996304067206225,
      "loss": 0.7958,
      "num_input_tokens_seen": 9359464,
      "step": 16130
    },
    {
      "epoch": 2.4031873696753054,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.029962903766170636,
      "loss": 0.8043,
      "num_input_tokens_seen": 9362600,
      "step": 16135
    },
    {
      "epoch": 2.4039320822162646,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029962766607495972,
      "loss": 0.8028,
      "num_input_tokens_seen": 9365352,
      "step": 16140
    },
    {
      "epoch": 2.404676794757224,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029962629196040577,
      "loss": 0.7999,
      "num_input_tokens_seen": 9368104,
      "step": 16145
    },
    {
      "epoch": 2.405421507298183,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029962491531806766,
      "loss": 0.803,
      "num_input_tokens_seen": 9370952,
      "step": 16150
    },
    {
      "epoch": 2.4061662198391423,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029962353614796867,
      "loss": 0.7964,
      "num_input_tokens_seen": 9373832,
      "step": 16155
    },
    {
      "epoch": 2.4069109323801015,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.029962215445013214,
      "loss": 0.8017,
      "num_input_tokens_seen": 9376808,
      "step": 16160
    },
    {
      "epoch": 2.4076556449210607,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029962077022458136,
      "loss": 0.7993,
      "num_input_tokens_seen": 9379400,
      "step": 16165
    },
    {
      "epoch": 2.4084003574620194,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02996193834713397,
      "loss": 0.7966,
      "num_input_tokens_seen": 9382024,
      "step": 16170
    },
    {
      "epoch": 2.409145070002979,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.029961799419043065,
      "loss": 0.7947,
      "num_input_tokens_seen": 9385032,
      "step": 16175
    },
    {
      "epoch": 2.409889782543938,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.029961660238187766,
      "loss": 0.8139,
      "num_input_tokens_seen": 9388040,
      "step": 16180
    },
    {
      "epoch": 2.4106344950848975,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029961520804570423,
      "loss": 0.7759,
      "num_input_tokens_seen": 9390856,
      "step": 16185
    },
    {
      "epoch": 2.4113792076258562,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029961381118193395,
      "loss": 0.7973,
      "num_input_tokens_seen": 9394088,
      "step": 16190
    },
    {
      "epoch": 2.4121239201668154,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.029961241179059035,
      "loss": 0.7846,
      "num_input_tokens_seen": 9397416,
      "step": 16195
    },
    {
      "epoch": 2.4128686327077746,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02996110098716971,
      "loss": 0.7799,
      "num_input_tokens_seen": 9400360,
      "step": 16200
    },
    {
      "epoch": 2.413613345248734,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.029960960542527795,
      "loss": 0.7866,
      "num_input_tokens_seen": 9403176,
      "step": 16205
    },
    {
      "epoch": 2.414358057789693,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02996081984513565,
      "loss": 0.7958,
      "num_input_tokens_seen": 9405896,
      "step": 16210
    },
    {
      "epoch": 2.4151027703306522,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029960678894995665,
      "loss": 0.8103,
      "num_input_tokens_seen": 9408872,
      "step": 16215
    },
    {
      "epoch": 2.4158474828716114,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029960537692110212,
      "loss": 0.8065,
      "num_input_tokens_seen": 9411976,
      "step": 16220
    },
    {
      "epoch": 2.4165921954125706,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.029960396236481685,
      "loss": 0.776,
      "num_input_tokens_seen": 9415112,
      "step": 16225
    },
    {
      "epoch": 2.41733690795353,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029960254528112466,
      "loss": 0.7972,
      "num_input_tokens_seen": 9417992,
      "step": 16230
    },
    {
      "epoch": 2.418081620494489,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02996011256700495,
      "loss": 0.7789,
      "num_input_tokens_seen": 9420808,
      "step": 16235
    },
    {
      "epoch": 2.4188263330354483,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.02995997035316154,
      "loss": 0.7525,
      "num_input_tokens_seen": 9423656,
      "step": 16240
    },
    {
      "epoch": 2.4195710455764075,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029959827886584633,
      "loss": 0.828,
      "num_input_tokens_seen": 9426600,
      "step": 16245
    },
    {
      "epoch": 2.4203157581173667,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029959685167276637,
      "loss": 0.8096,
      "num_input_tokens_seen": 9429608,
      "step": 16250
    },
    {
      "epoch": 2.421060470658326,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02995954219523997,
      "loss": 0.8104,
      "num_input_tokens_seen": 9432648,
      "step": 16255
    },
    {
      "epoch": 2.421805183199285,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02995939897047704,
      "loss": 0.8365,
      "num_input_tokens_seen": 9435624,
      "step": 16260
    },
    {
      "epoch": 2.4225498957402443,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029959255492990264,
      "loss": 0.7989,
      "num_input_tokens_seen": 9438632,
      "step": 16265
    },
    {
      "epoch": 2.4232946082812035,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029959111762782074,
      "loss": 0.7922,
      "num_input_tokens_seen": 9441416,
      "step": 16270
    },
    {
      "epoch": 2.4240393208221627,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029958967779854892,
      "loss": 0.8193,
      "num_input_tokens_seen": 9443944,
      "step": 16275
    },
    {
      "epoch": 2.424784033363122,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.029958823544211158,
      "loss": 0.8077,
      "num_input_tokens_seen": 9446952,
      "step": 16280
    },
    {
      "epoch": 2.425528745904081,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.029958679055853296,
      "loss": 0.783,
      "num_input_tokens_seen": 9449736,
      "step": 16285
    },
    {
      "epoch": 2.4262734584450403,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02995853431478376,
      "loss": 0.8209,
      "num_input_tokens_seen": 9452424,
      "step": 16290
    },
    {
      "epoch": 2.4270181709859995,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029958389321004992,
      "loss": 0.8209,
      "num_input_tokens_seen": 9455208,
      "step": 16295
    },
    {
      "epoch": 2.4277628835269587,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.029958244074519434,
      "loss": 0.775,
      "num_input_tokens_seen": 9458312,
      "step": 16300
    },
    {
      "epoch": 2.428507596067918,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02995809857532955,
      "loss": 0.8019,
      "num_input_tokens_seen": 9461160,
      "step": 16305
    },
    {
      "epoch": 2.429252308608877,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02995795282343779,
      "loss": 0.815,
      "num_input_tokens_seen": 9464040,
      "step": 16310
    },
    {
      "epoch": 2.4299970211498363,
      "grad_norm": 0.037109375,
      "learning_rate": 0.029957806818846622,
      "loss": 0.7965,
      "num_input_tokens_seen": 9466984,
      "step": 16315
    },
    {
      "epoch": 2.4307417336907955,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02995766056155851,
      "loss": 0.8122,
      "num_input_tokens_seen": 9470056,
      "step": 16320
    },
    {
      "epoch": 2.4314864462317547,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.02995751405157592,
      "loss": 0.7904,
      "num_input_tokens_seen": 9472936,
      "step": 16325
    },
    {
      "epoch": 2.432231158772714,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.029957367288901336,
      "loss": 0.8058,
      "num_input_tokens_seen": 9475592,
      "step": 16330
    },
    {
      "epoch": 2.432975871313673,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029957220273537234,
      "loss": 0.8122,
      "num_input_tokens_seen": 9478312,
      "step": 16335
    },
    {
      "epoch": 2.4337205838546323,
      "grad_norm": 0.025390625,
      "learning_rate": 0.0299570730054861,
      "loss": 0.8227,
      "num_input_tokens_seen": 9481384,
      "step": 16340
    },
    {
      "epoch": 2.434465296395591,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.029956925484750412,
      "loss": 0.816,
      "num_input_tokens_seen": 9484200,
      "step": 16345
    },
    {
      "epoch": 2.4352100089365507,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029956777711332677,
      "loss": 0.7909,
      "num_input_tokens_seen": 9487208,
      "step": 16350
    },
    {
      "epoch": 2.4359547214775095,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.02995662968523538,
      "loss": 0.8331,
      "num_input_tokens_seen": 9490152,
      "step": 16355
    },
    {
      "epoch": 2.436699434018469,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.029956481406461025,
      "loss": 0.8046,
      "num_input_tokens_seen": 9493064,
      "step": 16360
    },
    {
      "epoch": 2.437444146559428,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02995633287501212,
      "loss": 0.805,
      "num_input_tokens_seen": 9496008,
      "step": 16365
    },
    {
      "epoch": 2.438188859100387,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.029956184090891173,
      "loss": 0.797,
      "num_input_tokens_seen": 9498792,
      "step": 16370
    },
    {
      "epoch": 2.4389335716413463,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02995603505410069,
      "loss": 0.797,
      "num_input_tokens_seen": 9502056,
      "step": 16375
    },
    {
      "epoch": 2.4396782841823055,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029955885764643204,
      "loss": 0.7984,
      "num_input_tokens_seen": 9504744,
      "step": 16380
    },
    {
      "epoch": 2.4404229967232647,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.029955736222521224,
      "loss": 0.8109,
      "num_input_tokens_seen": 9507816,
      "step": 16385
    },
    {
      "epoch": 2.441167709264224,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029955586427737282,
      "loss": 0.8129,
      "num_input_tokens_seen": 9510888,
      "step": 16390
    },
    {
      "epoch": 2.441912421805183,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.02995543638029391,
      "loss": 0.7926,
      "num_input_tokens_seen": 9513768,
      "step": 16395
    },
    {
      "epoch": 2.4426571343461423,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029955286080193638,
      "loss": 0.8079,
      "num_input_tokens_seen": 9516648,
      "step": 16400
    },
    {
      "epoch": 2.4434018468871015,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029955135527439006,
      "loss": 0.807,
      "num_input_tokens_seen": 9519784,
      "step": 16405
    },
    {
      "epoch": 2.4441465594280607,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02995498472203256,
      "loss": 0.802,
      "num_input_tokens_seen": 9522888,
      "step": 16410
    },
    {
      "epoch": 2.44489127196902,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029954833663976846,
      "loss": 0.7876,
      "num_input_tokens_seen": 9525608,
      "step": 16415
    },
    {
      "epoch": 2.445635984509979,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.029954682353274417,
      "loss": 0.7871,
      "num_input_tokens_seen": 9528840,
      "step": 16420
    },
    {
      "epoch": 2.4463806970509383,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.029954530789927828,
      "loss": 0.8131,
      "num_input_tokens_seen": 9531752,
      "step": 16425
    },
    {
      "epoch": 2.4471254095918975,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.029954378973939644,
      "loss": 0.8038,
      "num_input_tokens_seen": 9534696,
      "step": 16430
    },
    {
      "epoch": 2.4478701221328567,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029954226905312423,
      "loss": 0.797,
      "num_input_tokens_seen": 9537352,
      "step": 16435
    },
    {
      "epoch": 2.448614834673816,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029954074584048737,
      "loss": 0.8043,
      "num_input_tokens_seen": 9540392,
      "step": 16440
    },
    {
      "epoch": 2.449359547214775,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029953922010151162,
      "loss": 0.7945,
      "num_input_tokens_seen": 9543144,
      "step": 16445
    },
    {
      "epoch": 2.4501042597557343,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029953769183622272,
      "loss": 0.8004,
      "num_input_tokens_seen": 9546184,
      "step": 16450
    },
    {
      "epoch": 2.4508489722966935,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02995361610446465,
      "loss": 0.8028,
      "num_input_tokens_seen": 9549032,
      "step": 16455
    },
    {
      "epoch": 2.4515936848376527,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02995346277268088,
      "loss": 0.8088,
      "num_input_tokens_seen": 9552296,
      "step": 16460
    },
    {
      "epoch": 2.452338397378612,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029953309188273557,
      "loss": 0.8139,
      "num_input_tokens_seen": 9555272,
      "step": 16465
    },
    {
      "epoch": 2.453083109919571,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.02995315535124527,
      "loss": 0.8091,
      "num_input_tokens_seen": 9558184,
      "step": 16470
    },
    {
      "epoch": 2.4538278224605303,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029953001261598625,
      "loss": 0.803,
      "num_input_tokens_seen": 9561064,
      "step": 16475
    },
    {
      "epoch": 2.4545725350014895,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02995284691933622,
      "loss": 0.8007,
      "num_input_tokens_seen": 9564072,
      "step": 16480
    },
    {
      "epoch": 2.4553172475424487,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029952692324460666,
      "loss": 0.805,
      "num_input_tokens_seen": 9566728,
      "step": 16485
    },
    {
      "epoch": 2.456061960083408,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02995253747697457,
      "loss": 0.8113,
      "num_input_tokens_seen": 9569608,
      "step": 16490
    },
    {
      "epoch": 2.456806672624367,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02995238237688055,
      "loss": 0.8,
      "num_input_tokens_seen": 9572392,
      "step": 16495
    },
    {
      "epoch": 2.4575513851653263,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.029952227024181224,
      "loss": 0.8242,
      "num_input_tokens_seen": 9575400,
      "step": 16500
    },
    {
      "epoch": 2.4582960977062855,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02995207141887922,
      "loss": 0.819,
      "num_input_tokens_seen": 9578312,
      "step": 16505
    },
    {
      "epoch": 2.4590408102472447,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02995191556097717,
      "loss": 0.8011,
      "num_input_tokens_seen": 9581000,
      "step": 16510
    },
    {
      "epoch": 2.459785522788204,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.0299517594504777,
      "loss": 0.7807,
      "num_input_tokens_seen": 9583688,
      "step": 16515
    },
    {
      "epoch": 2.4605302353291627,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029951603087383455,
      "loss": 0.8095,
      "num_input_tokens_seen": 9587080,
      "step": 16520
    },
    {
      "epoch": 2.4612749478701224,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.029951446471697066,
      "loss": 0.801,
      "num_input_tokens_seen": 9589960,
      "step": 16525
    },
    {
      "epoch": 2.462019660411081,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.029951289603421187,
      "loss": 0.8145,
      "num_input_tokens_seen": 9592936,
      "step": 16530
    },
    {
      "epoch": 2.4627643729520408,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.029951132482558467,
      "loss": 0.7942,
      "num_input_tokens_seen": 9595848,
      "step": 16535
    },
    {
      "epoch": 2.4635090854929995,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02995097510911156,
      "loss": 0.8052,
      "num_input_tokens_seen": 9598600,
      "step": 16540
    },
    {
      "epoch": 2.4642537980339587,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029950817483083123,
      "loss": 0.8034,
      "num_input_tokens_seen": 9601320,
      "step": 16545
    },
    {
      "epoch": 2.464998510574918,
      "grad_norm": 0.046875,
      "learning_rate": 0.029950659604475818,
      "loss": 0.7845,
      "num_input_tokens_seen": 9604584,
      "step": 16550
    },
    {
      "epoch": 2.465743223115877,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029950501473292316,
      "loss": 0.8035,
      "num_input_tokens_seen": 9607560,
      "step": 16555
    },
    {
      "epoch": 2.4664879356568363,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.029950343089535286,
      "loss": 0.8011,
      "num_input_tokens_seen": 9610216,
      "step": 16560
    },
    {
      "epoch": 2.4672326481977955,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.029950184453207407,
      "loss": 0.8182,
      "num_input_tokens_seen": 9613224,
      "step": 16565
    },
    {
      "epoch": 2.4679773607387547,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029950025564311353,
      "loss": 0.7867,
      "num_input_tokens_seen": 9615848,
      "step": 16570
    },
    {
      "epoch": 2.468722073279714,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02994986642284981,
      "loss": 0.817,
      "num_input_tokens_seen": 9618600,
      "step": 16575
    },
    {
      "epoch": 2.469466785820673,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029949707028825476,
      "loss": 0.7777,
      "num_input_tokens_seen": 9621480,
      "step": 16580
    },
    {
      "epoch": 2.4702114983616323,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029949547382241027,
      "loss": 0.8322,
      "num_input_tokens_seen": 9624488,
      "step": 16585
    },
    {
      "epoch": 2.4709562109025915,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.029949387483099176,
      "loss": 0.786,
      "num_input_tokens_seen": 9627720,
      "step": 16590
    },
    {
      "epoch": 2.4717009234435507,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.029949227331402612,
      "loss": 0.8309,
      "num_input_tokens_seen": 9630856,
      "step": 16595
    },
    {
      "epoch": 2.47244563598451,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02994906692715405,
      "loss": 0.7893,
      "num_input_tokens_seen": 9633928,
      "step": 16600
    },
    {
      "epoch": 2.473190348525469,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.029948906270356193,
      "loss": 0.7946,
      "num_input_tokens_seen": 9637064,
      "step": 16605
    },
    {
      "epoch": 2.4739350610664284,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029948745361011758,
      "loss": 0.7956,
      "num_input_tokens_seen": 9640168,
      "step": 16610
    },
    {
      "epoch": 2.4746797736073876,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029948584199123465,
      "loss": 0.8056,
      "num_input_tokens_seen": 9642888,
      "step": 16615
    },
    {
      "epoch": 2.4754244861483468,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02994842278469403,
      "loss": 0.808,
      "num_input_tokens_seen": 9645832,
      "step": 16620
    },
    {
      "epoch": 2.476169198689306,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029948261117726194,
      "loss": 0.7911,
      "num_input_tokens_seen": 9648680,
      "step": 16625
    },
    {
      "epoch": 2.476913911230265,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.029948099198222673,
      "loss": 0.795,
      "num_input_tokens_seen": 9651304,
      "step": 16630
    },
    {
      "epoch": 2.4776586237712244,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.02994793702618621,
      "loss": 0.8209,
      "num_input_tokens_seen": 9654184,
      "step": 16635
    },
    {
      "epoch": 2.4784033363121836,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029947774601619546,
      "loss": 0.8191,
      "num_input_tokens_seen": 9657000,
      "step": 16640
    },
    {
      "epoch": 2.4791480488531428,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029947611924525417,
      "loss": 0.804,
      "num_input_tokens_seen": 9660040,
      "step": 16645
    },
    {
      "epoch": 2.479892761394102,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02994744899490658,
      "loss": 0.7945,
      "num_input_tokens_seen": 9662984,
      "step": 16650
    },
    {
      "epoch": 2.480637473935061,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.029947285812765782,
      "loss": 0.7914,
      "num_input_tokens_seen": 9665704,
      "step": 16655
    },
    {
      "epoch": 2.4813821864760204,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029947122378105782,
      "loss": 0.7793,
      "num_input_tokens_seen": 9668296,
      "step": 16660
    },
    {
      "epoch": 2.4821268990169796,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.029946958690929344,
      "loss": 0.8015,
      "num_input_tokens_seen": 9671112,
      "step": 16665
    },
    {
      "epoch": 2.482871611557939,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02994679475123923,
      "loss": 0.7958,
      "num_input_tokens_seen": 9674248,
      "step": 16670
    },
    {
      "epoch": 2.483616324098898,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02994663055903821,
      "loss": 0.7791,
      "num_input_tokens_seen": 9677608,
      "step": 16675
    },
    {
      "epoch": 2.484361036639857,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02994646611432905,
      "loss": 0.8147,
      "num_input_tokens_seen": 9680712,
      "step": 16680
    },
    {
      "epoch": 2.4851057491808164,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029946301417114546,
      "loss": 0.8249,
      "num_input_tokens_seen": 9683496,
      "step": 16685
    },
    {
      "epoch": 2.4858504617217756,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.029946136467397466,
      "loss": 0.7869,
      "num_input_tokens_seen": 9686344,
      "step": 16690
    },
    {
      "epoch": 2.4865951742627344,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.0299459712651806,
      "loss": 0.7899,
      "num_input_tokens_seen": 9689576,
      "step": 16695
    },
    {
      "epoch": 2.487339886803694,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.029945805810466745,
      "loss": 0.8053,
      "num_input_tokens_seen": 9692136,
      "step": 16700
    },
    {
      "epoch": 2.4880845993446528,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.029945640103258685,
      "loss": 0.8536,
      "num_input_tokens_seen": 9694856,
      "step": 16705
    },
    {
      "epoch": 2.488829311885612,
      "grad_norm": 0.013671875,
      "learning_rate": 0.02994547414355923,
      "loss": 0.7898,
      "num_input_tokens_seen": 9697832,
      "step": 16710
    },
    {
      "epoch": 2.489574024426571,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029945307931371175,
      "loss": 0.7855,
      "num_input_tokens_seen": 9700456,
      "step": 16715
    },
    {
      "epoch": 2.4903187369675304,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029945141466697335,
      "loss": 0.8322,
      "num_input_tokens_seen": 9703272,
      "step": 16720
    },
    {
      "epoch": 2.4910634495084896,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.029944974749540523,
      "loss": 0.8101,
      "num_input_tokens_seen": 9706056,
      "step": 16725
    },
    {
      "epoch": 2.4918081620494488,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029944807779903545,
      "loss": 0.7977,
      "num_input_tokens_seen": 9708712,
      "step": 16730
    },
    {
      "epoch": 2.492552874590408,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.029944640557789233,
      "loss": 0.7859,
      "num_input_tokens_seen": 9711176,
      "step": 16735
    },
    {
      "epoch": 2.493297587131367,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029944473083200404,
      "loss": 0.8075,
      "num_input_tokens_seen": 9713768,
      "step": 16740
    },
    {
      "epoch": 2.4940422996723264,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.029944305356139896,
      "loss": 0.8009,
      "num_input_tokens_seen": 9716296,
      "step": 16745
    },
    {
      "epoch": 2.4947870122132856,
      "grad_norm": 0.015625,
      "learning_rate": 0.029944137376610537,
      "loss": 0.8029,
      "num_input_tokens_seen": 9719144,
      "step": 16750
    },
    {
      "epoch": 2.495531724754245,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029943969144615164,
      "loss": 0.8212,
      "num_input_tokens_seen": 9722472,
      "step": 16755
    },
    {
      "epoch": 2.496276437295204,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.02994380066015662,
      "loss": 0.7846,
      "num_input_tokens_seen": 9725288,
      "step": 16760
    },
    {
      "epoch": 2.497021149836163,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029943631923237752,
      "loss": 0.794,
      "num_input_tokens_seen": 9728136,
      "step": 16765
    },
    {
      "epoch": 2.4977658623771224,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02994346293386141,
      "loss": 0.8312,
      "num_input_tokens_seen": 9730984,
      "step": 16770
    },
    {
      "epoch": 2.4985105749180816,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.029943293692030453,
      "loss": 0.7856,
      "num_input_tokens_seen": 9733864,
      "step": 16775
    },
    {
      "epoch": 2.499255287459041,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029943124197747736,
      "loss": 0.8,
      "num_input_tokens_seen": 9736712,
      "step": 16780
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02994295445101612,
      "loss": 0.789,
      "num_input_tokens_seen": 9739400,
      "step": 16785
    },
    {
      "epoch": 2.500744712540959,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02994278445183848,
      "loss": 0.8034,
      "num_input_tokens_seen": 9742216,
      "step": 16790
    },
    {
      "epoch": 2.5014894250819184,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02994261420021768,
      "loss": 0.8108,
      "num_input_tokens_seen": 9745224,
      "step": 16795
    },
    {
      "epoch": 2.5022341376228776,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.0299424436961566,
      "loss": 0.8189,
      "num_input_tokens_seen": 9748328,
      "step": 16800
    },
    {
      "epoch": 2.502978850163837,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029942272939658125,
      "loss": 0.7749,
      "num_input_tokens_seen": 9751048,
      "step": 16805
    },
    {
      "epoch": 2.503723562704796,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02994210193072513,
      "loss": 0.8023,
      "num_input_tokens_seen": 9753704,
      "step": 16810
    },
    {
      "epoch": 2.504468275245755,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029941930669360517,
      "loss": 0.8241,
      "num_input_tokens_seen": 9756456,
      "step": 16815
    },
    {
      "epoch": 2.5052129877867144,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.02994175915556716,
      "loss": 0.8227,
      "num_input_tokens_seen": 9759400,
      "step": 16820
    },
    {
      "epoch": 2.5059577003276736,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.02994158738934798,
      "loss": 0.7885,
      "num_input_tokens_seen": 9762120,
      "step": 16825
    },
    {
      "epoch": 2.506702412868633,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02994141537070586,
      "loss": 0.7986,
      "num_input_tokens_seen": 9765096,
      "step": 16830
    },
    {
      "epoch": 2.507447125409592,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.029941243099643713,
      "loss": 0.805,
      "num_input_tokens_seen": 9767944,
      "step": 16835
    },
    {
      "epoch": 2.5081918379505512,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02994107057616445,
      "loss": 0.8096,
      "num_input_tokens_seen": 9770696,
      "step": 16840
    },
    {
      "epoch": 2.5089365504915104,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02994089780027099,
      "loss": 0.7869,
      "num_input_tokens_seen": 9773384,
      "step": 16845
    },
    {
      "epoch": 2.509681263032469,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029940724771966243,
      "loss": 0.8116,
      "num_input_tokens_seen": 9776232,
      "step": 16850
    },
    {
      "epoch": 2.510425975573429,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029940551491253138,
      "loss": 0.8147,
      "num_input_tokens_seen": 9778952,
      "step": 16855
    },
    {
      "epoch": 2.5111706881143876,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.029940377958134597,
      "loss": 0.7889,
      "num_input_tokens_seen": 9781864,
      "step": 16860
    },
    {
      "epoch": 2.5119154006553472,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.029940204172613557,
      "loss": 0.7977,
      "num_input_tokens_seen": 9784936,
      "step": 16865
    },
    {
      "epoch": 2.512660113196306,
      "grad_norm": 0.03125,
      "learning_rate": 0.029940030134692954,
      "loss": 0.8017,
      "num_input_tokens_seen": 9787976,
      "step": 16870
    },
    {
      "epoch": 2.5134048257372656,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.02993985584437572,
      "loss": 0.8194,
      "num_input_tokens_seen": 9790952,
      "step": 16875
    },
    {
      "epoch": 2.5141495382782244,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029939681301664814,
      "loss": 0.7904,
      "num_input_tokens_seen": 9793832,
      "step": 16880
    },
    {
      "epoch": 2.514894250819184,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02993950650656317,
      "loss": 0.7781,
      "num_input_tokens_seen": 9796872,
      "step": 16885
    },
    {
      "epoch": 2.515638963360143,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029939331459073754,
      "loss": 0.8145,
      "num_input_tokens_seen": 9799912,
      "step": 16890
    },
    {
      "epoch": 2.516383675901102,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02993915615919951,
      "loss": 0.7985,
      "num_input_tokens_seen": 9802696,
      "step": 16895
    },
    {
      "epoch": 2.517128388442061,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.02993898060694341,
      "loss": 0.7891,
      "num_input_tokens_seen": 9805384,
      "step": 16900
    },
    {
      "epoch": 2.5178731009830204,
      "grad_norm": 0.017578125,
      "learning_rate": 0.029938804802308414,
      "loss": 0.8216,
      "num_input_tokens_seen": 9808616,
      "step": 16905
    },
    {
      "epoch": 2.5186178135239796,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.029938628745297492,
      "loss": 0.8119,
      "num_input_tokens_seen": 9811720,
      "step": 16910
    },
    {
      "epoch": 2.519362526064939,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029938452435913625,
      "loss": 0.8043,
      "num_input_tokens_seen": 9814536,
      "step": 16915
    },
    {
      "epoch": 2.520107238605898,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029938275874159783,
      "loss": 0.8023,
      "num_input_tokens_seen": 9817256,
      "step": 16920
    },
    {
      "epoch": 2.5208519511468572,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.029938099060038956,
      "loss": 0.8392,
      "num_input_tokens_seen": 9820328,
      "step": 16925
    },
    {
      "epoch": 2.5215966636878164,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.029937921993554127,
      "loss": 0.7817,
      "num_input_tokens_seen": 9823304,
      "step": 16930
    },
    {
      "epoch": 2.5223413762287756,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029937744674708288,
      "loss": 0.8182,
      "num_input_tokens_seen": 9825992,
      "step": 16935
    },
    {
      "epoch": 2.523086088769735,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029937567103504432,
      "loss": 0.7777,
      "num_input_tokens_seen": 9828936,
      "step": 16940
    },
    {
      "epoch": 2.523830801310694,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029937389279945565,
      "loss": 0.7973,
      "num_input_tokens_seen": 9831656,
      "step": 16945
    },
    {
      "epoch": 2.5245755138516532,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029937211204034687,
      "loss": 0.7953,
      "num_input_tokens_seen": 9834632,
      "step": 16950
    },
    {
      "epoch": 2.5253202263926124,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.029937032875774806,
      "loss": 0.8352,
      "num_input_tokens_seen": 9837576,
      "step": 16955
    },
    {
      "epoch": 2.5260649389335716,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02993685429516894,
      "loss": 0.832,
      "num_input_tokens_seen": 9840616,
      "step": 16960
    },
    {
      "epoch": 2.526809651474531,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.029936675462220095,
      "loss": 0.8031,
      "num_input_tokens_seen": 9843720,
      "step": 16965
    },
    {
      "epoch": 2.52755436401549,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029936496376931306,
      "loss": 0.7964,
      "num_input_tokens_seen": 9846728,
      "step": 16970
    },
    {
      "epoch": 2.5282990765564493,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029936317039305584,
      "loss": 0.7992,
      "num_input_tokens_seen": 9849608,
      "step": 16975
    },
    {
      "epoch": 2.5290437890974085,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02993613744934597,
      "loss": 0.8183,
      "num_input_tokens_seen": 9852392,
      "step": 16980
    },
    {
      "epoch": 2.5297885016383677,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.029935957607055495,
      "loss": 0.8037,
      "num_input_tokens_seen": 9855048,
      "step": 16985
    },
    {
      "epoch": 2.530533214179327,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.029935777512437197,
      "loss": 0.7992,
      "num_input_tokens_seen": 9858056,
      "step": 16990
    },
    {
      "epoch": 2.531277926720286,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.029935597165494113,
      "loss": 0.8068,
      "num_input_tokens_seen": 9860776,
      "step": 16995
    },
    {
      "epoch": 2.5320226392612453,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029935416566229302,
      "loss": 0.7993,
      "num_input_tokens_seen": 9864008,
      "step": 17000
    },
    {
      "epoch": 2.5327673518022045,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029935235714645802,
      "loss": 0.7981,
      "num_input_tokens_seen": 9867144,
      "step": 17005
    },
    {
      "epoch": 2.5335120643431637,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.029935054610746678,
      "loss": 0.7931,
      "num_input_tokens_seen": 9869832,
      "step": 17010
    },
    {
      "epoch": 2.534256776884123,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029934873254534984,
      "loss": 0.8076,
      "num_input_tokens_seen": 9872936,
      "step": 17015
    },
    {
      "epoch": 2.535001489425082,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.029934691646013787,
      "loss": 0.795,
      "num_input_tokens_seen": 9875624,
      "step": 17020
    },
    {
      "epoch": 2.535746201966041,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02993450978518615,
      "loss": 0.7906,
      "num_input_tokens_seen": 9878472,
      "step": 17025
    },
    {
      "epoch": 2.5364909145070005,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.02993432767205515,
      "loss": 0.8217,
      "num_input_tokens_seen": 9881320,
      "step": 17030
    },
    {
      "epoch": 2.5372356270479592,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029934145306623867,
      "loss": 0.7784,
      "num_input_tokens_seen": 9884072,
      "step": 17035
    },
    {
      "epoch": 2.537980339588919,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.029933962688895373,
      "loss": 0.8016,
      "num_input_tokens_seen": 9886984,
      "step": 17040
    },
    {
      "epoch": 2.5387250521298776,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02993377981887276,
      "loss": 0.7762,
      "num_input_tokens_seen": 9889960,
      "step": 17045
    },
    {
      "epoch": 2.5394697646708373,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029933596696559112,
      "loss": 0.7947,
      "num_input_tokens_seen": 9892744,
      "step": 17050
    },
    {
      "epoch": 2.540214477211796,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.029933413321957526,
      "loss": 0.8124,
      "num_input_tokens_seen": 9895656,
      "step": 17055
    },
    {
      "epoch": 2.5409591897527557,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029933229695071104,
      "loss": 0.7973,
      "num_input_tokens_seen": 9898664,
      "step": 17060
    },
    {
      "epoch": 2.5417039022937145,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029933045815902937,
      "loss": 0.8093,
      "num_input_tokens_seen": 9901640,
      "step": 17065
    },
    {
      "epoch": 2.5424486148346737,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029932861684456146,
      "loss": 0.7979,
      "num_input_tokens_seen": 9904616,
      "step": 17070
    },
    {
      "epoch": 2.543193327375633,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02993267730073383,
      "loss": 0.8017,
      "num_input_tokens_seen": 9907240,
      "step": 17075
    },
    {
      "epoch": 2.543938039916592,
      "grad_norm": 0.017578125,
      "learning_rate": 0.029932492664739106,
      "loss": 0.7839,
      "num_input_tokens_seen": 9910440,
      "step": 17080
    },
    {
      "epoch": 2.5446827524575513,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.029932307776475098,
      "loss": 0.7823,
      "num_input_tokens_seen": 9913672,
      "step": 17085
    },
    {
      "epoch": 2.5454274649985105,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.029932122635944927,
      "loss": 0.7922,
      "num_input_tokens_seen": 9916584,
      "step": 17090
    },
    {
      "epoch": 2.5461721775394697,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029931937243151716,
      "loss": 0.8039,
      "num_input_tokens_seen": 9919432,
      "step": 17095
    },
    {
      "epoch": 2.546916890080429,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029931751598098605,
      "loss": 0.7881,
      "num_input_tokens_seen": 9922184,
      "step": 17100
    },
    {
      "epoch": 2.547661602621388,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.029931565700788726,
      "loss": 0.7894,
      "num_input_tokens_seen": 9925064,
      "step": 17105
    },
    {
      "epoch": 2.5484063151623473,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.029931379551225223,
      "loss": 0.791,
      "num_input_tokens_seen": 9928040,
      "step": 17110
    },
    {
      "epoch": 2.5491510277033065,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02993119314941123,
      "loss": 0.8215,
      "num_input_tokens_seen": 9930824,
      "step": 17115
    },
    {
      "epoch": 2.5498957402442657,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02993100649534991,
      "loss": 0.798,
      "num_input_tokens_seen": 9933480,
      "step": 17120
    },
    {
      "epoch": 2.550640452785225,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02993081958904441,
      "loss": 0.8479,
      "num_input_tokens_seen": 9936456,
      "step": 17125
    },
    {
      "epoch": 2.551385165326184,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02993063243049789,
      "loss": 0.8122,
      "num_input_tokens_seen": 9939368,
      "step": 17130
    },
    {
      "epoch": 2.5521298778671433,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.029930445019713505,
      "loss": 0.813,
      "num_input_tokens_seen": 9942216,
      "step": 17135
    },
    {
      "epoch": 2.5528745904081025,
      "grad_norm": 0.00982666015625,
      "learning_rate": 0.029930257356694433,
      "loss": 0.8017,
      "num_input_tokens_seen": 9944904,
      "step": 17140
    },
    {
      "epoch": 2.5536193029490617,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029930069441443833,
      "loss": 0.8024,
      "num_input_tokens_seen": 9947752,
      "step": 17145
    },
    {
      "epoch": 2.554364015490021,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029929881273964884,
      "loss": 0.7894,
      "num_input_tokens_seen": 9950536,
      "step": 17150
    },
    {
      "epoch": 2.55510872803098,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.029929692854260764,
      "loss": 0.802,
      "num_input_tokens_seen": 9953448,
      "step": 17155
    },
    {
      "epoch": 2.5558534405719393,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.029929504182334657,
      "loss": 0.8403,
      "num_input_tokens_seen": 9956424,
      "step": 17160
    },
    {
      "epoch": 2.5565981531128985,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.029929315258189752,
      "loss": 0.8411,
      "num_input_tokens_seen": 9959176,
      "step": 17165
    },
    {
      "epoch": 2.5573428656538577,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02992912608182924,
      "loss": 0.8176,
      "num_input_tokens_seen": 9962568,
      "step": 17170
    },
    {
      "epoch": 2.558087578194817,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029928936653256317,
      "loss": 0.795,
      "num_input_tokens_seen": 9965672,
      "step": 17175
    },
    {
      "epoch": 2.558832290735776,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02992874697247418,
      "loss": 0.7775,
      "num_input_tokens_seen": 9968520,
      "step": 17180
    },
    {
      "epoch": 2.5595770032767353,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029928557039486037,
      "loss": 0.8069,
      "num_input_tokens_seen": 9971240,
      "step": 17185
    },
    {
      "epoch": 2.5603217158176945,
      "grad_norm": 0.03125,
      "learning_rate": 0.029928366854295094,
      "loss": 0.7956,
      "num_input_tokens_seen": 9974120,
      "step": 17190
    },
    {
      "epoch": 2.5610664283586537,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.029928176416904564,
      "loss": 0.7713,
      "num_input_tokens_seen": 9977192,
      "step": 17195
    },
    {
      "epoch": 2.5618111408996125,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029927985727317672,
      "loss": 0.824,
      "num_input_tokens_seen": 9979848,
      "step": 17200
    },
    {
      "epoch": 2.562555853440572,
      "grad_norm": 0.01422119140625,
      "learning_rate": 0.02992779478553763,
      "loss": 0.8184,
      "num_input_tokens_seen": 9982984,
      "step": 17205
    },
    {
      "epoch": 2.563300565981531,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029927603591567663,
      "loss": 0.777,
      "num_input_tokens_seen": 9985960,
      "step": 17210
    },
    {
      "epoch": 2.5640452785224905,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02992741214541101,
      "loss": 0.7922,
      "num_input_tokens_seen": 9988872,
      "step": 17215
    },
    {
      "epoch": 2.5647899910634493,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.0299272204470709,
      "loss": 0.7879,
      "num_input_tokens_seen": 9991720,
      "step": 17220
    },
    {
      "epoch": 2.565534703604409,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.02992702849655057,
      "loss": 0.7918,
      "num_input_tokens_seen": 9994472,
      "step": 17225
    },
    {
      "epoch": 2.5662794161453677,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029926836293853266,
      "loss": 0.7902,
      "num_input_tokens_seen": 9997128,
      "step": 17230
    },
    {
      "epoch": 2.5670241286863273,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029926643838982234,
      "loss": 0.788,
      "num_input_tokens_seen": 9999848,
      "step": 17235
    },
    {
      "epoch": 2.567768841227286,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.029926451131940722,
      "loss": 0.8223,
      "num_input_tokens_seen": 10002664,
      "step": 17240
    },
    {
      "epoch": 2.5685135537682453,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029926258172731993,
      "loss": 0.8006,
      "num_input_tokens_seen": 10005320,
      "step": 17245
    },
    {
      "epoch": 2.5692582663092045,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.0299260649613593,
      "loss": 0.7806,
      "num_input_tokens_seen": 10008232,
      "step": 17250
    },
    {
      "epoch": 2.5700029788501637,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02992587149782591,
      "loss": 0.7903,
      "num_input_tokens_seen": 10011112,
      "step": 17255
    },
    {
      "epoch": 2.570747691391123,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.02992567778213509,
      "loss": 0.8121,
      "num_input_tokens_seen": 10014312,
      "step": 17260
    },
    {
      "epoch": 2.571492403932082,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029925483814290114,
      "loss": 0.82,
      "num_input_tokens_seen": 10017320,
      "step": 17265
    },
    {
      "epoch": 2.5722371164730413,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02992528959429426,
      "loss": 0.7837,
      "num_input_tokens_seen": 10020520,
      "step": 17270
    },
    {
      "epoch": 2.5729818290140005,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029925095122150806,
      "loss": 0.8268,
      "num_input_tokens_seen": 10023720,
      "step": 17275
    },
    {
      "epoch": 2.5737265415549597,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029924900397863042,
      "loss": 0.7686,
      "num_input_tokens_seen": 10026760,
      "step": 17280
    },
    {
      "epoch": 2.574471254095919,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.029924705421434255,
      "loss": 0.8071,
      "num_input_tokens_seen": 10029640,
      "step": 17285
    },
    {
      "epoch": 2.575215966636878,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029924510192867734,
      "loss": 0.7941,
      "num_input_tokens_seen": 10032744,
      "step": 17290
    },
    {
      "epoch": 2.5759606791778373,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.029924314712166784,
      "loss": 0.8061,
      "num_input_tokens_seen": 10035816,
      "step": 17295
    },
    {
      "epoch": 2.5767053917187965,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.029924118979334707,
      "loss": 0.8363,
      "num_input_tokens_seen": 10038792,
      "step": 17300
    },
    {
      "epoch": 2.5774501042597557,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.029923922994374803,
      "loss": 0.8135,
      "num_input_tokens_seen": 10042056,
      "step": 17305
    },
    {
      "epoch": 2.578194816800715,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029923726757290395,
      "loss": 0.7925,
      "num_input_tokens_seen": 10044968,
      "step": 17310
    },
    {
      "epoch": 2.578939529341674,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.02992353026808479,
      "loss": 0.7851,
      "num_input_tokens_seen": 10047816,
      "step": 17315
    },
    {
      "epoch": 2.5796842418826333,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.029923333526761305,
      "loss": 0.8232,
      "num_input_tokens_seen": 10050728,
      "step": 17320
    },
    {
      "epoch": 2.5804289544235925,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029923136533323267,
      "loss": 0.8323,
      "num_input_tokens_seen": 10053736,
      "step": 17325
    },
    {
      "epoch": 2.5811736669645517,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02992293928777401,
      "loss": 0.7752,
      "num_input_tokens_seen": 10056424,
      "step": 17330
    },
    {
      "epoch": 2.581918379505511,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.029922741790116857,
      "loss": 0.7876,
      "num_input_tokens_seen": 10059496,
      "step": 17335
    },
    {
      "epoch": 2.58266309204647,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029922544040355145,
      "loss": 0.782,
      "num_input_tokens_seen": 10062248,
      "step": 17340
    },
    {
      "epoch": 2.5834078045874294,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.029922346038492226,
      "loss": 0.7964,
      "num_input_tokens_seen": 10065128,
      "step": 17345
    },
    {
      "epoch": 2.5841525171283886,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.029922147784531433,
      "loss": 0.8199,
      "num_input_tokens_seen": 10068072,
      "step": 17350
    },
    {
      "epoch": 2.5848972296693478,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02992194927847612,
      "loss": 0.7643,
      "num_input_tokens_seen": 10071080,
      "step": 17355
    },
    {
      "epoch": 2.585641942210307,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.029921750520329637,
      "loss": 0.8004,
      "num_input_tokens_seen": 10073832,
      "step": 17360
    },
    {
      "epoch": 2.586386654751266,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.029921551510095346,
      "loss": 0.8302,
      "num_input_tokens_seen": 10076840,
      "step": 17365
    },
    {
      "epoch": 2.5871313672922254,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029921352247776607,
      "loss": 0.803,
      "num_input_tokens_seen": 10079560,
      "step": 17370
    },
    {
      "epoch": 2.587876079833184,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029921152733376793,
      "loss": 0.8034,
      "num_input_tokens_seen": 10082568,
      "step": 17375
    },
    {
      "epoch": 2.5886207923741438,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.029920952966899263,
      "loss": 0.8062,
      "num_input_tokens_seen": 10085352,
      "step": 17380
    },
    {
      "epoch": 2.5893655049151025,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0299207529483474,
      "loss": 0.8043,
      "num_input_tokens_seen": 10088232,
      "step": 17385
    },
    {
      "epoch": 2.590110217456062,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02992055267772458,
      "loss": 0.8291,
      "num_input_tokens_seen": 10091272,
      "step": 17390
    },
    {
      "epoch": 2.590854929997021,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029920352155034188,
      "loss": 0.8124,
      "num_input_tokens_seen": 10094088,
      "step": 17395
    },
    {
      "epoch": 2.5915996425379806,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02992015138027961,
      "loss": 0.8167,
      "num_input_tokens_seen": 10096808,
      "step": 17400
    },
    {
      "epoch": 2.5923443550789393,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029919950353464238,
      "loss": 0.7951,
      "num_input_tokens_seen": 10099752,
      "step": 17405
    },
    {
      "epoch": 2.593089067619899,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.02991974907459147,
      "loss": 0.8123,
      "num_input_tokens_seen": 10102632,
      "step": 17410
    },
    {
      "epoch": 2.5938337801608577,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.0299195475436647,
      "loss": 0.8044,
      "num_input_tokens_seen": 10105480,
      "step": 17415
    },
    {
      "epoch": 2.594578492701817,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029919345760687343,
      "loss": 0.784,
      "num_input_tokens_seen": 10108616,
      "step": 17420
    },
    {
      "epoch": 2.595323205242776,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029919143725662806,
      "loss": 0.7971,
      "num_input_tokens_seen": 10111560,
      "step": 17425
    },
    {
      "epoch": 2.5960679177837354,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02991894143859449,
      "loss": 0.8103,
      "num_input_tokens_seen": 10114824,
      "step": 17430
    },
    {
      "epoch": 2.5968126303246946,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02991873889948583,
      "loss": 0.8143,
      "num_input_tokens_seen": 10117768,
      "step": 17435
    },
    {
      "epoch": 2.5975573428656538,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029918536108340235,
      "loss": 0.8001,
      "num_input_tokens_seen": 10121064,
      "step": 17440
    },
    {
      "epoch": 2.598302055406613,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.029918333065161132,
      "loss": 0.7822,
      "num_input_tokens_seen": 10123912,
      "step": 17445
    },
    {
      "epoch": 2.599046767947572,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029918129769951958,
      "loss": 0.7879,
      "num_input_tokens_seen": 10126760,
      "step": 17450
    },
    {
      "epoch": 2.5997914804885314,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029917926222716145,
      "loss": 0.7852,
      "num_input_tokens_seen": 10129512,
      "step": 17455
    },
    {
      "epoch": 2.6005361930294906,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.029917722423457128,
      "loss": 0.8028,
      "num_input_tokens_seen": 10132424,
      "step": 17460
    },
    {
      "epoch": 2.6012809055704498,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02991751837217835,
      "loss": 0.814,
      "num_input_tokens_seen": 10135304,
      "step": 17465
    },
    {
      "epoch": 2.602025618111409,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02991731406888326,
      "loss": 0.8053,
      "num_input_tokens_seen": 10137864,
      "step": 17470
    },
    {
      "epoch": 2.602770330652368,
      "grad_norm": 0.01171875,
      "learning_rate": 0.029917109513575315,
      "loss": 0.7927,
      "num_input_tokens_seen": 10140648,
      "step": 17475
    },
    {
      "epoch": 2.6035150431933274,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029916904706257963,
      "loss": 0.8163,
      "num_input_tokens_seen": 10143464,
      "step": 17480
    },
    {
      "epoch": 2.6042597557342866,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.02991669964693467,
      "loss": 0.8095,
      "num_input_tokens_seen": 10146536,
      "step": 17485
    },
    {
      "epoch": 2.605004468275246,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029916494335608897,
      "loss": 0.7917,
      "num_input_tokens_seen": 10149480,
      "step": 17490
    },
    {
      "epoch": 2.605749180816205,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02991628877228411,
      "loss": 0.7902,
      "num_input_tokens_seen": 10152456,
      "step": 17495
    },
    {
      "epoch": 2.606493893357164,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029916082956963786,
      "loss": 0.7972,
      "num_input_tokens_seen": 10155304,
      "step": 17500
    },
    {
      "epoch": 2.6072386058981234,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.0299158768896514,
      "loss": 0.8045,
      "num_input_tokens_seen": 10158312,
      "step": 17505
    },
    {
      "epoch": 2.6079833184390826,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029915670570350432,
      "loss": 0.8167,
      "num_input_tokens_seen": 10161128,
      "step": 17510
    },
    {
      "epoch": 2.608728030980042,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.029915463999064373,
      "loss": 0.8047,
      "num_input_tokens_seen": 10163880,
      "step": 17515
    },
    {
      "epoch": 2.609472743521001,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.029915257175796708,
      "loss": 0.7931,
      "num_input_tokens_seen": 10166568,
      "step": 17520
    },
    {
      "epoch": 2.61021745606196,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.02991505010055093,
      "loss": 0.794,
      "num_input_tokens_seen": 10169512,
      "step": 17525
    },
    {
      "epoch": 2.6109621686029194,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02991484277333054,
      "loss": 0.7796,
      "num_input_tokens_seen": 10172264,
      "step": 17530
    },
    {
      "epoch": 2.6117068811438786,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02991463519413904,
      "loss": 0.8038,
      "num_input_tokens_seen": 10175112,
      "step": 17535
    },
    {
      "epoch": 2.612451593684838,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029914427362979935,
      "loss": 0.7948,
      "num_input_tokens_seen": 10177928,
      "step": 17540
    },
    {
      "epoch": 2.613196306225797,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.02991421927985674,
      "loss": 0.7834,
      "num_input_tokens_seen": 10180840,
      "step": 17545
    },
    {
      "epoch": 2.6139410187667558,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.02991401094477297,
      "loss": 0.7958,
      "num_input_tokens_seen": 10183624,
      "step": 17550
    },
    {
      "epoch": 2.6146857313077154,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.029913802357732142,
      "loss": 0.785,
      "num_input_tokens_seen": 10186728,
      "step": 17555
    },
    {
      "epoch": 2.615430443848674,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02991359351873778,
      "loss": 0.8288,
      "num_input_tokens_seen": 10189512,
      "step": 17560
    },
    {
      "epoch": 2.616175156389634,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029913384427793416,
      "loss": 0.8524,
      "num_input_tokens_seen": 10192104,
      "step": 17565
    },
    {
      "epoch": 2.6169198689305926,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029913175084902578,
      "loss": 0.7943,
      "num_input_tokens_seen": 10194568,
      "step": 17570
    },
    {
      "epoch": 2.6176645814715522,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.0299129654900688,
      "loss": 0.8072,
      "num_input_tokens_seen": 10197224,
      "step": 17575
    },
    {
      "epoch": 2.618409294012511,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02991275564329563,
      "loss": 0.7918,
      "num_input_tokens_seen": 10200168,
      "step": 17580
    },
    {
      "epoch": 2.6191540065534706,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.029912545544586614,
      "loss": 0.817,
      "num_input_tokens_seen": 10202792,
      "step": 17585
    },
    {
      "epoch": 2.6198987190944294,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029912335193945296,
      "loss": 0.8219,
      "num_input_tokens_seen": 10205640,
      "step": 17590
    },
    {
      "epoch": 2.6206434316353886,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029912124591375225,
      "loss": 0.805,
      "num_input_tokens_seen": 10208200,
      "step": 17595
    },
    {
      "epoch": 2.621388144176348,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02991191373687997,
      "loss": 0.7984,
      "num_input_tokens_seen": 10210920,
      "step": 17600
    },
    {
      "epoch": 2.622132856717307,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.029911702630463086,
      "loss": 0.8189,
      "num_input_tokens_seen": 10213960,
      "step": 17605
    },
    {
      "epoch": 2.622877569258266,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.029911491272128148,
      "loss": 0.8096,
      "num_input_tokens_seen": 10216616,
      "step": 17610
    },
    {
      "epoch": 2.6236222817992254,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.02991127966187871,
      "loss": 0.8124,
      "num_input_tokens_seen": 10219336,
      "step": 17615
    },
    {
      "epoch": 2.6243669943401846,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02991106779971836,
      "loss": 0.8058,
      "num_input_tokens_seen": 10222120,
      "step": 17620
    },
    {
      "epoch": 2.625111706881144,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029910855685650682,
      "loss": 0.835,
      "num_input_tokens_seen": 10224904,
      "step": 17625
    },
    {
      "epoch": 2.625856419422103,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.029910643319679242,
      "loss": 0.7908,
      "num_input_tokens_seen": 10227816,
      "step": 17630
    },
    {
      "epoch": 2.626601131963062,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02991043070180765,
      "loss": 0.7938,
      "num_input_tokens_seen": 10230568,
      "step": 17635
    },
    {
      "epoch": 2.6273458445040214,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029910217832039475,
      "loss": 0.7795,
      "num_input_tokens_seen": 10233960,
      "step": 17640
    },
    {
      "epoch": 2.6280905570449806,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.029910004710378323,
      "loss": 0.8,
      "num_input_tokens_seen": 10236776,
      "step": 17645
    },
    {
      "epoch": 2.62883526958594,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.0299097913368278,
      "loss": 0.7975,
      "num_input_tokens_seen": 10240072,
      "step": 17650
    },
    {
      "epoch": 2.629579982126899,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.029909577711391508,
      "loss": 0.806,
      "num_input_tokens_seen": 10242920,
      "step": 17655
    },
    {
      "epoch": 2.6303246946678582,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029909363834073048,
      "loss": 0.8191,
      "num_input_tokens_seen": 10246408,
      "step": 17660
    },
    {
      "epoch": 2.6310694072088174,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029909149704876046,
      "loss": 0.8059,
      "num_input_tokens_seen": 10249256,
      "step": 17665
    },
    {
      "epoch": 2.6318141197497766,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029908935323804106,
      "loss": 0.812,
      "num_input_tokens_seen": 10252264,
      "step": 17670
    },
    {
      "epoch": 2.632558832290736,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.029908720690860858,
      "loss": 0.7891,
      "num_input_tokens_seen": 10255304,
      "step": 17675
    },
    {
      "epoch": 2.633303544831695,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02990850580604993,
      "loss": 0.7974,
      "num_input_tokens_seen": 10258120,
      "step": 17680
    },
    {
      "epoch": 2.6340482573726542,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029908290669374945,
      "loss": 0.8008,
      "num_input_tokens_seen": 10260872,
      "step": 17685
    },
    {
      "epoch": 2.6347929699136134,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029908075280839545,
      "loss": 0.7959,
      "num_input_tokens_seen": 10263752,
      "step": 17690
    },
    {
      "epoch": 2.6355376824545726,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02990785964044736,
      "loss": 0.7935,
      "num_input_tokens_seen": 10266824,
      "step": 17695
    },
    {
      "epoch": 2.636282394995532,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02990764374820204,
      "loss": 0.8007,
      "num_input_tokens_seen": 10269960,
      "step": 17700
    },
    {
      "epoch": 2.637027107536491,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029907427604107233,
      "loss": 0.8042,
      "num_input_tokens_seen": 10272968,
      "step": 17705
    },
    {
      "epoch": 2.6377718200774503,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029907211208166584,
      "loss": 0.82,
      "num_input_tokens_seen": 10276104,
      "step": 17710
    },
    {
      "epoch": 2.6385165326184095,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02990699456038376,
      "loss": 0.8047,
      "num_input_tokens_seen": 10278888,
      "step": 17715
    },
    {
      "epoch": 2.6392612451593687,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02990677766076241,
      "loss": 0.8335,
      "num_input_tokens_seen": 10281832,
      "step": 17720
    },
    {
      "epoch": 2.6400059577003274,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029906560509306197,
      "loss": 0.7961,
      "num_input_tokens_seen": 10284680,
      "step": 17725
    },
    {
      "epoch": 2.640750670241287,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0299063431060188,
      "loss": 0.8291,
      "num_input_tokens_seen": 10287912,
      "step": 17730
    },
    {
      "epoch": 2.641495382782246,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.029906125450903882,
      "loss": 0.8202,
      "num_input_tokens_seen": 10290856,
      "step": 17735
    },
    {
      "epoch": 2.6422400953232055,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.02990590754396513,
      "loss": 0.7891,
      "num_input_tokens_seen": 10294024,
      "step": 17740
    },
    {
      "epoch": 2.6429848078641642,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.029905689385206213,
      "loss": 0.7976,
      "num_input_tokens_seen": 10296904,
      "step": 17745
    },
    {
      "epoch": 2.643729520405124,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029905470974630824,
      "loss": 0.7976,
      "num_input_tokens_seen": 10299656,
      "step": 17750
    },
    {
      "epoch": 2.6444742329460826,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02990525231224266,
      "loss": 0.7788,
      "num_input_tokens_seen": 10302568,
      "step": 17755
    },
    {
      "epoch": 2.645218945487042,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.0299050333980454,
      "loss": 0.8077,
      "num_input_tokens_seen": 10305416,
      "step": 17760
    },
    {
      "epoch": 2.645963658028001,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02990481423204275,
      "loss": 0.8301,
      "num_input_tokens_seen": 10308360,
      "step": 17765
    },
    {
      "epoch": 2.6467083705689602,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029904594814238412,
      "loss": 0.7963,
      "num_input_tokens_seen": 10311208,
      "step": 17770
    },
    {
      "epoch": 2.6474530831099194,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.0299043751446361,
      "loss": 0.7939,
      "num_input_tokens_seen": 10314280,
      "step": 17775
    },
    {
      "epoch": 2.6481977956508786,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029904155223239512,
      "loss": 0.8053,
      "num_input_tokens_seen": 10317384,
      "step": 17780
    },
    {
      "epoch": 2.648942508191838,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.029903935050052368,
      "loss": 0.7968,
      "num_input_tokens_seen": 10320200,
      "step": 17785
    },
    {
      "epoch": 2.649687220732797,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.029903714625078393,
      "loss": 0.7937,
      "num_input_tokens_seen": 10323464,
      "step": 17790
    },
    {
      "epoch": 2.6504319332737563,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.029903493948321303,
      "loss": 0.8237,
      "num_input_tokens_seen": 10326408,
      "step": 17795
    },
    {
      "epoch": 2.6511766458147155,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029903273019784835,
      "loss": 0.7951,
      "num_input_tokens_seen": 10329384,
      "step": 17800
    },
    {
      "epoch": 2.6519213583556747,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02990305183947271,
      "loss": 0.7916,
      "num_input_tokens_seen": 10332520,
      "step": 17805
    },
    {
      "epoch": 2.652666070896634,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029902830407388677,
      "loss": 0.8075,
      "num_input_tokens_seen": 10335496,
      "step": 17810
    },
    {
      "epoch": 2.653410783437593,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.029902608723536468,
      "loss": 0.8075,
      "num_input_tokens_seen": 10338376,
      "step": 17815
    },
    {
      "epoch": 2.6541554959785523,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02990238678791983,
      "loss": 0.8073,
      "num_input_tokens_seen": 10341288,
      "step": 17820
    },
    {
      "epoch": 2.6549002085195115,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.029902164600542517,
      "loss": 0.7993,
      "num_input_tokens_seen": 10344552,
      "step": 17825
    },
    {
      "epoch": 2.6556449210604707,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.029901942161408273,
      "loss": 0.8086,
      "num_input_tokens_seen": 10347528,
      "step": 17830
    },
    {
      "epoch": 2.65638963360143,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.029901719470520866,
      "loss": 0.7914,
      "num_input_tokens_seen": 10350344,
      "step": 17835
    },
    {
      "epoch": 2.657134346142389,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029901496527884056,
      "loss": 0.7988,
      "num_input_tokens_seen": 10353256,
      "step": 17840
    },
    {
      "epoch": 2.6578790586833483,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029901273333501604,
      "loss": 0.8074,
      "num_input_tokens_seen": 10356040,
      "step": 17845
    },
    {
      "epoch": 2.6586237712243075,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029901049887377285,
      "loss": 0.7911,
      "num_input_tokens_seen": 10359144,
      "step": 17850
    },
    {
      "epoch": 2.6593684837652667,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02990082618951487,
      "loss": 0.8136,
      "num_input_tokens_seen": 10362440,
      "step": 17855
    },
    {
      "epoch": 2.660113196306226,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029900602239918146,
      "loss": 0.7835,
      "num_input_tokens_seen": 10365000,
      "step": 17860
    },
    {
      "epoch": 2.660857908847185,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.02990037803859089,
      "loss": 0.7956,
      "num_input_tokens_seen": 10368040,
      "step": 17865
    },
    {
      "epoch": 2.6616026213881443,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02990015358553689,
      "loss": 0.817,
      "num_input_tokens_seen": 10370792,
      "step": 17870
    },
    {
      "epoch": 2.6623473339291035,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029899928880759936,
      "loss": 0.8067,
      "num_input_tokens_seen": 10373352,
      "step": 17875
    },
    {
      "epoch": 2.6630920464700627,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029899703924263832,
      "loss": 0.7934,
      "num_input_tokens_seen": 10376296,
      "step": 17880
    },
    {
      "epoch": 2.663836759011022,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02989947871605237,
      "loss": 0.7989,
      "num_input_tokens_seen": 10378984,
      "step": 17885
    },
    {
      "epoch": 2.6645814715519807,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02989925325612936,
      "loss": 0.8134,
      "num_input_tokens_seen": 10382120,
      "step": 17890
    },
    {
      "epoch": 2.6653261840929403,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02989902754449861,
      "loss": 0.8159,
      "num_input_tokens_seen": 10384872,
      "step": 17895
    },
    {
      "epoch": 2.666070896633899,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.029898801581163932,
      "loss": 0.8326,
      "num_input_tokens_seen": 10387560,
      "step": 17900
    },
    {
      "epoch": 2.6668156091748587,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029898575366129145,
      "loss": 0.7918,
      "num_input_tokens_seen": 10390440,
      "step": 17905
    },
    {
      "epoch": 2.6675603217158175,
      "grad_norm": 0.025390625,
      "learning_rate": 0.029898348899398065,
      "loss": 0.796,
      "num_input_tokens_seen": 10393672,
      "step": 17910
    },
    {
      "epoch": 2.668305034256777,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029898122180974526,
      "loss": 0.7971,
      "num_input_tokens_seen": 10396424,
      "step": 17915
    },
    {
      "epoch": 2.669049746797736,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029897895210862356,
      "loss": 0.7772,
      "num_input_tokens_seen": 10399368,
      "step": 17920
    },
    {
      "epoch": 2.6697944593386955,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029897667989065384,
      "loss": 0.8152,
      "num_input_tokens_seen": 10402120,
      "step": 17925
    },
    {
      "epoch": 2.6705391718796543,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029897440515587455,
      "loss": 0.8391,
      "num_input_tokens_seen": 10404712,
      "step": 17930
    },
    {
      "epoch": 2.6712838844206135,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.029897212790432406,
      "loss": 0.83,
      "num_input_tokens_seen": 10407656,
      "step": 17935
    },
    {
      "epoch": 2.6720285969615727,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029896984813604092,
      "loss": 0.8044,
      "num_input_tokens_seen": 10410472,
      "step": 17940
    },
    {
      "epoch": 2.672773309502532,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029896756585106363,
      "loss": 0.8108,
      "num_input_tokens_seen": 10413352,
      "step": 17945
    },
    {
      "epoch": 2.673518022043491,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029896528104943067,
      "loss": 0.7968,
      "num_input_tokens_seen": 10416456,
      "step": 17950
    },
    {
      "epoch": 2.6742627345844503,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.029896299373118065,
      "loss": 0.7994,
      "num_input_tokens_seen": 10419240,
      "step": 17955
    },
    {
      "epoch": 2.6750074471254095,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.02989607038963523,
      "loss": 0.7955,
      "num_input_tokens_seen": 10422696,
      "step": 17960
    },
    {
      "epoch": 2.6757521596663687,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029895841154498424,
      "loss": 0.8106,
      "num_input_tokens_seen": 10425672,
      "step": 17965
    },
    {
      "epoch": 2.676496872207328,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029895611667711524,
      "loss": 0.7987,
      "num_input_tokens_seen": 10428744,
      "step": 17970
    },
    {
      "epoch": 2.677241584748287,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.0298953819292784,
      "loss": 0.801,
      "num_input_tokens_seen": 10431816,
      "step": 17975
    },
    {
      "epoch": 2.6779862972892463,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029895151939202943,
      "loss": 0.8057,
      "num_input_tokens_seen": 10434696,
      "step": 17980
    },
    {
      "epoch": 2.6787310098302055,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02989492169748903,
      "loss": 0.7969,
      "num_input_tokens_seen": 10437640,
      "step": 17985
    },
    {
      "epoch": 2.6794757223711647,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029894691204140553,
      "loss": 0.7911,
      "num_input_tokens_seen": 10440392,
      "step": 17990
    },
    {
      "epoch": 2.680220434912124,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029894460459161405,
      "loss": 0.799,
      "num_input_tokens_seen": 10443368,
      "step": 17995
    },
    {
      "epoch": 2.680965147453083,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02989422946255549,
      "loss": 0.8204,
      "num_input_tokens_seen": 10446216,
      "step": 18000
    },
    {
      "epoch": 2.6817098599940423,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.0298939982143267,
      "loss": 0.8216,
      "num_input_tokens_seen": 10448936,
      "step": 18005
    },
    {
      "epoch": 2.6824545725350015,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.02989376671447895,
      "loss": 0.8164,
      "num_input_tokens_seen": 10451880,
      "step": 18010
    },
    {
      "epoch": 2.6831992850759607,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02989353496301615,
      "loss": 0.7926,
      "num_input_tokens_seen": 10454792,
      "step": 18015
    },
    {
      "epoch": 2.68394399761692,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02989330295994221,
      "loss": 0.7914,
      "num_input_tokens_seen": 10457480,
      "step": 18020
    },
    {
      "epoch": 2.684688710157879,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.029893070705261055,
      "loss": 0.8005,
      "num_input_tokens_seen": 10460360,
      "step": 18025
    },
    {
      "epoch": 2.6854334226988383,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02989283819897661,
      "loss": 0.7992,
      "num_input_tokens_seen": 10462920,
      "step": 18030
    },
    {
      "epoch": 2.6861781352397975,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0298926054410928,
      "loss": 0.8065,
      "num_input_tokens_seen": 10465416,
      "step": 18035
    },
    {
      "epoch": 2.6869228477807567,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029892372431613558,
      "loss": 0.7889,
      "num_input_tokens_seen": 10468136,
      "step": 18040
    },
    {
      "epoch": 2.687667560321716,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.029892139170542814,
      "loss": 0.8284,
      "num_input_tokens_seen": 10470952,
      "step": 18045
    },
    {
      "epoch": 2.688412272862675,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.029891905657884518,
      "loss": 0.7876,
      "num_input_tokens_seen": 10473832,
      "step": 18050
    },
    {
      "epoch": 2.6891569854036343,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029891671893642613,
      "loss": 0.8104,
      "num_input_tokens_seen": 10476744,
      "step": 18055
    },
    {
      "epoch": 2.6899016979445936,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.029891437877821045,
      "loss": 0.7988,
      "num_input_tokens_seen": 10479432,
      "step": 18060
    },
    {
      "epoch": 2.6906464104855523,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02989120361042377,
      "loss": 0.8023,
      "num_input_tokens_seen": 10482152,
      "step": 18065
    },
    {
      "epoch": 2.691391123026512,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.02989096909145474,
      "loss": 0.8282,
      "num_input_tokens_seen": 10485128,
      "step": 18070
    },
    {
      "epoch": 2.6921358355674707,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.02989073432091793,
      "loss": 0.8022,
      "num_input_tokens_seen": 10488296,
      "step": 18075
    },
    {
      "epoch": 2.6928805481084304,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.029890499298817288,
      "loss": 0.8,
      "num_input_tokens_seen": 10490984,
      "step": 18080
    },
    {
      "epoch": 2.693625260649389,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029890264025156803,
      "loss": 0.8079,
      "num_input_tokens_seen": 10493832,
      "step": 18085
    },
    {
      "epoch": 2.6943699731903488,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.02989002849994044,
      "loss": 0.7956,
      "num_input_tokens_seen": 10496616,
      "step": 18090
    },
    {
      "epoch": 2.6951146857313075,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029889792723172177,
      "loss": 0.7837,
      "num_input_tokens_seen": 10499400,
      "step": 18095
    },
    {
      "epoch": 2.695859398272267,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029889556694855997,
      "loss": 0.7968,
      "num_input_tokens_seen": 10502248,
      "step": 18100
    },
    {
      "epoch": 2.696604110813226,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029889320414995892,
      "loss": 0.7827,
      "num_input_tokens_seen": 10504936,
      "step": 18105
    },
    {
      "epoch": 2.697348823354185,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02988908388359585,
      "loss": 0.8388,
      "num_input_tokens_seen": 10507944,
      "step": 18110
    },
    {
      "epoch": 2.6980935358951443,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02988884710065987,
      "loss": 0.8176,
      "num_input_tokens_seen": 10510696,
      "step": 18115
    },
    {
      "epoch": 2.6988382484361035,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029888610066191947,
      "loss": 0.7858,
      "num_input_tokens_seen": 10513576,
      "step": 18120
    },
    {
      "epoch": 2.6995829609770627,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029888372780196095,
      "loss": 0.8063,
      "num_input_tokens_seen": 10516520,
      "step": 18125
    },
    {
      "epoch": 2.700327673518022,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.029888135242676316,
      "loss": 0.7998,
      "num_input_tokens_seen": 10519400,
      "step": 18130
    },
    {
      "epoch": 2.701072386058981,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02988789745363662,
      "loss": 0.8061,
      "num_input_tokens_seen": 10522696,
      "step": 18135
    },
    {
      "epoch": 2.7018170985999403,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02988765941308103,
      "loss": 0.8013,
      "num_input_tokens_seen": 10525736,
      "step": 18140
    },
    {
      "epoch": 2.7025618111408996,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029887421121013565,
      "loss": 0.81,
      "num_input_tokens_seen": 10528808,
      "step": 18145
    },
    {
      "epoch": 2.7033065236818588,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02988718257743825,
      "loss": 0.7743,
      "num_input_tokens_seen": 10531880,
      "step": 18150
    },
    {
      "epoch": 2.704051236222818,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.029886943782359114,
      "loss": 0.8134,
      "num_input_tokens_seen": 10534696,
      "step": 18155
    },
    {
      "epoch": 2.704795948763777,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029886704735780197,
      "loss": 0.8315,
      "num_input_tokens_seen": 10537608,
      "step": 18160
    },
    {
      "epoch": 2.7055406613047364,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029886465437705536,
      "loss": 0.7939,
      "num_input_tokens_seen": 10540040,
      "step": 18165
    },
    {
      "epoch": 2.7062853738456956,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029886225888139165,
      "loss": 0.7995,
      "num_input_tokens_seen": 10542632,
      "step": 18170
    },
    {
      "epoch": 2.7070300863866548,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.029885986087085144,
      "loss": 0.7958,
      "num_input_tokens_seen": 10545512,
      "step": 18175
    },
    {
      "epoch": 2.707774798927614,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029885746034547512,
      "loss": 0.7994,
      "num_input_tokens_seen": 10548200,
      "step": 18180
    },
    {
      "epoch": 2.708519511468573,
      "grad_norm": 0.0123291015625,
      "learning_rate": 0.02988550573053033,
      "loss": 0.8051,
      "num_input_tokens_seen": 10551016,
      "step": 18185
    },
    {
      "epoch": 2.7092642240095324,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029885265175037663,
      "loss": 0.8061,
      "num_input_tokens_seen": 10553416,
      "step": 18190
    },
    {
      "epoch": 2.7100089365504916,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.029885024368073564,
      "loss": 0.8079,
      "num_input_tokens_seen": 10556456,
      "step": 18195
    },
    {
      "epoch": 2.710753649091451,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.029884783309642112,
      "loss": 0.815,
      "num_input_tokens_seen": 10559528,
      "step": 18200
    },
    {
      "epoch": 2.71149836163241,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02988454199974737,
      "loss": 0.7941,
      "num_input_tokens_seen": 10562664,
      "step": 18205
    },
    {
      "epoch": 2.712243074173369,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02988430043839342,
      "loss": 0.8025,
      "num_input_tokens_seen": 10565576,
      "step": 18210
    },
    {
      "epoch": 2.7129877867143284,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.029884058625584343,
      "loss": 0.7844,
      "num_input_tokens_seen": 10568360,
      "step": 18215
    },
    {
      "epoch": 2.7137324992552876,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.029883816561324224,
      "loss": 0.8247,
      "num_input_tokens_seen": 10571432,
      "step": 18220
    },
    {
      "epoch": 2.714477211796247,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029883574245617148,
      "loss": 0.7998,
      "num_input_tokens_seen": 10574600,
      "step": 18225
    },
    {
      "epoch": 2.715221924337206,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02988333167846721,
      "loss": 0.8152,
      "num_input_tokens_seen": 10577672,
      "step": 18230
    },
    {
      "epoch": 2.715966636878165,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.029883088859878514,
      "loss": 0.7996,
      "num_input_tokens_seen": 10580392,
      "step": 18235
    },
    {
      "epoch": 2.716711349419124,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02988284578985516,
      "loss": 0.7824,
      "num_input_tokens_seen": 10583304,
      "step": 18240
    },
    {
      "epoch": 2.7174560619600836,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029882602468401246,
      "loss": 0.8068,
      "num_input_tokens_seen": 10585960,
      "step": 18245
    },
    {
      "epoch": 2.7182007745010424,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029882358895520895,
      "loss": 0.8159,
      "num_input_tokens_seen": 10588648,
      "step": 18250
    },
    {
      "epoch": 2.718945487042002,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.029882115071218216,
      "loss": 0.7936,
      "num_input_tokens_seen": 10591560,
      "step": 18255
    },
    {
      "epoch": 2.7196901995829608,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029881870995497325,
      "loss": 0.792,
      "num_input_tokens_seen": 10594248,
      "step": 18260
    },
    {
      "epoch": 2.7204349121239204,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029881626668362353,
      "loss": 0.8112,
      "num_input_tokens_seen": 10597064,
      "step": 18265
    },
    {
      "epoch": 2.721179624664879,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029881382089817422,
      "loss": 0.8005,
      "num_input_tokens_seen": 10599944,
      "step": 18270
    },
    {
      "epoch": 2.721924337205839,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.029881137259866665,
      "loss": 0.8142,
      "num_input_tokens_seen": 10602760,
      "step": 18275
    },
    {
      "epoch": 2.7226690497467976,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02988089217851422,
      "loss": 0.8028,
      "num_input_tokens_seen": 10606184,
      "step": 18280
    },
    {
      "epoch": 2.723413762287757,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02988064684576422,
      "loss": 0.8041,
      "num_input_tokens_seen": 10609064,
      "step": 18285
    },
    {
      "epoch": 2.724158474828716,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029880401261620823,
      "loss": 0.8094,
      "num_input_tokens_seen": 10611912,
      "step": 18290
    },
    {
      "epoch": 2.724903187369675,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.029880155426088167,
      "loss": 0.7962,
      "num_input_tokens_seen": 10614984,
      "step": 18295
    },
    {
      "epoch": 2.7256478999106344,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02987990933917041,
      "loss": 0.7907,
      "num_input_tokens_seen": 10617640,
      "step": 18300
    },
    {
      "epoch": 2.7263926124515936,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.029879663000871703,
      "loss": 0.8159,
      "num_input_tokens_seen": 10620680,
      "step": 18305
    },
    {
      "epoch": 2.727137324992553,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.02987941641119622,
      "loss": 0.8199,
      "num_input_tokens_seen": 10623464,
      "step": 18310
    },
    {
      "epoch": 2.727882037533512,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.029879169570148113,
      "loss": 0.7892,
      "num_input_tokens_seen": 10626216,
      "step": 18315
    },
    {
      "epoch": 2.728626750074471,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02987892247773156,
      "loss": 0.8119,
      "num_input_tokens_seen": 10629512,
      "step": 18320
    },
    {
      "epoch": 2.7293714626154304,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029878675133950736,
      "loss": 0.8118,
      "num_input_tokens_seen": 10632264,
      "step": 18325
    },
    {
      "epoch": 2.7301161751563896,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02987842753880982,
      "loss": 0.7845,
      "num_input_tokens_seen": 10635304,
      "step": 18330
    },
    {
      "epoch": 2.730860887697349,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02987817969231299,
      "loss": 0.82,
      "num_input_tokens_seen": 10637960,
      "step": 18335
    },
    {
      "epoch": 2.731605600238308,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02987793159446443,
      "loss": 0.779,
      "num_input_tokens_seen": 10640680,
      "step": 18340
    },
    {
      "epoch": 2.732350312779267,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.029877683245268342,
      "loss": 0.792,
      "num_input_tokens_seen": 10643496,
      "step": 18345
    },
    {
      "epoch": 2.7330950253202264,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02987743464472892,
      "loss": 0.7938,
      "num_input_tokens_seen": 10646408,
      "step": 18350
    },
    {
      "epoch": 2.7338397378611856,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.029877185792850355,
      "loss": 0.787,
      "num_input_tokens_seen": 10649064,
      "step": 18355
    },
    {
      "epoch": 2.734584450402145,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.02987693668963686,
      "loss": 0.8132,
      "num_input_tokens_seen": 10651752,
      "step": 18360
    },
    {
      "epoch": 2.735329162943104,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029876687335092637,
      "loss": 0.8022,
      "num_input_tokens_seen": 10654376,
      "step": 18365
    },
    {
      "epoch": 2.7360738754840632,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.0298764377292219,
      "loss": 0.7928,
      "num_input_tokens_seen": 10657448,
      "step": 18370
    },
    {
      "epoch": 2.7368185880250224,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029876187872028875,
      "loss": 0.7861,
      "num_input_tokens_seen": 10660264,
      "step": 18375
    },
    {
      "epoch": 2.7375633005659816,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02987593776351777,
      "loss": 0.8203,
      "num_input_tokens_seen": 10663336,
      "step": 18380
    },
    {
      "epoch": 2.738308013106941,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.029875687403692817,
      "loss": 0.8006,
      "num_input_tokens_seen": 10666216,
      "step": 18385
    },
    {
      "epoch": 2.7390527256479,
      "grad_norm": 0.015625,
      "learning_rate": 0.029875436792558243,
      "loss": 0.8012,
      "num_input_tokens_seen": 10669384,
      "step": 18390
    },
    {
      "epoch": 2.7397974381888592,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029875185930118283,
      "loss": 0.7641,
      "num_input_tokens_seen": 10672232,
      "step": 18395
    },
    {
      "epoch": 2.7405421507298184,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.029874934816377176,
      "loss": 0.8143,
      "num_input_tokens_seen": 10675112,
      "step": 18400
    },
    {
      "epoch": 2.7412868632707776,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.029874683451339163,
      "loss": 0.8055,
      "num_input_tokens_seen": 10678088,
      "step": 18405
    },
    {
      "epoch": 2.742031575811737,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029874431835008493,
      "loss": 0.8322,
      "num_input_tokens_seen": 10681256,
      "step": 18410
    },
    {
      "epoch": 2.7427762883526956,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02987417996738941,
      "loss": 0.7972,
      "num_input_tokens_seen": 10684008,
      "step": 18415
    },
    {
      "epoch": 2.7435210008936552,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02987392784848618,
      "loss": 0.8029,
      "num_input_tokens_seen": 10686728,
      "step": 18420
    },
    {
      "epoch": 2.744265713434614,
      "grad_norm": 0.025390625,
      "learning_rate": 0.029873675478303052,
      "loss": 0.8098,
      "num_input_tokens_seen": 10689384,
      "step": 18425
    },
    {
      "epoch": 2.7450104259755737,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029873422856844297,
      "loss": 0.794,
      "num_input_tokens_seen": 10692744,
      "step": 18430
    },
    {
      "epoch": 2.7457551385165324,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.029873169984114177,
      "loss": 0.7695,
      "num_input_tokens_seen": 10695656,
      "step": 18435
    },
    {
      "epoch": 2.746499851057492,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.029872916860116967,
      "loss": 0.8021,
      "num_input_tokens_seen": 10698472,
      "step": 18440
    },
    {
      "epoch": 2.747244563598451,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02987266348485694,
      "loss": 0.8181,
      "num_input_tokens_seen": 10701640,
      "step": 18445
    },
    {
      "epoch": 2.7479892761394105,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02987240985833839,
      "loss": 0.8082,
      "num_input_tokens_seen": 10704392,
      "step": 18450
    },
    {
      "epoch": 2.7487339886803692,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029872155980565578,
      "loss": 0.7971,
      "num_input_tokens_seen": 10707080,
      "step": 18455
    },
    {
      "epoch": 2.7494787012213284,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02987190185154281,
      "loss": 0.7878,
      "num_input_tokens_seen": 10709928,
      "step": 18460
    },
    {
      "epoch": 2.7502234137622876,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.02987164747127438,
      "loss": 0.8221,
      "num_input_tokens_seen": 10713128,
      "step": 18465
    },
    {
      "epoch": 2.750968126303247,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029871392839764577,
      "loss": 0.7872,
      "num_input_tokens_seen": 10716168,
      "step": 18470
    },
    {
      "epoch": 2.751712838844206,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.02987113795701771,
      "loss": 0.8228,
      "num_input_tokens_seen": 10718888,
      "step": 18475
    },
    {
      "epoch": 2.7524575513851652,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029870882823038077,
      "loss": 0.812,
      "num_input_tokens_seen": 10722056,
      "step": 18480
    },
    {
      "epoch": 2.7532022639261244,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02987062743782999,
      "loss": 0.7939,
      "num_input_tokens_seen": 10725064,
      "step": 18485
    },
    {
      "epoch": 2.7539469764670836,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.029870371801397774,
      "loss": 0.7939,
      "num_input_tokens_seen": 10728008,
      "step": 18490
    },
    {
      "epoch": 2.754691689008043,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.029870115913745734,
      "loss": 0.8147,
      "num_input_tokens_seen": 10730824,
      "step": 18495
    },
    {
      "epoch": 2.755436401549002,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.029869859774878205,
      "loss": 0.8098,
      "num_input_tokens_seen": 10733640,
      "step": 18500
    },
    {
      "epoch": 2.7561811140899612,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029869603384799506,
      "loss": 0.8169,
      "num_input_tokens_seen": 10736360,
      "step": 18505
    },
    {
      "epoch": 2.7569258266309205,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029869346743513967,
      "loss": 0.8096,
      "num_input_tokens_seen": 10739272,
      "step": 18510
    },
    {
      "epoch": 2.7576705391718797,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02986908985102593,
      "loss": 0.8088,
      "num_input_tokens_seen": 10742184,
      "step": 18515
    },
    {
      "epoch": 2.758415251712839,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.029868832707339737,
      "loss": 0.8009,
      "num_input_tokens_seen": 10745000,
      "step": 18520
    },
    {
      "epoch": 2.759159964253798,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02986857531245972,
      "loss": 0.7867,
      "num_input_tokens_seen": 10747752,
      "step": 18525
    },
    {
      "epoch": 2.7599046767947573,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029868317666390237,
      "loss": 0.8029,
      "num_input_tokens_seen": 10750792,
      "step": 18530
    },
    {
      "epoch": 2.7606493893357165,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029868059769135636,
      "loss": 0.7989,
      "num_input_tokens_seen": 10753832,
      "step": 18535
    },
    {
      "epoch": 2.7613941018766757,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02986780162070028,
      "loss": 0.8223,
      "num_input_tokens_seen": 10756584,
      "step": 18540
    },
    {
      "epoch": 2.762138814417635,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029867543221088526,
      "loss": 0.8122,
      "num_input_tokens_seen": 10759432,
      "step": 18545
    },
    {
      "epoch": 2.762883526958594,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029867284570304742,
      "loss": 0.817,
      "num_input_tokens_seen": 10762472,
      "step": 18550
    },
    {
      "epoch": 2.7636282394995533,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029867025668353295,
      "loss": 0.8062,
      "num_input_tokens_seen": 10765256,
      "step": 18555
    },
    {
      "epoch": 2.7643729520405125,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029866766515238556,
      "loss": 0.8115,
      "num_input_tokens_seen": 10767880,
      "step": 18560
    },
    {
      "epoch": 2.7651176645814717,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.029866507110964908,
      "loss": 0.8183,
      "num_input_tokens_seen": 10770664,
      "step": 18565
    },
    {
      "epoch": 2.765862377122431,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029866247455536732,
      "loss": 0.81,
      "num_input_tokens_seen": 10773352,
      "step": 18570
    },
    {
      "epoch": 2.76660708966339,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029865987548958417,
      "loss": 0.804,
      "num_input_tokens_seen": 10776360,
      "step": 18575
    },
    {
      "epoch": 2.7673518022043493,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02986572739123435,
      "loss": 0.7939,
      "num_input_tokens_seen": 10779336,
      "step": 18580
    },
    {
      "epoch": 2.7680965147453085,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02986546698236893,
      "loss": 0.7854,
      "num_input_tokens_seen": 10782152,
      "step": 18585
    },
    {
      "epoch": 2.7688412272862672,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029865206322366545,
      "loss": 0.7888,
      "num_input_tokens_seen": 10785096,
      "step": 18590
    },
    {
      "epoch": 2.769585939827227,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029864945411231615,
      "loss": 0.7897,
      "num_input_tokens_seen": 10787944,
      "step": 18595
    },
    {
      "epoch": 2.7703306523681857,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02986468424896854,
      "loss": 0.8163,
      "num_input_tokens_seen": 10791176,
      "step": 18600
    },
    {
      "epoch": 2.7710753649091453,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02986442283558173,
      "loss": 0.8127,
      "num_input_tokens_seen": 10793864,
      "step": 18605
    },
    {
      "epoch": 2.771820077450104,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.029864161171075605,
      "loss": 0.816,
      "num_input_tokens_seen": 10796328,
      "step": 18610
    },
    {
      "epoch": 2.7725647899910637,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.029863899255454584,
      "loss": 0.803,
      "num_input_tokens_seen": 10798984,
      "step": 18615
    },
    {
      "epoch": 2.7733095025320225,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029863637088723093,
      "loss": 0.8046,
      "num_input_tokens_seen": 10801768,
      "step": 18620
    },
    {
      "epoch": 2.7740542150729817,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02986337467088556,
      "loss": 0.8121,
      "num_input_tokens_seen": 10804744,
      "step": 18625
    },
    {
      "epoch": 2.774798927613941,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02986311200194642,
      "loss": 0.7976,
      "num_input_tokens_seen": 10807688,
      "step": 18630
    },
    {
      "epoch": 2.7755436401549,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.029862849081910104,
      "loss": 0.8029,
      "num_input_tokens_seen": 10810472,
      "step": 18635
    },
    {
      "epoch": 2.7762883526958593,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.029862585910781062,
      "loss": 0.8096,
      "num_input_tokens_seen": 10813224,
      "step": 18640
    },
    {
      "epoch": 2.7770330652368185,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.029862322488563738,
      "loss": 0.7903,
      "num_input_tokens_seen": 10815784,
      "step": 18645
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.029862058815262577,
      "loss": 0.8003,
      "num_input_tokens_seen": 10818792,
      "step": 18650
    },
    {
      "epoch": 2.778522490318737,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.029861794890882042,
      "loss": 0.8123,
      "num_input_tokens_seen": 10821800,
      "step": 18655
    },
    {
      "epoch": 2.779267202859696,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029861530715426583,
      "loss": 0.785,
      "num_input_tokens_seen": 10824968,
      "step": 18660
    },
    {
      "epoch": 2.7800119154006553,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.02986126628890067,
      "loss": 0.7924,
      "num_input_tokens_seen": 10827912,
      "step": 18665
    },
    {
      "epoch": 2.7807566279416145,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.029861001611308766,
      "loss": 0.7857,
      "num_input_tokens_seen": 10830728,
      "step": 18670
    },
    {
      "epoch": 2.7815013404825737,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.029860736682655344,
      "loss": 0.8034,
      "num_input_tokens_seen": 10833544,
      "step": 18675
    },
    {
      "epoch": 2.782246053023533,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.029860471502944883,
      "loss": 0.8076,
      "num_input_tokens_seen": 10836296,
      "step": 18680
    },
    {
      "epoch": 2.782990765564492,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.029860206072181854,
      "loss": 0.8082,
      "num_input_tokens_seen": 10839336,
      "step": 18685
    },
    {
      "epoch": 2.7837354781054513,
      "grad_norm": 0.013671875,
      "learning_rate": 0.029859940390370748,
      "loss": 0.797,
      "num_input_tokens_seen": 10841864,
      "step": 18690
    },
    {
      "epoch": 2.7844801906464105,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029859674457516052,
      "loss": 0.7804,
      "num_input_tokens_seen": 10844680,
      "step": 18695
    },
    {
      "epoch": 2.7852249031873697,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02985940827362226,
      "loss": 0.8116,
      "num_input_tokens_seen": 10847368,
      "step": 18700
    },
    {
      "epoch": 2.785969615728329,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029859141838693867,
      "loss": 0.808,
      "num_input_tokens_seen": 10850216,
      "step": 18705
    },
    {
      "epoch": 2.786714328269288,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.029858875152735373,
      "loss": 0.7804,
      "num_input_tokens_seen": 10853032,
      "step": 18710
    },
    {
      "epoch": 2.7874590408102473,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02985860821575129,
      "loss": 0.8282,
      "num_input_tokens_seen": 10856072,
      "step": 18715
    },
    {
      "epoch": 2.7882037533512065,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02985834102774612,
      "loss": 0.805,
      "num_input_tokens_seen": 10858632,
      "step": 18720
    },
    {
      "epoch": 2.7889484658921657,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02985807358872438,
      "loss": 0.8026,
      "num_input_tokens_seen": 10861704,
      "step": 18725
    },
    {
      "epoch": 2.789693178433125,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02985780589869059,
      "loss": 0.7995,
      "num_input_tokens_seen": 10864680,
      "step": 18730
    },
    {
      "epoch": 2.790437890974084,
      "grad_norm": 0.03125,
      "learning_rate": 0.029857537957649265,
      "loss": 0.8101,
      "num_input_tokens_seen": 10867720,
      "step": 18735
    },
    {
      "epoch": 2.7911826035150433,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02985726976560494,
      "loss": 0.8016,
      "num_input_tokens_seen": 10870408,
      "step": 18740
    },
    {
      "epoch": 2.7919273160560025,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02985700132256214,
      "loss": 0.7893,
      "num_input_tokens_seen": 10873320,
      "step": 18745
    },
    {
      "epoch": 2.7926720285969617,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.02985673262852541,
      "loss": 0.8038,
      "num_input_tokens_seen": 10876552,
      "step": 18750
    },
    {
      "epoch": 2.7934167411379205,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02985646368349927,
      "loss": 0.8047,
      "num_input_tokens_seen": 10879592,
      "step": 18755
    },
    {
      "epoch": 2.79416145367888,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02985619448748828,
      "loss": 0.7855,
      "num_input_tokens_seen": 10882472,
      "step": 18760
    },
    {
      "epoch": 2.794906166219839,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02985592504049699,
      "loss": 0.7812,
      "num_input_tokens_seen": 10885352,
      "step": 18765
    },
    {
      "epoch": 2.7956508787607985,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.02985565534252994,
      "loss": 0.8337,
      "num_input_tokens_seen": 10888488,
      "step": 18770
    },
    {
      "epoch": 2.7963955913017573,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.029855385393591696,
      "loss": 0.7785,
      "num_input_tokens_seen": 10891304,
      "step": 18775
    },
    {
      "epoch": 2.797140303842717,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029855115193686808,
      "loss": 0.8135,
      "num_input_tokens_seen": 10893928,
      "step": 18780
    },
    {
      "epoch": 2.7978850163836757,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02985484474281985,
      "loss": 0.7831,
      "num_input_tokens_seen": 10896936,
      "step": 18785
    },
    {
      "epoch": 2.7986297289246354,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.029854574040995386,
      "loss": 0.8024,
      "num_input_tokens_seen": 10899880,
      "step": 18790
    },
    {
      "epoch": 2.799374441465594,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.02985430308821799,
      "loss": 0.8273,
      "num_input_tokens_seen": 10903016,
      "step": 18795
    },
    {
      "epoch": 2.8001191540065533,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.029854031884492246,
      "loss": 0.8085,
      "num_input_tokens_seen": 10905960,
      "step": 18800
    },
    {
      "epoch": 2.8008638665475125,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.029853760429822726,
      "loss": 0.7973,
      "num_input_tokens_seen": 10908808,
      "step": 18805
    },
    {
      "epoch": 2.8016085790884717,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02985348872421402,
      "loss": 0.7901,
      "num_input_tokens_seen": 10911720,
      "step": 18810
    },
    {
      "epoch": 2.802353291629431,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02985321676767072,
      "loss": 0.7885,
      "num_input_tokens_seen": 10914920,
      "step": 18815
    },
    {
      "epoch": 2.80309800417039,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02985294456019742,
      "loss": 0.798,
      "num_input_tokens_seen": 10917832,
      "step": 18820
    },
    {
      "epoch": 2.8038427167113493,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02985267210179872,
      "loss": 0.7891,
      "num_input_tokens_seen": 10920616,
      "step": 18825
    },
    {
      "epoch": 2.8045874292523085,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02985239939247921,
      "loss": 0.7937,
      "num_input_tokens_seen": 10923656,
      "step": 18830
    },
    {
      "epoch": 2.8053321417932677,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029852126432243515,
      "loss": 0.7915,
      "num_input_tokens_seen": 10926664,
      "step": 18835
    },
    {
      "epoch": 2.806076854334227,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.029851853221096236,
      "loss": 0.8056,
      "num_input_tokens_seen": 10929576,
      "step": 18840
    },
    {
      "epoch": 2.806821566875186,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.029851579759041996,
      "loss": 0.7903,
      "num_input_tokens_seen": 10932424,
      "step": 18845
    },
    {
      "epoch": 2.8075662794161453,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02985130604608541,
      "loss": 0.813,
      "num_input_tokens_seen": 10935208,
      "step": 18850
    },
    {
      "epoch": 2.8083109919571045,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.0298510320822311,
      "loss": 0.8056,
      "num_input_tokens_seen": 10938024,
      "step": 18855
    },
    {
      "epoch": 2.8090557044980637,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.029850757867483698,
      "loss": 0.8085,
      "num_input_tokens_seen": 10941160,
      "step": 18860
    },
    {
      "epoch": 2.809800417039023,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.029850483401847833,
      "loss": 0.819,
      "num_input_tokens_seen": 10943784,
      "step": 18865
    },
    {
      "epoch": 2.810545129579982,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.029850208685328148,
      "loss": 0.8168,
      "num_input_tokens_seen": 10946792,
      "step": 18870
    },
    {
      "epoch": 2.8112898421209414,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.029849933717929277,
      "loss": 0.7863,
      "num_input_tokens_seen": 10949576,
      "step": 18875
    },
    {
      "epoch": 2.8120345546619006,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029849658499655873,
      "loss": 0.833,
      "num_input_tokens_seen": 10952104,
      "step": 18880
    },
    {
      "epoch": 2.8127792672028598,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.029849383030512575,
      "loss": 0.7865,
      "num_input_tokens_seen": 10954888,
      "step": 18885
    },
    {
      "epoch": 2.813523979743819,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.029849107310504046,
      "loss": 0.8317,
      "num_input_tokens_seen": 10957768,
      "step": 18890
    },
    {
      "epoch": 2.814268692284778,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02984883133963494,
      "loss": 0.8066,
      "num_input_tokens_seen": 10960488,
      "step": 18895
    },
    {
      "epoch": 2.8150134048257374,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02984855511790992,
      "loss": 0.7901,
      "num_input_tokens_seen": 10963464,
      "step": 18900
    },
    {
      "epoch": 2.8157581173666966,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029848278645333657,
      "loss": 0.7978,
      "num_input_tokens_seen": 10966376,
      "step": 18905
    },
    {
      "epoch": 2.8165028299076558,
      "grad_norm": 0.029296875,
      "learning_rate": 0.029848001921910814,
      "loss": 0.8071,
      "num_input_tokens_seen": 10969224,
      "step": 18910
    },
    {
      "epoch": 2.817247542448615,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02984772494764607,
      "loss": 0.7896,
      "num_input_tokens_seen": 10971944,
      "step": 18915
    },
    {
      "epoch": 2.817992254989574,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.0298474477225441,
      "loss": 0.797,
      "num_input_tokens_seen": 10975016,
      "step": 18920
    },
    {
      "epoch": 2.8187369675305334,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.0298471702466096,
      "loss": 0.7881,
      "num_input_tokens_seen": 10977672,
      "step": 18925
    },
    {
      "epoch": 2.819481680071492,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02984689251984724,
      "loss": 0.8241,
      "num_input_tokens_seen": 10980360,
      "step": 18930
    },
    {
      "epoch": 2.820226392612452,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029846614542261726,
      "loss": 0.8006,
      "num_input_tokens_seen": 10983272,
      "step": 18935
    },
    {
      "epoch": 2.8209711051534105,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029846336313857746,
      "loss": 0.8094,
      "num_input_tokens_seen": 10985992,
      "step": 18940
    },
    {
      "epoch": 2.82171581769437,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02984605783464,
      "loss": 0.8187,
      "num_input_tokens_seen": 10988872,
      "step": 18945
    },
    {
      "epoch": 2.822460530235329,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.0298457791046132,
      "loss": 0.8224,
      "num_input_tokens_seen": 10991784,
      "step": 18950
    },
    {
      "epoch": 2.8232052427762886,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.02984550012378205,
      "loss": 0.8298,
      "num_input_tokens_seen": 10994728,
      "step": 18955
    },
    {
      "epoch": 2.8239499553172474,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029845220892151265,
      "loss": 0.7942,
      "num_input_tokens_seen": 10997576,
      "step": 18960
    },
    {
      "epoch": 2.824694667858207,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02984494140972556,
      "loss": 0.8209,
      "num_input_tokens_seen": 11000168,
      "step": 18965
    },
    {
      "epoch": 2.8254393803991658,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02984466167650966,
      "loss": 0.8056,
      "num_input_tokens_seen": 11002984,
      "step": 18970
    },
    {
      "epoch": 2.826184092940125,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.029844381692508283,
      "loss": 0.7989,
      "num_input_tokens_seen": 11005768,
      "step": 18975
    },
    {
      "epoch": 2.826928805481084,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02984410145772617,
      "loss": 0.7975,
      "num_input_tokens_seen": 11008744,
      "step": 18980
    },
    {
      "epoch": 2.8276735180220434,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029843820972168044,
      "loss": 0.8097,
      "num_input_tokens_seen": 11011720,
      "step": 18985
    },
    {
      "epoch": 2.8284182305630026,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029843540235838652,
      "loss": 0.7987,
      "num_input_tokens_seen": 11014312,
      "step": 18990
    },
    {
      "epoch": 2.8291629431039618,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029843259248742735,
      "loss": 0.7941,
      "num_input_tokens_seen": 11017128,
      "step": 18995
    },
    {
      "epoch": 2.829907655644921,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.02984297801088504,
      "loss": 0.8082,
      "num_input_tokens_seen": 11020072,
      "step": 19000
    },
    {
      "epoch": 2.83065236818588,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029842696522270312,
      "loss": 0.7982,
      "num_input_tokens_seen": 11023080,
      "step": 19005
    },
    {
      "epoch": 2.8313970807268394,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029842414782903313,
      "loss": 0.8116,
      "num_input_tokens_seen": 11025960,
      "step": 19010
    },
    {
      "epoch": 2.8321417932677986,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.029842132792788807,
      "loss": 0.8037,
      "num_input_tokens_seen": 11028488,
      "step": 19015
    },
    {
      "epoch": 2.832886505808758,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02984185055193155,
      "loss": 0.7995,
      "num_input_tokens_seen": 11031464,
      "step": 19020
    },
    {
      "epoch": 2.833631218349717,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02984156806033631,
      "loss": 0.8048,
      "num_input_tokens_seen": 11034440,
      "step": 19025
    },
    {
      "epoch": 2.834375930890676,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029841285318007862,
      "loss": 0.8107,
      "num_input_tokens_seen": 11037384,
      "step": 19030
    },
    {
      "epoch": 2.8351206434316354,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.029841002324950985,
      "loss": 0.8045,
      "num_input_tokens_seen": 11040392,
      "step": 19035
    },
    {
      "epoch": 2.8358653559725946,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.029840719081170457,
      "loss": 0.8141,
      "num_input_tokens_seen": 11043528,
      "step": 19040
    },
    {
      "epoch": 2.836610068513554,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02984043558667106,
      "loss": 0.8003,
      "num_input_tokens_seen": 11046216,
      "step": 19045
    },
    {
      "epoch": 2.837354781054513,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02984015184145759,
      "loss": 0.7952,
      "num_input_tokens_seen": 11049128,
      "step": 19050
    },
    {
      "epoch": 2.838099493595472,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02983986784553484,
      "loss": 0.7929,
      "num_input_tokens_seen": 11051976,
      "step": 19055
    },
    {
      "epoch": 2.8388442061364314,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.029839583598907603,
      "loss": 0.8258,
      "num_input_tokens_seen": 11054888,
      "step": 19060
    },
    {
      "epoch": 2.8395889186773906,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02983929910158068,
      "loss": 0.7967,
      "num_input_tokens_seen": 11057896,
      "step": 19065
    },
    {
      "epoch": 2.84033363121835,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.029839014353558884,
      "loss": 0.7858,
      "num_input_tokens_seen": 11060808,
      "step": 19070
    },
    {
      "epoch": 2.841078343759309,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.02983872935484702,
      "loss": 0.7941,
      "num_input_tokens_seen": 11063592,
      "step": 19075
    },
    {
      "epoch": 2.841823056300268,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029838444105449908,
      "loss": 0.8144,
      "num_input_tokens_seen": 11066376,
      "step": 19080
    },
    {
      "epoch": 2.8425677688412274,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029838158605372367,
      "loss": 0.7963,
      "num_input_tokens_seen": 11069160,
      "step": 19085
    },
    {
      "epoch": 2.8433124813821866,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02983787285461921,
      "loss": 0.7969,
      "num_input_tokens_seen": 11072136,
      "step": 19090
    },
    {
      "epoch": 2.844057193923146,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.029837586853195278,
      "loss": 0.8099,
      "num_input_tokens_seen": 11075368,
      "step": 19095
    },
    {
      "epoch": 2.844801906464105,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029837300601105392,
      "loss": 0.7923,
      "num_input_tokens_seen": 11078312,
      "step": 19100
    },
    {
      "epoch": 2.845546619005064,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02983701409835439,
      "loss": 0.8199,
      "num_input_tokens_seen": 11081288,
      "step": 19105
    },
    {
      "epoch": 2.8462913315460234,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02983672734494712,
      "loss": 0.8057,
      "num_input_tokens_seen": 11084232,
      "step": 19110
    },
    {
      "epoch": 2.847036044086982,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029836440340888417,
      "loss": 0.7989,
      "num_input_tokens_seen": 11087240,
      "step": 19115
    },
    {
      "epoch": 2.847780756627942,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029836153086183134,
      "loss": 0.7993,
      "num_input_tokens_seen": 11090184,
      "step": 19120
    },
    {
      "epoch": 2.8485254691689006,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.029835865580836125,
      "loss": 0.8023,
      "num_input_tokens_seen": 11093032,
      "step": 19125
    },
    {
      "epoch": 2.8492701817098602,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029835577824852244,
      "loss": 0.7978,
      "num_input_tokens_seen": 11095848,
      "step": 19130
    },
    {
      "epoch": 2.850014894250819,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.029835289818236355,
      "loss": 0.8157,
      "num_input_tokens_seen": 11099016,
      "step": 19135
    },
    {
      "epoch": 2.8507596067917786,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.029835001560993325,
      "loss": 0.8142,
      "num_input_tokens_seen": 11102472,
      "step": 19140
    },
    {
      "epoch": 2.8515043193327374,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029834713053128018,
      "loss": 0.8018,
      "num_input_tokens_seen": 11105608,
      "step": 19145
    },
    {
      "epoch": 2.8522490318736966,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.02983442429464531,
      "loss": 0.8004,
      "num_input_tokens_seen": 11108392,
      "step": 19150
    },
    {
      "epoch": 2.852993744414656,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02983413528555008,
      "loss": 0.7911,
      "num_input_tokens_seen": 11111496,
      "step": 19155
    },
    {
      "epoch": 2.853738456955615,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029833846025847217,
      "loss": 0.8226,
      "num_input_tokens_seen": 11114216,
      "step": 19160
    },
    {
      "epoch": 2.854483169496574,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029833556515541596,
      "loss": 0.796,
      "num_input_tokens_seen": 11117448,
      "step": 19165
    },
    {
      "epoch": 2.8552278820375334,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029833266754638117,
      "loss": 0.804,
      "num_input_tokens_seen": 11120296,
      "step": 19170
    },
    {
      "epoch": 2.8559725945784926,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02983297674314167,
      "loss": 0.7794,
      "num_input_tokens_seen": 11123208,
      "step": 19175
    },
    {
      "epoch": 2.856717307119452,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029832686481057157,
      "loss": 0.7957,
      "num_input_tokens_seen": 11126056,
      "step": 19180
    },
    {
      "epoch": 2.857462019660411,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.029832395968389484,
      "loss": 0.8113,
      "num_input_tokens_seen": 11128744,
      "step": 19185
    },
    {
      "epoch": 2.8582067322013702,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.029832105205143553,
      "loss": 0.8125,
      "num_input_tokens_seen": 11131464,
      "step": 19190
    },
    {
      "epoch": 2.8589514447423294,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.02983181419132428,
      "loss": 0.8047,
      "num_input_tokens_seen": 11134344,
      "step": 19195
    },
    {
      "epoch": 2.8596961572832886,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029831522926936582,
      "loss": 0.7923,
      "num_input_tokens_seen": 11137032,
      "step": 19200
    },
    {
      "epoch": 2.860440869824248,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02983123141198538,
      "loss": 0.7968,
      "num_input_tokens_seen": 11140040,
      "step": 19205
    },
    {
      "epoch": 2.861185582365207,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.029830939646475593,
      "loss": 0.8118,
      "num_input_tokens_seen": 11143048,
      "step": 19210
    },
    {
      "epoch": 2.8619302949061662,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02983064763041215,
      "loss": 0.7992,
      "num_input_tokens_seen": 11145768,
      "step": 19215
    },
    {
      "epoch": 2.8626750074471254,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029830355363799996,
      "loss": 0.7947,
      "num_input_tokens_seen": 11148424,
      "step": 19220
    },
    {
      "epoch": 2.8634197199880846,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.02983006284664406,
      "loss": 0.8143,
      "num_input_tokens_seen": 11151272,
      "step": 19225
    },
    {
      "epoch": 2.864164432529044,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029829770078949287,
      "loss": 0.7955,
      "num_input_tokens_seen": 11154440,
      "step": 19230
    },
    {
      "epoch": 2.864909145070003,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.029829477060720613,
      "loss": 0.8001,
      "num_input_tokens_seen": 11157384,
      "step": 19235
    },
    {
      "epoch": 2.8656538576109623,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029829183791963,
      "loss": 0.8247,
      "num_input_tokens_seen": 11160200,
      "step": 19240
    },
    {
      "epoch": 2.8663985701519215,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0298288902726814,
      "loss": 0.7833,
      "num_input_tokens_seen": 11163208,
      "step": 19245
    },
    {
      "epoch": 2.8671432826928807,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029828596502880767,
      "loss": 0.7973,
      "num_input_tokens_seen": 11165736,
      "step": 19250
    },
    {
      "epoch": 2.86788799523384,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02982830248256607,
      "loss": 0.7953,
      "num_input_tokens_seen": 11168488,
      "step": 19255
    },
    {
      "epoch": 2.868632707774799,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029828008211742276,
      "loss": 0.8218,
      "num_input_tokens_seen": 11171496,
      "step": 19260
    },
    {
      "epoch": 2.8693774203157583,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029827713690414353,
      "loss": 0.7839,
      "num_input_tokens_seen": 11174568,
      "step": 19265
    },
    {
      "epoch": 2.8701221328567175,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02982741891858727,
      "loss": 0.8008,
      "num_input_tokens_seen": 11177544,
      "step": 19270
    },
    {
      "epoch": 2.8708668453976767,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02982712389626602,
      "loss": 0.8219,
      "num_input_tokens_seen": 11180200,
      "step": 19275
    },
    {
      "epoch": 2.8716115579386354,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.029826828623455582,
      "loss": 0.814,
      "num_input_tokens_seen": 11183048,
      "step": 19280
    },
    {
      "epoch": 2.872356270479595,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02982653310016094,
      "loss": 0.79,
      "num_input_tokens_seen": 11185768,
      "step": 19285
    },
    {
      "epoch": 2.873100983020554,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029826237326387093,
      "loss": 0.794,
      "num_input_tokens_seen": 11188776,
      "step": 19290
    },
    {
      "epoch": 2.8738456955615135,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029825941302139038,
      "loss": 0.7998,
      "num_input_tokens_seen": 11191528,
      "step": 19295
    },
    {
      "epoch": 2.8745904081024722,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.029825645027421768,
      "loss": 0.7922,
      "num_input_tokens_seen": 11194312,
      "step": 19300
    },
    {
      "epoch": 2.875335120643432,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.029825348502240296,
      "loss": 0.8133,
      "num_input_tokens_seen": 11197320,
      "step": 19305
    },
    {
      "epoch": 2.8760798331843906,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.02982505172659963,
      "loss": 0.7832,
      "num_input_tokens_seen": 11200360,
      "step": 19310
    },
    {
      "epoch": 2.8768245457253503,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029824754700504778,
      "loss": 0.799,
      "num_input_tokens_seen": 11202952,
      "step": 19315
    },
    {
      "epoch": 2.877569258266309,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029824457423960768,
      "loss": 0.8158,
      "num_input_tokens_seen": 11206056,
      "step": 19320
    },
    {
      "epoch": 2.8783139708072683,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029824159896972614,
      "loss": 0.7861,
      "num_input_tokens_seen": 11208904,
      "step": 19325
    },
    {
      "epoch": 2.8790586833482275,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.029823862119545345,
      "loss": 0.7885,
      "num_input_tokens_seen": 11211944,
      "step": 19330
    },
    {
      "epoch": 2.8798033958891867,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029823564091683987,
      "loss": 0.804,
      "num_input_tokens_seen": 11214856,
      "step": 19335
    },
    {
      "epoch": 2.880548108430146,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.029823265813393587,
      "loss": 0.8033,
      "num_input_tokens_seen": 11217832,
      "step": 19340
    },
    {
      "epoch": 2.881292820971105,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029822967284679175,
      "loss": 0.8003,
      "num_input_tokens_seen": 11220520,
      "step": 19345
    },
    {
      "epoch": 2.8820375335120643,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.029822668505545796,
      "loss": 0.826,
      "num_input_tokens_seen": 11223432,
      "step": 19350
    },
    {
      "epoch": 2.8827822460530235,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029822369475998495,
      "loss": 0.8132,
      "num_input_tokens_seen": 11226024,
      "step": 19355
    },
    {
      "epoch": 2.8835269585939827,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02982207019604233,
      "loss": 0.8047,
      "num_input_tokens_seen": 11229160,
      "step": 19360
    },
    {
      "epoch": 2.884271671134942,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029821770665682354,
      "loss": 0.7978,
      "num_input_tokens_seen": 11232168,
      "step": 19365
    },
    {
      "epoch": 2.885016383675901,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029821470884923627,
      "loss": 0.8103,
      "num_input_tokens_seen": 11235080,
      "step": 19370
    },
    {
      "epoch": 2.8857610962168603,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02982117085377121,
      "loss": 0.7861,
      "num_input_tokens_seen": 11237992,
      "step": 19375
    },
    {
      "epoch": 2.8865058087578195,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.029820870572230176,
      "loss": 0.7908,
      "num_input_tokens_seen": 11241416,
      "step": 19380
    },
    {
      "epoch": 2.8872505212987787,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029820570040305597,
      "loss": 0.8037,
      "num_input_tokens_seen": 11244584,
      "step": 19385
    },
    {
      "epoch": 2.887995233839738,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02982026925800255,
      "loss": 0.8046,
      "num_input_tokens_seen": 11247976,
      "step": 19390
    },
    {
      "epoch": 2.888739946380697,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.029819968225326118,
      "loss": 0.8277,
      "num_input_tokens_seen": 11251080,
      "step": 19395
    },
    {
      "epoch": 2.8894846589216563,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02981966694228138,
      "loss": 0.8037,
      "num_input_tokens_seen": 11253928,
      "step": 19400
    },
    {
      "epoch": 2.8902293714626155,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02981936540887344,
      "loss": 0.7947,
      "num_input_tokens_seen": 11256840,
      "step": 19405
    },
    {
      "epoch": 2.8909740840035747,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.029819063625107375,
      "loss": 0.7825,
      "num_input_tokens_seen": 11259496,
      "step": 19410
    },
    {
      "epoch": 2.891718796544534,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029818761590988292,
      "loss": 0.8209,
      "num_input_tokens_seen": 11262344,
      "step": 19415
    },
    {
      "epoch": 2.892463509085493,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.0298184593065213,
      "loss": 0.7917,
      "num_input_tokens_seen": 11265608,
      "step": 19420
    },
    {
      "epoch": 2.8932082216264523,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.029818156771711492,
      "loss": 0.8058,
      "num_input_tokens_seen": 11268584,
      "step": 19425
    },
    {
      "epoch": 2.8939529341674115,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02981785398656399,
      "loss": 0.7946,
      "num_input_tokens_seen": 11271208,
      "step": 19430
    },
    {
      "epoch": 2.8946976467083707,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.0298175509510839,
      "loss": 0.797,
      "num_input_tokens_seen": 11274056,
      "step": 19435
    },
    {
      "epoch": 2.89544235924933,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029817247665276353,
      "loss": 0.7905,
      "num_input_tokens_seen": 11277096,
      "step": 19440
    },
    {
      "epoch": 2.896187071790289,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.029816944129146462,
      "loss": 0.825,
      "num_input_tokens_seen": 11279944,
      "step": 19445
    },
    {
      "epoch": 2.8969317843312483,
      "grad_norm": 0.015625,
      "learning_rate": 0.029816640342699364,
      "loss": 0.8038,
      "num_input_tokens_seen": 11282728,
      "step": 19450
    },
    {
      "epoch": 2.897676496872207,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.029816336305940182,
      "loss": 0.8146,
      "num_input_tokens_seen": 11285448,
      "step": 19455
    },
    {
      "epoch": 2.8984212094131667,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.029816032018874058,
      "loss": 0.8059,
      "num_input_tokens_seen": 11288168,
      "step": 19460
    },
    {
      "epoch": 2.8991659219541255,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029815727481506127,
      "loss": 0.7894,
      "num_input_tokens_seen": 11291048,
      "step": 19465
    },
    {
      "epoch": 2.899910634495085,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.029815422693841544,
      "loss": 0.7873,
      "num_input_tokens_seen": 11293768,
      "step": 19470
    },
    {
      "epoch": 2.900655347036044,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.029815117655885452,
      "loss": 0.7898,
      "num_input_tokens_seen": 11296808,
      "step": 19475
    },
    {
      "epoch": 2.9014000595770035,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029814812367643002,
      "loss": 0.7907,
      "num_input_tokens_seen": 11299432,
      "step": 19480
    },
    {
      "epoch": 2.9021447721179623,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029814506829119358,
      "loss": 0.8125,
      "num_input_tokens_seen": 11302632,
      "step": 19485
    },
    {
      "epoch": 2.9028894846589215,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029814201040319673,
      "loss": 0.7857,
      "num_input_tokens_seen": 11305480,
      "step": 19490
    },
    {
      "epoch": 2.9036341971998807,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02981389500124912,
      "loss": 0.8158,
      "num_input_tokens_seen": 11308200,
      "step": 19495
    },
    {
      "epoch": 2.90437890974084,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.029813588711912867,
      "loss": 0.8202,
      "num_input_tokens_seen": 11311112,
      "step": 19500
    },
    {
      "epoch": 2.905123622281799,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029813282172316092,
      "loss": 0.8034,
      "num_input_tokens_seen": 11314152,
      "step": 19505
    },
    {
      "epoch": 2.9058683348227583,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029812975382463962,
      "loss": 0.817,
      "num_input_tokens_seen": 11316904,
      "step": 19510
    },
    {
      "epoch": 2.9066130473637175,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029812668342361678,
      "loss": 0.8184,
      "num_input_tokens_seen": 11320040,
      "step": 19515
    },
    {
      "epoch": 2.9073577599046767,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029812361052014413,
      "loss": 0.7867,
      "num_input_tokens_seen": 11322920,
      "step": 19520
    },
    {
      "epoch": 2.908102472445636,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029812053511427364,
      "loss": 0.836,
      "num_input_tokens_seen": 11325768,
      "step": 19525
    },
    {
      "epoch": 2.908847184986595,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029811745720605718,
      "loss": 0.8039,
      "num_input_tokens_seen": 11328488,
      "step": 19530
    },
    {
      "epoch": 2.9095918975275543,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02981143767955469,
      "loss": 0.8113,
      "num_input_tokens_seen": 11331304,
      "step": 19535
    },
    {
      "epoch": 2.9103366100685135,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02981112938827947,
      "loss": 0.806,
      "num_input_tokens_seen": 11334120,
      "step": 19540
    },
    {
      "epoch": 2.9110813226094727,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.029810820846785276,
      "loss": 0.787,
      "num_input_tokens_seen": 11336712,
      "step": 19545
    },
    {
      "epoch": 2.911826035150432,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02981051205507732,
      "loss": 0.8044,
      "num_input_tokens_seen": 11339656,
      "step": 19550
    },
    {
      "epoch": 2.912570747691391,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.029810203013160813,
      "loss": 0.7882,
      "num_input_tokens_seen": 11342632,
      "step": 19555
    },
    {
      "epoch": 2.9133154602323503,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029809893721040977,
      "loss": 0.8041,
      "num_input_tokens_seen": 11345768,
      "step": 19560
    },
    {
      "epoch": 2.9140601727733095,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.02980958417872304,
      "loss": 0.7982,
      "num_input_tokens_seen": 11348552,
      "step": 19565
    },
    {
      "epoch": 2.9148048853142687,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02980927438621223,
      "loss": 0.8059,
      "num_input_tokens_seen": 11351336,
      "step": 19570
    },
    {
      "epoch": 2.915549597855228,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029808964343513777,
      "loss": 0.7869,
      "num_input_tokens_seen": 11354024,
      "step": 19575
    },
    {
      "epoch": 2.916294310396187,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029808654050632925,
      "loss": 0.8229,
      "num_input_tokens_seen": 11356840,
      "step": 19580
    },
    {
      "epoch": 2.9170390229371463,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029808343507574917,
      "loss": 0.7893,
      "num_input_tokens_seen": 11359656,
      "step": 19585
    },
    {
      "epoch": 2.9177837354781055,
      "grad_norm": 0.025390625,
      "learning_rate": 0.029808032714344995,
      "loss": 0.7785,
      "num_input_tokens_seen": 11362472,
      "step": 19590
    },
    {
      "epoch": 2.9185284480190647,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029807721670948407,
      "loss": 0.8036,
      "num_input_tokens_seen": 11365672,
      "step": 19595
    },
    {
      "epoch": 2.919273160560024,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029807410377390414,
      "loss": 0.8028,
      "num_input_tokens_seen": 11368552,
      "step": 19600
    },
    {
      "epoch": 2.920017873100983,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02980709883367627,
      "loss": 0.8092,
      "num_input_tokens_seen": 11371528,
      "step": 19605
    },
    {
      "epoch": 2.9207625856419424,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029806787039811242,
      "loss": 0.7878,
      "num_input_tokens_seen": 11374440,
      "step": 19610
    },
    {
      "epoch": 2.9215072981829016,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029806474995800598,
      "loss": 0.8228,
      "num_input_tokens_seen": 11377064,
      "step": 19615
    },
    {
      "epoch": 2.9222520107238603,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029806162701649607,
      "loss": 0.8162,
      "num_input_tokens_seen": 11379944,
      "step": 19620
    },
    {
      "epoch": 2.92299672326482,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02980585015736354,
      "loss": 0.807,
      "num_input_tokens_seen": 11382760,
      "step": 19625
    },
    {
      "epoch": 2.9237414358057787,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029805537362947686,
      "loss": 0.8048,
      "num_input_tokens_seen": 11385704,
      "step": 19630
    },
    {
      "epoch": 2.9244861483467384,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02980522431840733,
      "loss": 0.8058,
      "num_input_tokens_seen": 11388392,
      "step": 19635
    },
    {
      "epoch": 2.925230860887697,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02980491102374775,
      "loss": 0.7879,
      "num_input_tokens_seen": 11391208,
      "step": 19640
    },
    {
      "epoch": 2.9259755734286568,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02980459747897425,
      "loss": 0.7826,
      "num_input_tokens_seen": 11394280,
      "step": 19645
    },
    {
      "epoch": 2.9267202859696155,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02980428368409212,
      "loss": 0.7901,
      "num_input_tokens_seen": 11397256,
      "step": 19650
    },
    {
      "epoch": 2.927464998510575,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029803969639106663,
      "loss": 0.8059,
      "num_input_tokens_seen": 11400328,
      "step": 19655
    },
    {
      "epoch": 2.928209711051534,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.02980365534402319,
      "loss": 0.8279,
      "num_input_tokens_seen": 11403176,
      "step": 19660
    },
    {
      "epoch": 2.928954423592493,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.029803340798847,
      "loss": 0.7797,
      "num_input_tokens_seen": 11405992,
      "step": 19665
    },
    {
      "epoch": 2.9296991361334523,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029803026003583415,
      "loss": 0.8145,
      "num_input_tokens_seen": 11408808,
      "step": 19670
    },
    {
      "epoch": 2.9304438486744115,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02980271095823775,
      "loss": 0.8153,
      "num_input_tokens_seen": 11411624,
      "step": 19675
    },
    {
      "epoch": 2.9311885612153707,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029802395662815326,
      "loss": 0.8177,
      "num_input_tokens_seen": 11414536,
      "step": 19680
    },
    {
      "epoch": 2.93193327375633,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029802080117321477,
      "loss": 0.815,
      "num_input_tokens_seen": 11417384,
      "step": 19685
    },
    {
      "epoch": 2.932677986297289,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.02980176432176152,
      "loss": 0.7934,
      "num_input_tokens_seen": 11420296,
      "step": 19690
    },
    {
      "epoch": 2.9334226988382484,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02980144827614081,
      "loss": 0.7946,
      "num_input_tokens_seen": 11423144,
      "step": 19695
    },
    {
      "epoch": 2.9341674113792076,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02980113198046467,
      "loss": 0.8124,
      "num_input_tokens_seen": 11426120,
      "step": 19700
    },
    {
      "epoch": 2.9349121239201668,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029800815434738448,
      "loss": 0.7901,
      "num_input_tokens_seen": 11429000,
      "step": 19705
    },
    {
      "epoch": 2.935656836461126,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.02980049863896749,
      "loss": 0.803,
      "num_input_tokens_seen": 11431688,
      "step": 19710
    },
    {
      "epoch": 2.936401549002085,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029800181593157154,
      "loss": 0.8218,
      "num_input_tokens_seen": 11434376,
      "step": 19715
    },
    {
      "epoch": 2.9371462615430444,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.029799864297312788,
      "loss": 0.779,
      "num_input_tokens_seen": 11437320,
      "step": 19720
    },
    {
      "epoch": 2.9378909740840036,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.029799546751439757,
      "loss": 0.8262,
      "num_input_tokens_seen": 11440168,
      "step": 19725
    },
    {
      "epoch": 2.9386356866249628,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02979922895554343,
      "loss": 0.8193,
      "num_input_tokens_seen": 11442984,
      "step": 19730
    },
    {
      "epoch": 2.939380399165922,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02979891090962917,
      "loss": 0.8093,
      "num_input_tokens_seen": 11446056,
      "step": 19735
    },
    {
      "epoch": 2.940125111706881,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029798592613702345,
      "loss": 0.7947,
      "num_input_tokens_seen": 11449000,
      "step": 19740
    },
    {
      "epoch": 2.9408698242478404,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029798274067768343,
      "loss": 0.816,
      "num_input_tokens_seen": 11451560,
      "step": 19745
    },
    {
      "epoch": 2.9416145367887996,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.02979795527183254,
      "loss": 0.8176,
      "num_input_tokens_seen": 11454856,
      "step": 19750
    },
    {
      "epoch": 2.942359249329759,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.029797636225900325,
      "loss": 0.7943,
      "num_input_tokens_seen": 11457832,
      "step": 19755
    },
    {
      "epoch": 2.943103961870718,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029797316929977082,
      "loss": 0.8052,
      "num_input_tokens_seen": 11460904,
      "step": 19760
    },
    {
      "epoch": 2.943848674411677,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02979699738406821,
      "loss": 0.8106,
      "num_input_tokens_seen": 11465064,
      "step": 19765
    },
    {
      "epoch": 2.9445933869526364,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.029796677588179107,
      "loss": 0.7981,
      "num_input_tokens_seen": 11468168,
      "step": 19770
    },
    {
      "epoch": 2.9453380994935956,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029796357542315174,
      "loss": 0.7995,
      "num_input_tokens_seen": 11471208,
      "step": 19775
    },
    {
      "epoch": 2.946082812034555,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02979603724648182,
      "loss": 0.801,
      "num_input_tokens_seen": 11474376,
      "step": 19780
    },
    {
      "epoch": 2.946827524575514,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029795716700684455,
      "loss": 0.7999,
      "num_input_tokens_seen": 11477320,
      "step": 19785
    },
    {
      "epoch": 2.947572237116473,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029795395904928492,
      "loss": 0.7948,
      "num_input_tokens_seen": 11480136,
      "step": 19790
    },
    {
      "epoch": 2.948316949657432,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029795074859219353,
      "loss": 0.7822,
      "num_input_tokens_seen": 11483176,
      "step": 19795
    },
    {
      "epoch": 2.9490616621983916,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02979475356356246,
      "loss": 0.7956,
      "num_input_tokens_seen": 11485704,
      "step": 19800
    },
    {
      "epoch": 2.9498063747393504,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.029794432017963245,
      "loss": 0.8032,
      "num_input_tokens_seen": 11488552,
      "step": 19805
    },
    {
      "epoch": 2.95055108728031,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02979411022242714,
      "loss": 0.7957,
      "num_input_tokens_seen": 11491336,
      "step": 19810
    },
    {
      "epoch": 2.9512957998212688,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029793788176959574,
      "loss": 0.7807,
      "num_input_tokens_seen": 11494376,
      "step": 19815
    },
    {
      "epoch": 2.9520405123622284,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029793465881565992,
      "loss": 0.8393,
      "num_input_tokens_seen": 11497352,
      "step": 19820
    },
    {
      "epoch": 2.952785224903187,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029793143336251843,
      "loss": 0.8044,
      "num_input_tokens_seen": 11500008,
      "step": 19825
    },
    {
      "epoch": 2.953529937444147,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02979282054102257,
      "loss": 0.7988,
      "num_input_tokens_seen": 11502824,
      "step": 19830
    },
    {
      "epoch": 2.9542746499851056,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.02979249749588363,
      "loss": 0.8103,
      "num_input_tokens_seen": 11505832,
      "step": 19835
    },
    {
      "epoch": 2.955019362526065,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.029792174200840477,
      "loss": 0.7889,
      "num_input_tokens_seen": 11508872,
      "step": 19840
    },
    {
      "epoch": 2.955764075067024,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029791850655898577,
      "loss": 0.8138,
      "num_input_tokens_seen": 11511464,
      "step": 19845
    },
    {
      "epoch": 2.956508787607983,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.029791526861063393,
      "loss": 0.7885,
      "num_input_tokens_seen": 11514216,
      "step": 19850
    },
    {
      "epoch": 2.9572535001489424,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.029791202816340397,
      "loss": 0.7777,
      "num_input_tokens_seen": 11517480,
      "step": 19855
    },
    {
      "epoch": 2.9579982126899016,
      "grad_norm": 0.01129150390625,
      "learning_rate": 0.02979087852173506,
      "loss": 0.7957,
      "num_input_tokens_seen": 11520136,
      "step": 19860
    },
    {
      "epoch": 2.958742925230861,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029790553977252866,
      "loss": 0.8167,
      "num_input_tokens_seen": 11522952,
      "step": 19865
    },
    {
      "epoch": 2.95948763777182,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029790229182899294,
      "loss": 0.8229,
      "num_input_tokens_seen": 11525960,
      "step": 19870
    },
    {
      "epoch": 2.960232350312779,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029789904138679834,
      "loss": 0.8015,
      "num_input_tokens_seen": 11528488,
      "step": 19875
    },
    {
      "epoch": 2.9609770628537384,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.029789578844599974,
      "loss": 0.7934,
      "num_input_tokens_seen": 11531432,
      "step": 19880
    },
    {
      "epoch": 2.9617217753946976,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029789253300665213,
      "loss": 0.7958,
      "num_input_tokens_seen": 11534248,
      "step": 19885
    },
    {
      "epoch": 2.962466487935657,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02978892750688104,
      "loss": 0.782,
      "num_input_tokens_seen": 11537128,
      "step": 19890
    },
    {
      "epoch": 2.963211200476616,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02978860146325298,
      "loss": 0.818,
      "num_input_tokens_seen": 11540264,
      "step": 19895
    },
    {
      "epoch": 2.963955913017575,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029788275169786518,
      "loss": 0.8335,
      "num_input_tokens_seen": 11543144,
      "step": 19900
    },
    {
      "epoch": 2.9647006255585344,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.029787948626487187,
      "loss": 0.8288,
      "num_input_tokens_seen": 11546152,
      "step": 19905
    },
    {
      "epoch": 2.9654453380994936,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029787621833360485,
      "loss": 0.8264,
      "num_input_tokens_seen": 11548744,
      "step": 19910
    },
    {
      "epoch": 2.966190050640453,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029787294790411943,
      "loss": 0.8034,
      "num_input_tokens_seen": 11551560,
      "step": 19915
    },
    {
      "epoch": 2.966934763181412,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029786967497647088,
      "loss": 0.8051,
      "num_input_tokens_seen": 11554600,
      "step": 19920
    },
    {
      "epoch": 2.9676794757223712,
      "grad_norm": 0.02734375,
      "learning_rate": 0.029786639955071446,
      "loss": 0.7971,
      "num_input_tokens_seen": 11557704,
      "step": 19925
    },
    {
      "epoch": 2.9684241882633304,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029786312162690547,
      "loss": 0.8084,
      "num_input_tokens_seen": 11560424,
      "step": 19930
    },
    {
      "epoch": 2.9691689008042896,
      "grad_norm": 0.025390625,
      "learning_rate": 0.029785984120509933,
      "loss": 0.8015,
      "num_input_tokens_seen": 11563496,
      "step": 19935
    },
    {
      "epoch": 2.969913613345249,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029785655828535148,
      "loss": 0.7984,
      "num_input_tokens_seen": 11566376,
      "step": 19940
    },
    {
      "epoch": 2.970658325886208,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02978532728677173,
      "loss": 0.8047,
      "num_input_tokens_seen": 11569128,
      "step": 19945
    },
    {
      "epoch": 2.9714030384271672,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029784998495225243,
      "loss": 0.8099,
      "num_input_tokens_seen": 11571912,
      "step": 19950
    },
    {
      "epoch": 2.9721477509681264,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029784669453901225,
      "loss": 0.7892,
      "num_input_tokens_seen": 11574696,
      "step": 19955
    },
    {
      "epoch": 2.9728924635090856,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029784340162805248,
      "loss": 0.8012,
      "num_input_tokens_seen": 11577512,
      "step": 19960
    },
    {
      "epoch": 2.973637176050045,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02978401062194287,
      "loss": 0.8114,
      "num_input_tokens_seen": 11580616,
      "step": 19965
    },
    {
      "epoch": 2.9743818885910036,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029783680831319655,
      "loss": 0.7947,
      "num_input_tokens_seen": 11583368,
      "step": 19970
    },
    {
      "epoch": 2.9751266011319633,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029783350790941177,
      "loss": 0.789,
      "num_input_tokens_seen": 11586504,
      "step": 19975
    },
    {
      "epoch": 2.975871313672922,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029783020500813016,
      "loss": 0.811,
      "num_input_tokens_seen": 11589352,
      "step": 19980
    },
    {
      "epoch": 2.9766160262138817,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.02978268996094075,
      "loss": 0.8092,
      "num_input_tokens_seen": 11592360,
      "step": 19985
    },
    {
      "epoch": 2.9773607387548404,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029782359171329962,
      "loss": 0.8086,
      "num_input_tokens_seen": 11595272,
      "step": 19990
    },
    {
      "epoch": 2.9781054512958,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029782028131986233,
      "loss": 0.7982,
      "num_input_tokens_seen": 11598408,
      "step": 19995
    },
    {
      "epoch": 2.978850163836759,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029781696842915168,
      "loss": 0.8011,
      "num_input_tokens_seen": 11601000,
      "step": 20000
    },
    {
      "epoch": 2.9795948763777185,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029781365304122356,
      "loss": 0.781,
      "num_input_tokens_seen": 11603784,
      "step": 20005
    },
    {
      "epoch": 2.9803395889186772,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.0297810335156134,
      "loss": 0.8169,
      "num_input_tokens_seen": 11606728,
      "step": 20010
    },
    {
      "epoch": 2.9810843014596364,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02978070147739391,
      "loss": 0.8077,
      "num_input_tokens_seen": 11609672,
      "step": 20015
    },
    {
      "epoch": 2.9818290140005956,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029780369189469486,
      "loss": 0.8027,
      "num_input_tokens_seen": 11612584,
      "step": 20020
    },
    {
      "epoch": 2.982573726541555,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.029780036651845747,
      "loss": 0.7903,
      "num_input_tokens_seen": 11615560,
      "step": 20025
    },
    {
      "epoch": 2.983318439082514,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02977970386452831,
      "loss": 0.8199,
      "num_input_tokens_seen": 11618664,
      "step": 20030
    },
    {
      "epoch": 2.9840631516234732,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.0297793708275228,
      "loss": 0.7794,
      "num_input_tokens_seen": 11621512,
      "step": 20035
    },
    {
      "epoch": 2.9848078641644324,
      "grad_norm": 0.01171875,
      "learning_rate": 0.02977903754083484,
      "loss": 0.8089,
      "num_input_tokens_seen": 11624296,
      "step": 20040
    },
    {
      "epoch": 2.9855525767053916,
      "grad_norm": 0.01171875,
      "learning_rate": 0.029778704004470063,
      "loss": 0.7788,
      "num_input_tokens_seen": 11627112,
      "step": 20045
    },
    {
      "epoch": 2.986297289246351,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.0297783702184341,
      "loss": 0.8052,
      "num_input_tokens_seen": 11630088,
      "step": 20050
    },
    {
      "epoch": 2.98704200178731,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029778036182732594,
      "loss": 0.8131,
      "num_input_tokens_seen": 11632744,
      "step": 20055
    },
    {
      "epoch": 2.9877867143282693,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029777701897371188,
      "loss": 0.7956,
      "num_input_tokens_seen": 11636104,
      "step": 20060
    },
    {
      "epoch": 2.9885314268692285,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029777367362355527,
      "loss": 0.8088,
      "num_input_tokens_seen": 11639112,
      "step": 20065
    },
    {
      "epoch": 2.9892761394101877,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029777032577691258,
      "loss": 0.7902,
      "num_input_tokens_seen": 11641864,
      "step": 20070
    },
    {
      "epoch": 2.990020851951147,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029776697543384047,
      "loss": 0.7827,
      "num_input_tokens_seen": 11644776,
      "step": 20075
    },
    {
      "epoch": 2.990765564492106,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02977636225943955,
      "loss": 0.8016,
      "num_input_tokens_seen": 11647752,
      "step": 20080
    },
    {
      "epoch": 2.9915102770330653,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029776026725863425,
      "loss": 0.8228,
      "num_input_tokens_seen": 11650536,
      "step": 20085
    },
    {
      "epoch": 2.9922549895740245,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.029775690942661356,
      "loss": 0.7805,
      "num_input_tokens_seen": 11653256,
      "step": 20090
    },
    {
      "epoch": 2.9929997021149837,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029775354909839,
      "loss": 0.7997,
      "num_input_tokens_seen": 11656072,
      "step": 20095
    },
    {
      "epoch": 2.993744414655943,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029775018627402037,
      "loss": 0.7825,
      "num_input_tokens_seen": 11658984,
      "step": 20100
    },
    {
      "epoch": 2.994489127196902,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029774682095356154,
      "loss": 0.8068,
      "num_input_tokens_seen": 11662120,
      "step": 20105
    },
    {
      "epoch": 2.9952338397378613,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.029774345313707033,
      "loss": 0.8025,
      "num_input_tokens_seen": 11665512,
      "step": 20110
    },
    {
      "epoch": 2.9959785522788205,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02977400828246037,
      "loss": 0.8087,
      "num_input_tokens_seen": 11668392,
      "step": 20115
    },
    {
      "epoch": 2.9967232648197797,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029773671001621843,
      "loss": 0.8124,
      "num_input_tokens_seen": 11671624,
      "step": 20120
    },
    {
      "epoch": 2.997467977360739,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029773333471197164,
      "loss": 0.7966,
      "num_input_tokens_seen": 11674408,
      "step": 20125
    },
    {
      "epoch": 2.998212689901698,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02977299569119203,
      "loss": 0.7986,
      "num_input_tokens_seen": 11677416,
      "step": 20130
    },
    {
      "epoch": 2.9989574024426573,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.029772657661612148,
      "loss": 0.7871,
      "num_input_tokens_seen": 11680424,
      "step": 20135
    },
    {
      "epoch": 2.9997021149836165,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02977231938246323,
      "loss": 0.792,
      "num_input_tokens_seen": 11683176,
      "step": 20140
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.802687406539917,
      "eval_runtime": 70.7039,
      "eval_samples_per_second": 42.204,
      "eval_steps_per_second": 10.551,
      "num_input_tokens_seen": 11683856,
      "step": 20142
    },
    {
      "epoch": 3.0004468275245757,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02977198085375099,
      "loss": 0.8022,
      "num_input_tokens_seen": 11685584,
      "step": 20145
    },
    {
      "epoch": 3.001191540065535,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029771642075481147,
      "loss": 0.7899,
      "num_input_tokens_seen": 11688464,
      "step": 20150
    },
    {
      "epoch": 3.001936252606494,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029771303047659424,
      "loss": 0.7894,
      "num_input_tokens_seen": 11691280,
      "step": 20155
    },
    {
      "epoch": 3.002680965147453,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02977096377029155,
      "loss": 0.8257,
      "num_input_tokens_seen": 11694000,
      "step": 20160
    },
    {
      "epoch": 3.003425677688412,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.029770624243383257,
      "loss": 0.8264,
      "num_input_tokens_seen": 11697072,
      "step": 20165
    },
    {
      "epoch": 3.0041703902293713,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029770284466940276,
      "loss": 0.7695,
      "num_input_tokens_seen": 11700144,
      "step": 20170
    },
    {
      "epoch": 3.0049151027703305,
      "grad_norm": 0.01544189453125,
      "learning_rate": 0.02976994444096835,
      "loss": 0.804,
      "num_input_tokens_seen": 11703120,
      "step": 20175
    },
    {
      "epoch": 3.0056598153112897,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029769604165473227,
      "loss": 0.8019,
      "num_input_tokens_seen": 11705776,
      "step": 20180
    },
    {
      "epoch": 3.006404527852249,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.029769263640460654,
      "loss": 0.809,
      "num_input_tokens_seen": 11708592,
      "step": 20185
    },
    {
      "epoch": 3.007149240393208,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.02976892286593638,
      "loss": 0.8117,
      "num_input_tokens_seen": 11711472,
      "step": 20190
    },
    {
      "epoch": 3.0078939529341673,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.029768581841906165,
      "loss": 0.8046,
      "num_input_tokens_seen": 11714800,
      "step": 20195
    },
    {
      "epoch": 3.0086386654751265,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02976824056837577,
      "loss": 0.8246,
      "num_input_tokens_seen": 11717424,
      "step": 20200
    },
    {
      "epoch": 3.0093833780160857,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029767899045350963,
      "loss": 0.8106,
      "num_input_tokens_seen": 11720240,
      "step": 20205
    },
    {
      "epoch": 3.010128090557045,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029767557272837506,
      "loss": 0.7849,
      "num_input_tokens_seen": 11723248,
      "step": 20210
    },
    {
      "epoch": 3.010872803098004,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02976721525084118,
      "loss": 0.7768,
      "num_input_tokens_seen": 11726128,
      "step": 20215
    },
    {
      "epoch": 3.0116175156389633,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029766872979367767,
      "loss": 0.7914,
      "num_input_tokens_seen": 11729008,
      "step": 20220
    },
    {
      "epoch": 3.0123622281799225,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02976653045842304,
      "loss": 0.7946,
      "num_input_tokens_seen": 11732016,
      "step": 20225
    },
    {
      "epoch": 3.0131069407208817,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02976618768801278,
      "loss": 0.8268,
      "num_input_tokens_seen": 11734928,
      "step": 20230
    },
    {
      "epoch": 3.013851653261841,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.029765844668142798,
      "loss": 0.8229,
      "num_input_tokens_seen": 11737968,
      "step": 20235
    },
    {
      "epoch": 3.0145963658028,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.029765501398818877,
      "loss": 0.8245,
      "num_input_tokens_seen": 11740752,
      "step": 20240
    },
    {
      "epoch": 3.0153410783437593,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02976515788004681,
      "loss": 0.7973,
      "num_input_tokens_seen": 11743664,
      "step": 20245
    },
    {
      "epoch": 3.0160857908847185,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029764814111832412,
      "loss": 0.8087,
      "num_input_tokens_seen": 11746608,
      "step": 20250
    },
    {
      "epoch": 3.0168305034256777,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029764470094181483,
      "loss": 0.7988,
      "num_input_tokens_seen": 11749712,
      "step": 20255
    },
    {
      "epoch": 3.017575215966637,
      "grad_norm": 0.010986328125,
      "learning_rate": 0.029764125827099842,
      "loss": 0.8086,
      "num_input_tokens_seen": 11752528,
      "step": 20260
    },
    {
      "epoch": 3.018319928507596,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.029763781310593294,
      "loss": 0.8123,
      "num_input_tokens_seen": 11755344,
      "step": 20265
    },
    {
      "epoch": 3.0190646410485553,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029763436544667673,
      "loss": 0.7924,
      "num_input_tokens_seen": 11758064,
      "step": 20270
    },
    {
      "epoch": 3.0198093535895145,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029763091529328795,
      "loss": 0.795,
      "num_input_tokens_seen": 11760848,
      "step": 20275
    },
    {
      "epoch": 3.0205540661304737,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.029762746264582488,
      "loss": 0.791,
      "num_input_tokens_seen": 11764048,
      "step": 20280
    },
    {
      "epoch": 3.021298778671433,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029762400750434584,
      "loss": 0.7803,
      "num_input_tokens_seen": 11767248,
      "step": 20285
    },
    {
      "epoch": 3.022043491212392,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.02976205498689093,
      "loss": 0.7827,
      "num_input_tokens_seen": 11770256,
      "step": 20290
    },
    {
      "epoch": 3.0227882037533513,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029761708973957354,
      "loss": 0.7972,
      "num_input_tokens_seen": 11773264,
      "step": 20295
    },
    {
      "epoch": 3.0235329162943105,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029761362711639715,
      "loss": 0.7999,
      "num_input_tokens_seen": 11776336,
      "step": 20300
    },
    {
      "epoch": 3.0242776288352697,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02976101619994385,
      "loss": 0.8272,
      "num_input_tokens_seen": 11779408,
      "step": 20305
    },
    {
      "epoch": 3.025022341376229,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02976066943887562,
      "loss": 0.8003,
      "num_input_tokens_seen": 11782320,
      "step": 20310
    },
    {
      "epoch": 3.025767053917188,
      "grad_norm": 0.02734375,
      "learning_rate": 0.029760322428440886,
      "loss": 0.8173,
      "num_input_tokens_seen": 11786032,
      "step": 20315
    },
    {
      "epoch": 3.0265117664581473,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.029759975168645498,
      "loss": 0.8115,
      "num_input_tokens_seen": 11789360,
      "step": 20320
    },
    {
      "epoch": 3.0272564789991065,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.029759627659495337,
      "loss": 0.7982,
      "num_input_tokens_seen": 11792336,
      "step": 20325
    },
    {
      "epoch": 3.0280011915400658,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029759279900996267,
      "loss": 0.8238,
      "num_input_tokens_seen": 11794992,
      "step": 20330
    },
    {
      "epoch": 3.0287459040810245,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029758931893154164,
      "loss": 0.8023,
      "num_input_tokens_seen": 11797712,
      "step": 20335
    },
    {
      "epoch": 3.0294906166219837,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029758583635974906,
      "loss": 0.8038,
      "num_input_tokens_seen": 11800496,
      "step": 20340
    },
    {
      "epoch": 3.030235329162943,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029758235129464378,
      "loss": 0.7936,
      "num_input_tokens_seen": 11803248,
      "step": 20345
    },
    {
      "epoch": 3.030980041703902,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029757886373628466,
      "loss": 0.7947,
      "num_input_tokens_seen": 11806000,
      "step": 20350
    },
    {
      "epoch": 3.0317247542448613,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029757537368473065,
      "loss": 0.7764,
      "num_input_tokens_seen": 11808976,
      "step": 20355
    },
    {
      "epoch": 3.0324694667858205,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029757188114004067,
      "loss": 0.7865,
      "num_input_tokens_seen": 11811888,
      "step": 20360
    },
    {
      "epoch": 3.0332141793267797,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029756838610227376,
      "loss": 0.8083,
      "num_input_tokens_seen": 11814928,
      "step": 20365
    },
    {
      "epoch": 3.033958891867739,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.029756488857148895,
      "loss": 0.8166,
      "num_input_tokens_seen": 11818320,
      "step": 20370
    },
    {
      "epoch": 3.034703604408698,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02975613885477453,
      "loss": 0.7925,
      "num_input_tokens_seen": 11821072,
      "step": 20375
    },
    {
      "epoch": 3.0354483169496573,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.0297557886031102,
      "loss": 0.792,
      "num_input_tokens_seen": 11823792,
      "step": 20380
    },
    {
      "epoch": 3.0361930294906165,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.029755438102161816,
      "loss": 0.8146,
      "num_input_tokens_seen": 11826832,
      "step": 20385
    },
    {
      "epoch": 3.0369377420315757,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029755087351935303,
      "loss": 0.7939,
      "num_input_tokens_seen": 11829520,
      "step": 20390
    },
    {
      "epoch": 3.037682454572535,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029754736352436584,
      "loss": 0.7922,
      "num_input_tokens_seen": 11832496,
      "step": 20395
    },
    {
      "epoch": 3.038427167113494,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02975438510367159,
      "loss": 0.7914,
      "num_input_tokens_seen": 11835376,
      "step": 20400
    },
    {
      "epoch": 3.0391718796544533,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.029754033605646258,
      "loss": 0.8175,
      "num_input_tokens_seen": 11838384,
      "step": 20405
    },
    {
      "epoch": 3.0399165921954125,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.029753681858366524,
      "loss": 0.8158,
      "num_input_tokens_seen": 11841168,
      "step": 20410
    },
    {
      "epoch": 3.0406613047363718,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029753329861838332,
      "loss": 0.8003,
      "num_input_tokens_seen": 11844464,
      "step": 20415
    },
    {
      "epoch": 3.041406017277331,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.029752977616067617,
      "loss": 0.8249,
      "num_input_tokens_seen": 11847312,
      "step": 20420
    },
    {
      "epoch": 3.04215072981829,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029752625121060347,
      "loss": 0.8052,
      "num_input_tokens_seen": 11850256,
      "step": 20425
    },
    {
      "epoch": 3.0428954423592494,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02975227237682247,
      "loss": 0.7853,
      "num_input_tokens_seen": 11853136,
      "step": 20430
    },
    {
      "epoch": 3.0436401549002086,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029751919383359943,
      "loss": 0.7928,
      "num_input_tokens_seen": 11856304,
      "step": 20435
    },
    {
      "epoch": 3.0443848674411678,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02975156614067873,
      "loss": 0.8302,
      "num_input_tokens_seen": 11859248,
      "step": 20440
    },
    {
      "epoch": 3.045129579982127,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.0297512126487848,
      "loss": 0.7849,
      "num_input_tokens_seen": 11862032,
      "step": 20445
    },
    {
      "epoch": 3.045874292523086,
      "grad_norm": 0.01123046875,
      "learning_rate": 0.029750858907684127,
      "loss": 0.7891,
      "num_input_tokens_seen": 11864912,
      "step": 20450
    },
    {
      "epoch": 3.0466190050640454,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029750504917382684,
      "loss": 0.7955,
      "num_input_tokens_seen": 11867856,
      "step": 20455
    },
    {
      "epoch": 3.0473637176050046,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029750150677886455,
      "loss": 0.802,
      "num_input_tokens_seen": 11870832,
      "step": 20460
    },
    {
      "epoch": 3.0481084301459638,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029749796189201413,
      "loss": 0.7813,
      "num_input_tokens_seen": 11874640,
      "step": 20465
    },
    {
      "epoch": 3.048853142686923,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029749441451333566,
      "loss": 0.8226,
      "num_input_tokens_seen": 11877296,
      "step": 20470
    },
    {
      "epoch": 3.049597855227882,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02974908646428889,
      "loss": 0.8329,
      "num_input_tokens_seen": 11879792,
      "step": 20475
    },
    {
      "epoch": 3.0503425677688414,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029748731228073387,
      "loss": 0.8147,
      "num_input_tokens_seen": 11882512,
      "step": 20480
    },
    {
      "epoch": 3.0510872803098006,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.029748375742693067,
      "loss": 0.7808,
      "num_input_tokens_seen": 11885520,
      "step": 20485
    },
    {
      "epoch": 3.05183199285076,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029748020008153922,
      "loss": 0.7823,
      "num_input_tokens_seen": 11888304,
      "step": 20490
    },
    {
      "epoch": 3.052576705391719,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02974766402446197,
      "loss": 0.795,
      "num_input_tokens_seen": 11891056,
      "step": 20495
    },
    {
      "epoch": 3.053321417932678,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.029747307791623222,
      "loss": 0.7908,
      "num_input_tokens_seen": 11893776,
      "step": 20500
    },
    {
      "epoch": 3.054066130473637,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.0297469513096437,
      "loss": 0.8207,
      "num_input_tokens_seen": 11897008,
      "step": 20505
    },
    {
      "epoch": 3.054810843014596,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02974659457852942,
      "loss": 0.797,
      "num_input_tokens_seen": 11900080,
      "step": 20510
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02974623759828642,
      "loss": 0.7796,
      "num_input_tokens_seen": 11902992,
      "step": 20515
    },
    {
      "epoch": 3.0563002680965146,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.029745880368920716,
      "loss": 0.7995,
      "num_input_tokens_seen": 11905840,
      "step": 20520
    },
    {
      "epoch": 3.0570449806374738,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029745522890438348,
      "loss": 0.8025,
      "num_input_tokens_seen": 11908464,
      "step": 20525
    },
    {
      "epoch": 3.057789693178433,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029745165162845363,
      "loss": 0.772,
      "num_input_tokens_seen": 11911312,
      "step": 20530
    },
    {
      "epoch": 3.058534405719392,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.0297448071861478,
      "loss": 0.7974,
      "num_input_tokens_seen": 11914192,
      "step": 20535
    },
    {
      "epoch": 3.0592791182603514,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029744448960351704,
      "loss": 0.7714,
      "num_input_tokens_seen": 11917328,
      "step": 20540
    },
    {
      "epoch": 3.0600238308013106,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029744090485463125,
      "loss": 0.7447,
      "num_input_tokens_seen": 11920048,
      "step": 20545
    },
    {
      "epoch": 3.0607685433422698,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.029743731761488125,
      "loss": 0.8059,
      "num_input_tokens_seen": 11922800,
      "step": 20550
    },
    {
      "epoch": 3.061513255883229,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.029743372788432764,
      "loss": 0.8027,
      "num_input_tokens_seen": 11926096,
      "step": 20555
    },
    {
      "epoch": 3.062257968424188,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.0297430135663031,
      "loss": 0.8007,
      "num_input_tokens_seen": 11928944,
      "step": 20560
    },
    {
      "epoch": 3.0630026809651474,
      "grad_norm": 0.025390625,
      "learning_rate": 0.029742654095105205,
      "loss": 0.7869,
      "num_input_tokens_seen": 11932240,
      "step": 20565
    },
    {
      "epoch": 3.0637473935061066,
      "grad_norm": 0.017578125,
      "learning_rate": 0.02974229437484516,
      "loss": 0.8257,
      "num_input_tokens_seen": 11935216,
      "step": 20570
    },
    {
      "epoch": 3.064492106047066,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.029741934405529028,
      "loss": 0.7903,
      "num_input_tokens_seen": 11938288,
      "step": 20575
    },
    {
      "epoch": 3.065236818588025,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.0297415741871629,
      "loss": 0.8036,
      "num_input_tokens_seen": 11940816,
      "step": 20580
    },
    {
      "epoch": 3.065981531128984,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02974121371975286,
      "loss": 0.8341,
      "num_input_tokens_seen": 11943728,
      "step": 20585
    },
    {
      "epoch": 3.0667262436699434,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029740853003304993,
      "loss": 0.7803,
      "num_input_tokens_seen": 11946544,
      "step": 20590
    },
    {
      "epoch": 3.0674709562109026,
      "grad_norm": 0.015625,
      "learning_rate": 0.0297404920378254,
      "loss": 0.8121,
      "num_input_tokens_seen": 11949488,
      "step": 20595
    },
    {
      "epoch": 3.068215668751862,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.029740130823320176,
      "loss": 0.8474,
      "num_input_tokens_seen": 11952240,
      "step": 20600
    },
    {
      "epoch": 3.068960381292821,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02973976935979542,
      "loss": 0.8286,
      "num_input_tokens_seen": 11954896,
      "step": 20605
    },
    {
      "epoch": 3.06970509383378,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.029739407647257243,
      "loss": 0.794,
      "num_input_tokens_seen": 11957712,
      "step": 20610
    },
    {
      "epoch": 3.0704498063747394,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.029739045685711753,
      "loss": 0.7896,
      "num_input_tokens_seen": 11960592,
      "step": 20615
    },
    {
      "epoch": 3.0711945189156986,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.029738683475165065,
      "loss": 0.8068,
      "num_input_tokens_seen": 11963664,
      "step": 20620
    },
    {
      "epoch": 3.071939231456658,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029738321015623306,
      "loss": 0.8096,
      "num_input_tokens_seen": 11966928,
      "step": 20625
    },
    {
      "epoch": 3.072683943997617,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02973795830709259,
      "loss": 0.8291,
      "num_input_tokens_seen": 11970000,
      "step": 20630
    },
    {
      "epoch": 3.073428656538576,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029737595349579045,
      "loss": 0.82,
      "num_input_tokens_seen": 11973040,
      "step": 20635
    },
    {
      "epoch": 3.0741733690795354,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.0297372321430888,
      "loss": 0.8124,
      "num_input_tokens_seen": 11976016,
      "step": 20640
    },
    {
      "epoch": 3.0749180816204946,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.029736868687628006,
      "loss": 0.8184,
      "num_input_tokens_seen": 11979120,
      "step": 20645
    },
    {
      "epoch": 3.075662794161454,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029736504983202787,
      "loss": 0.835,
      "num_input_tokens_seen": 11981776,
      "step": 20650
    },
    {
      "epoch": 3.076407506702413,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029736141029819297,
      "loss": 0.8194,
      "num_input_tokens_seen": 11984400,
      "step": 20655
    },
    {
      "epoch": 3.0771522192433722,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.02973577682748368,
      "loss": 0.7991,
      "num_input_tokens_seen": 11987248,
      "step": 20660
    },
    {
      "epoch": 3.0778969317843314,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029735412376202093,
      "loss": 0.8066,
      "num_input_tokens_seen": 11990128,
      "step": 20665
    },
    {
      "epoch": 3.0786416443252906,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02973504767598069,
      "loss": 0.8059,
      "num_input_tokens_seen": 11993392,
      "step": 20670
    },
    {
      "epoch": 3.07938635686625,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02973468272682563,
      "loss": 0.7976,
      "num_input_tokens_seen": 11996464,
      "step": 20675
    },
    {
      "epoch": 3.0801310694072086,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02973431752874308,
      "loss": 0.8056,
      "num_input_tokens_seen": 11999472,
      "step": 20680
    },
    {
      "epoch": 3.080875781948168,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.029733952081739214,
      "loss": 0.7924,
      "num_input_tokens_seen": 12002416,
      "step": 20685
    },
    {
      "epoch": 3.081620494489127,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.0297335863858202,
      "loss": 0.7864,
      "num_input_tokens_seen": 12005456,
      "step": 20690
    },
    {
      "epoch": 3.082365207030086,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.02973322044099222,
      "loss": 0.7857,
      "num_input_tokens_seen": 12008464,
      "step": 20695
    },
    {
      "epoch": 3.0831099195710454,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.029732854247261454,
      "loss": 0.8215,
      "num_input_tokens_seen": 12011312,
      "step": 20700
    },
    {
      "epoch": 3.0838546321120046,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.029732487804634087,
      "loss": 0.7897,
      "num_input_tokens_seen": 12014512,
      "step": 20705
    },
    {
      "epoch": 3.084599344652964,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02973212111311632,
      "loss": 0.8045,
      "num_input_tokens_seen": 12017360,
      "step": 20710
    },
    {
      "epoch": 3.085344057193923,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02973175417271433,
      "loss": 0.8257,
      "num_input_tokens_seen": 12020112,
      "step": 20715
    },
    {
      "epoch": 3.086088769734882,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029731386983434333,
      "loss": 0.797,
      "num_input_tokens_seen": 12022800,
      "step": 20720
    },
    {
      "epoch": 3.0868334822758414,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02973101954528252,
      "loss": 0.8076,
      "num_input_tokens_seen": 12025840,
      "step": 20725
    },
    {
      "epoch": 3.0875781948168006,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.02973065185826511,
      "loss": 0.7742,
      "num_input_tokens_seen": 12028784,
      "step": 20730
    },
    {
      "epoch": 3.08832290735776,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.0297302839223883,
      "loss": 0.7792,
      "num_input_tokens_seen": 12031408,
      "step": 20735
    },
    {
      "epoch": 3.089067619898719,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.029729915737658326,
      "loss": 0.8182,
      "num_input_tokens_seen": 12034224,
      "step": 20740
    },
    {
      "epoch": 3.0898123324396782,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029729547304081384,
      "loss": 0.797,
      "num_input_tokens_seen": 12036880,
      "step": 20745
    },
    {
      "epoch": 3.0905570449806374,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02972917862166372,
      "loss": 0.7901,
      "num_input_tokens_seen": 12039920,
      "step": 20750
    },
    {
      "epoch": 3.0913017575215966,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029728809690411546,
      "loss": 0.7774,
      "num_input_tokens_seen": 12043088,
      "step": 20755
    },
    {
      "epoch": 3.092046470062556,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02972844051033111,
      "loss": 0.7966,
      "num_input_tokens_seen": 12045968,
      "step": 20760
    },
    {
      "epoch": 3.092791182603515,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029728071081428633,
      "loss": 0.8242,
      "num_input_tokens_seen": 12049008,
      "step": 20765
    },
    {
      "epoch": 3.0935358951444742,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02972770140371037,
      "loss": 0.7866,
      "num_input_tokens_seen": 12051856,
      "step": 20770
    },
    {
      "epoch": 3.0942806076854334,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029727331477182557,
      "loss": 0.8008,
      "num_input_tokens_seen": 12054608,
      "step": 20775
    },
    {
      "epoch": 3.0950253202263927,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.029726961301851446,
      "loss": 0.8443,
      "num_input_tokens_seen": 12057584,
      "step": 20780
    },
    {
      "epoch": 3.095770032767352,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.029726590877723294,
      "loss": 0.8115,
      "num_input_tokens_seen": 12060272,
      "step": 20785
    },
    {
      "epoch": 3.096514745308311,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.029726220204804355,
      "loss": 0.8083,
      "num_input_tokens_seen": 12063120,
      "step": 20790
    },
    {
      "epoch": 3.0972594578492703,
      "grad_norm": 0.01171875,
      "learning_rate": 0.029725849283100896,
      "loss": 0.8124,
      "num_input_tokens_seen": 12066032,
      "step": 20795
    },
    {
      "epoch": 3.0980041703902295,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029725478112619177,
      "loss": 0.7886,
      "num_input_tokens_seen": 12068816,
      "step": 20800
    },
    {
      "epoch": 3.0987488829311887,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02972510669336547,
      "loss": 0.8032,
      "num_input_tokens_seen": 12071792,
      "step": 20805
    },
    {
      "epoch": 3.099493595472148,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029724735025346057,
      "loss": 0.8198,
      "num_input_tokens_seen": 12074576,
      "step": 20810
    },
    {
      "epoch": 3.100238308013107,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.029724363108567207,
      "loss": 0.8145,
      "num_input_tokens_seen": 12077200,
      "step": 20815
    },
    {
      "epoch": 3.1009830205540663,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029723990943035208,
      "loss": 0.8042,
      "num_input_tokens_seen": 12080048,
      "step": 20820
    },
    {
      "epoch": 3.1017277330950255,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029723618528756345,
      "loss": 0.7993,
      "num_input_tokens_seen": 12082736,
      "step": 20825
    },
    {
      "epoch": 3.1024724456359847,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.029723245865736915,
      "loss": 0.8077,
      "num_input_tokens_seen": 12085456,
      "step": 20830
    },
    {
      "epoch": 3.103217158176944,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02972287295398321,
      "loss": 0.7949,
      "num_input_tokens_seen": 12088464,
      "step": 20835
    },
    {
      "epoch": 3.103961870717903,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.029722499793501528,
      "loss": 0.7977,
      "num_input_tokens_seen": 12091312,
      "step": 20840
    },
    {
      "epoch": 3.1047065832588623,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.029722126384298178,
      "loss": 0.8185,
      "num_input_tokens_seen": 12094256,
      "step": 20845
    },
    {
      "epoch": 3.1054512957998215,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029721752726379464,
      "loss": 0.8045,
      "num_input_tokens_seen": 12097040,
      "step": 20850
    },
    {
      "epoch": 3.1061960083407802,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.029721378819751697,
      "loss": 0.8058,
      "num_input_tokens_seen": 12100080,
      "step": 20855
    },
    {
      "epoch": 3.1069407208817394,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029721004664421197,
      "loss": 0.8125,
      "num_input_tokens_seen": 12102864,
      "step": 20860
    },
    {
      "epoch": 3.1076854334226987,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02972063026039429,
      "loss": 0.7935,
      "num_input_tokens_seen": 12106064,
      "step": 20865
    },
    {
      "epoch": 3.108430145963658,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029720255607677295,
      "loss": 0.8117,
      "num_input_tokens_seen": 12108784,
      "step": 20870
    },
    {
      "epoch": 3.109174858504617,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.029719880706276545,
      "loss": 0.797,
      "num_input_tokens_seen": 12111344,
      "step": 20875
    },
    {
      "epoch": 3.1099195710455763,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.029719505556198365,
      "loss": 0.7946,
      "num_input_tokens_seen": 12114224,
      "step": 20880
    },
    {
      "epoch": 3.1106642835865355,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.029719130157449102,
      "loss": 0.7905,
      "num_input_tokens_seen": 12117328,
      "step": 20885
    },
    {
      "epoch": 3.1114089961274947,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.029718754510035094,
      "loss": 0.818,
      "num_input_tokens_seen": 12120016,
      "step": 20890
    },
    {
      "epoch": 3.112153708668454,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.029718378613962686,
      "loss": 0.8083,
      "num_input_tokens_seen": 12122672,
      "step": 20895
    },
    {
      "epoch": 3.112898421209413,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029718002469238235,
      "loss": 0.7911,
      "num_input_tokens_seen": 12125648,
      "step": 20900
    },
    {
      "epoch": 3.1136431337503723,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.029717626075868087,
      "loss": 0.8066,
      "num_input_tokens_seen": 12128464,
      "step": 20905
    },
    {
      "epoch": 3.1143878462913315,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029717249433858603,
      "loss": 0.7925,
      "num_input_tokens_seen": 12131376,
      "step": 20910
    },
    {
      "epoch": 3.1151325588322907,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029716872543216158,
      "loss": 0.8007,
      "num_input_tokens_seen": 12134672,
      "step": 20915
    },
    {
      "epoch": 3.11587727137325,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029716495403947097,
      "loss": 0.7978,
      "num_input_tokens_seen": 12137616,
      "step": 20920
    },
    {
      "epoch": 3.116621983914209,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02971611801605781,
      "loss": 0.7989,
      "num_input_tokens_seen": 12140496,
      "step": 20925
    },
    {
      "epoch": 3.1173666964551683,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029715740379554666,
      "loss": 0.7975,
      "num_input_tokens_seen": 12143632,
      "step": 20930
    },
    {
      "epoch": 3.1181114089961275,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029715362494444043,
      "loss": 0.7946,
      "num_input_tokens_seen": 12146448,
      "step": 20935
    },
    {
      "epoch": 3.1188561215370867,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.029714984360732326,
      "loss": 0.7886,
      "num_input_tokens_seen": 12149296,
      "step": 20940
    },
    {
      "epoch": 3.119600834078046,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029714605978425904,
      "loss": 0.7921,
      "num_input_tokens_seen": 12152144,
      "step": 20945
    },
    {
      "epoch": 3.120345546619005,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.02971422734753117,
      "loss": 0.7932,
      "num_input_tokens_seen": 12155248,
      "step": 20950
    },
    {
      "epoch": 3.1210902591599643,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02971384846805452,
      "loss": 0.8003,
      "num_input_tokens_seen": 12158192,
      "step": 20955
    },
    {
      "epoch": 3.1218349717009235,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029713469340002352,
      "loss": 0.8074,
      "num_input_tokens_seen": 12161232,
      "step": 20960
    },
    {
      "epoch": 3.1225796842418827,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029713089963381078,
      "loss": 0.8083,
      "num_input_tokens_seen": 12163984,
      "step": 20965
    },
    {
      "epoch": 3.123324396782842,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.0297127103381971,
      "loss": 0.7936,
      "num_input_tokens_seen": 12167152,
      "step": 20970
    },
    {
      "epoch": 3.124069109323801,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029712330464456836,
      "loss": 0.7777,
      "num_input_tokens_seen": 12169968,
      "step": 20975
    },
    {
      "epoch": 3.1248138218647603,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.0297119503421667,
      "loss": 0.7893,
      "num_input_tokens_seen": 12173072,
      "step": 20980
    },
    {
      "epoch": 3.1255585344057195,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.029711569971333115,
      "loss": 0.8247,
      "num_input_tokens_seen": 12176144,
      "step": 20985
    },
    {
      "epoch": 3.1263032469466787,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02971118935196251,
      "loss": 0.782,
      "num_input_tokens_seen": 12178832,
      "step": 20990
    },
    {
      "epoch": 3.127047959487638,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.029710808484061314,
      "loss": 0.7809,
      "num_input_tokens_seen": 12181392,
      "step": 20995
    },
    {
      "epoch": 3.127792672028597,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.029710427367635957,
      "loss": 0.8127,
      "num_input_tokens_seen": 12184272,
      "step": 21000
    },
    {
      "epoch": 3.1285373845695563,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029710046002692878,
      "loss": 0.8282,
      "num_input_tokens_seen": 12187152,
      "step": 21005
    },
    {
      "epoch": 3.1292820971105155,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02970966438923853,
      "loss": 0.8282,
      "num_input_tokens_seen": 12190352,
      "step": 21010
    },
    {
      "epoch": 3.1300268096514747,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029709282527279342,
      "loss": 0.79,
      "num_input_tokens_seen": 12193168,
      "step": 21015
    },
    {
      "epoch": 3.1307715221924335,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.029708900416821785,
      "loss": 0.8291,
      "num_input_tokens_seen": 12195984,
      "step": 21020
    },
    {
      "epoch": 3.131516234733393,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029708518057872297,
      "loss": 0.7945,
      "num_input_tokens_seen": 12198960,
      "step": 21025
    },
    {
      "epoch": 3.132260947274352,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029708135450437354,
      "loss": 0.8169,
      "num_input_tokens_seen": 12201712,
      "step": 21030
    },
    {
      "epoch": 3.133005659815311,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.029707752594523407,
      "loss": 0.8383,
      "num_input_tokens_seen": 12204528,
      "step": 21035
    },
    {
      "epoch": 3.1337503723562703,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02970736949013693,
      "loss": 0.8214,
      "num_input_tokens_seen": 12207376,
      "step": 21040
    },
    {
      "epoch": 3.1344950848972295,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02970698613728439,
      "loss": 0.8142,
      "num_input_tokens_seen": 12210256,
      "step": 21045
    },
    {
      "epoch": 3.1352397974381887,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02970660253597227,
      "loss": 0.81,
      "num_input_tokens_seen": 12213488,
      "step": 21050
    },
    {
      "epoch": 3.135984509979148,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.029706218686207046,
      "loss": 0.8001,
      "num_input_tokens_seen": 12216272,
      "step": 21055
    },
    {
      "epoch": 3.136729222520107,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.0297058345879952,
      "loss": 0.7967,
      "num_input_tokens_seen": 12219120,
      "step": 21060
    },
    {
      "epoch": 3.1374739350610663,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.02970545024134323,
      "loss": 0.7949,
      "num_input_tokens_seen": 12221840,
      "step": 21065
    },
    {
      "epoch": 3.1382186476020255,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02970506564625762,
      "loss": 0.8028,
      "num_input_tokens_seen": 12225136,
      "step": 21070
    },
    {
      "epoch": 3.1389633601429847,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029704680802744878,
      "loss": 0.8136,
      "num_input_tokens_seen": 12227856,
      "step": 21075
    },
    {
      "epoch": 3.139708072683944,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029704295710811495,
      "loss": 0.7932,
      "num_input_tokens_seen": 12230704,
      "step": 21080
    },
    {
      "epoch": 3.140452785224903,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02970391037046398,
      "loss": 0.7958,
      "num_input_tokens_seen": 12233872,
      "step": 21085
    },
    {
      "epoch": 3.1411974977658623,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029703524781708845,
      "loss": 0.7905,
      "num_input_tokens_seen": 12236496,
      "step": 21090
    },
    {
      "epoch": 3.1419422103068215,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029703138944552603,
      "loss": 0.7995,
      "num_input_tokens_seen": 12239216,
      "step": 21095
    },
    {
      "epoch": 3.1426869228477807,
      "grad_norm": 0.015625,
      "learning_rate": 0.02970275285900177,
      "loss": 0.8004,
      "num_input_tokens_seen": 12242032,
      "step": 21100
    },
    {
      "epoch": 3.14343163538874,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029702366525062873,
      "loss": 0.8064,
      "num_input_tokens_seen": 12245232,
      "step": 21105
    },
    {
      "epoch": 3.144176347929699,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.029701979942742436,
      "loss": 0.7988,
      "num_input_tokens_seen": 12247856,
      "step": 21110
    },
    {
      "epoch": 3.1449210604706583,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.029701593112046988,
      "loss": 0.8235,
      "num_input_tokens_seen": 12250480,
      "step": 21115
    },
    {
      "epoch": 3.1456657730116175,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02970120603298307,
      "loss": 0.8026,
      "num_input_tokens_seen": 12253232,
      "step": 21120
    },
    {
      "epoch": 3.1464104855525767,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029700818705557214,
      "loss": 0.793,
      "num_input_tokens_seen": 12256240,
      "step": 21125
    },
    {
      "epoch": 3.147155198093536,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.02970043112977597,
      "loss": 0.7954,
      "num_input_tokens_seen": 12259344,
      "step": 21130
    },
    {
      "epoch": 3.147899910634495,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029700043305645882,
      "loss": 0.8081,
      "num_input_tokens_seen": 12262192,
      "step": 21135
    },
    {
      "epoch": 3.1486446231754543,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029699655233173503,
      "loss": 0.8197,
      "num_input_tokens_seen": 12265136,
      "step": 21140
    },
    {
      "epoch": 3.1493893357164136,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02969926691236539,
      "loss": 0.7956,
      "num_input_tokens_seen": 12268240,
      "step": 21145
    },
    {
      "epoch": 3.1501340482573728,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029698878343228104,
      "loss": 0.7959,
      "num_input_tokens_seen": 12271120,
      "step": 21150
    },
    {
      "epoch": 3.150878760798332,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.0296984895257682,
      "loss": 0.7889,
      "num_input_tokens_seen": 12273776,
      "step": 21155
    },
    {
      "epoch": 3.151623473339291,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02969810045999226,
      "loss": 0.8093,
      "num_input_tokens_seen": 12277008,
      "step": 21160
    },
    {
      "epoch": 3.1523681858802504,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.02969771114590685,
      "loss": 0.7849,
      "num_input_tokens_seen": 12279568,
      "step": 21165
    },
    {
      "epoch": 3.1531128984212096,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02969732158351855,
      "loss": 0.8253,
      "num_input_tokens_seen": 12282352,
      "step": 21170
    },
    {
      "epoch": 3.1538576109621688,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.02969693177283394,
      "loss": 0.8045,
      "num_input_tokens_seen": 12285296,
      "step": 21175
    },
    {
      "epoch": 3.154602323503128,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029696541713859603,
      "loss": 0.7996,
      "num_input_tokens_seen": 12287984,
      "step": 21180
    },
    {
      "epoch": 3.155347036044087,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.029696151406602127,
      "loss": 0.8004,
      "num_input_tokens_seen": 12290928,
      "step": 21185
    },
    {
      "epoch": 3.1560917485850464,
      "grad_norm": 0.02734375,
      "learning_rate": 0.029695760851068113,
      "loss": 0.7976,
      "num_input_tokens_seen": 12294192,
      "step": 21190
    },
    {
      "epoch": 3.156836461126005,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029695370047264157,
      "loss": 0.7981,
      "num_input_tokens_seen": 12296976,
      "step": 21195
    },
    {
      "epoch": 3.157581173666965,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02969497899519686,
      "loss": 0.7681,
      "num_input_tokens_seen": 12299984,
      "step": 21200
    },
    {
      "epoch": 3.1583258862079235,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029694587694872827,
      "loss": 0.8084,
      "num_input_tokens_seen": 12303152,
      "step": 21205
    },
    {
      "epoch": 3.1590705987488827,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02969419614629867,
      "loss": 0.7661,
      "num_input_tokens_seen": 12306032,
      "step": 21210
    },
    {
      "epoch": 3.159815311289842,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029693804349481,
      "loss": 0.797,
      "num_input_tokens_seen": 12308656,
      "step": 21215
    },
    {
      "epoch": 3.160560023830801,
      "grad_norm": 0.029296875,
      "learning_rate": 0.029693412304426445,
      "loss": 0.7726,
      "num_input_tokens_seen": 12311600,
      "step": 21220
    },
    {
      "epoch": 3.1613047363717603,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.02969302001114162,
      "loss": 0.8402,
      "num_input_tokens_seen": 12314416,
      "step": 21225
    },
    {
      "epoch": 3.1620494489127196,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.029692627469633158,
      "loss": 0.8017,
      "num_input_tokens_seen": 12317520,
      "step": 21230
    },
    {
      "epoch": 3.1627941614536788,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.029692234679907684,
      "loss": 0.8306,
      "num_input_tokens_seen": 12320496,
      "step": 21235
    },
    {
      "epoch": 3.163538873994638,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02969184164197184,
      "loss": 0.7907,
      "num_input_tokens_seen": 12323856,
      "step": 21240
    },
    {
      "epoch": 3.164283586535597,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02969144835583226,
      "loss": 0.7879,
      "num_input_tokens_seen": 12326736,
      "step": 21245
    },
    {
      "epoch": 3.1650282990765564,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029691054821495595,
      "loss": 0.8249,
      "num_input_tokens_seen": 12329456,
      "step": 21250
    },
    {
      "epoch": 3.1657730116175156,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.02969066103896849,
      "loss": 0.8153,
      "num_input_tokens_seen": 12332240,
      "step": 21255
    },
    {
      "epoch": 3.1665177241584748,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.029690267008257595,
      "loss": 0.8133,
      "num_input_tokens_seen": 12335120,
      "step": 21260
    },
    {
      "epoch": 3.167262436699434,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029689872729369574,
      "loss": 0.7894,
      "num_input_tokens_seen": 12338384,
      "step": 21265
    },
    {
      "epoch": 3.168007149240393,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029689478202311083,
      "loss": 0.8123,
      "num_input_tokens_seen": 12341200,
      "step": 21270
    },
    {
      "epoch": 3.1687518617813524,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.029689083427088783,
      "loss": 0.7965,
      "num_input_tokens_seen": 12344080,
      "step": 21275
    },
    {
      "epoch": 3.1694965743223116,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02968868840370935,
      "loss": 0.7908,
      "num_input_tokens_seen": 12347120,
      "step": 21280
    },
    {
      "epoch": 3.170241286863271,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029688293132179456,
      "loss": 0.7806,
      "num_input_tokens_seen": 12350128,
      "step": 21285
    },
    {
      "epoch": 3.17098599940423,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02968789761250578,
      "loss": 0.796,
      "num_input_tokens_seen": 12352848,
      "step": 21290
    },
    {
      "epoch": 3.171730711945189,
      "grad_norm": 0.015625,
      "learning_rate": 0.029687501844695,
      "loss": 0.7968,
      "num_input_tokens_seen": 12355920,
      "step": 21295
    },
    {
      "epoch": 3.1724754244861484,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029687105828753804,
      "loss": 0.8106,
      "num_input_tokens_seen": 12359120,
      "step": 21300
    },
    {
      "epoch": 3.1732201370271076,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029686709564688878,
      "loss": 0.8074,
      "num_input_tokens_seen": 12361808,
      "step": 21305
    },
    {
      "epoch": 3.173964849568067,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.029686313052506928,
      "loss": 0.7908,
      "num_input_tokens_seen": 12364816,
      "step": 21310
    },
    {
      "epoch": 3.174709562109026,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.029685916292214645,
      "loss": 0.7775,
      "num_input_tokens_seen": 12367696,
      "step": 21315
    },
    {
      "epoch": 3.175454274649985,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.029685519283818726,
      "loss": 0.8055,
      "num_input_tokens_seen": 12370704,
      "step": 21320
    },
    {
      "epoch": 3.1761989871909444,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.02968512202732589,
      "loss": 0.7918,
      "num_input_tokens_seen": 12373584,
      "step": 21325
    },
    {
      "epoch": 3.1769436997319036,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.02968472452274284,
      "loss": 0.7896,
      "num_input_tokens_seen": 12376912,
      "step": 21330
    },
    {
      "epoch": 3.177688412272863,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029684326770076295,
      "loss": 0.8314,
      "num_input_tokens_seen": 12379920,
      "step": 21335
    },
    {
      "epoch": 3.178433124813822,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029683928769332974,
      "loss": 0.8021,
      "num_input_tokens_seen": 12382864,
      "step": 21340
    },
    {
      "epoch": 3.179177837354781,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029683530520519603,
      "loss": 0.8319,
      "num_input_tokens_seen": 12385744,
      "step": 21345
    },
    {
      "epoch": 3.1799225498957404,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.0296831320236429,
      "loss": 0.7937,
      "num_input_tokens_seen": 12388624,
      "step": 21350
    },
    {
      "epoch": 3.1806672624366996,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.02968273327870961,
      "loss": 0.7954,
      "num_input_tokens_seen": 12391408,
      "step": 21355
    },
    {
      "epoch": 3.181411974977659,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029682334285726465,
      "loss": 0.7958,
      "num_input_tokens_seen": 12394320,
      "step": 21360
    },
    {
      "epoch": 3.182156687518618,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.0296819350447002,
      "loss": 0.8357,
      "num_input_tokens_seen": 12397264,
      "step": 21365
    },
    {
      "epoch": 3.182901400059577,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029681535555637573,
      "loss": 0.8059,
      "num_input_tokens_seen": 12399984,
      "step": 21370
    },
    {
      "epoch": 3.1836461126005364,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.02968113581854532,
      "loss": 0.8141,
      "num_input_tokens_seen": 12402832,
      "step": 21375
    },
    {
      "epoch": 3.184390825141495,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029680735833430196,
      "loss": 0.7999,
      "num_input_tokens_seen": 12405712,
      "step": 21380
    },
    {
      "epoch": 3.1851355376824544,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02968033560029896,
      "loss": 0.7978,
      "num_input_tokens_seen": 12408752,
      "step": 21385
    },
    {
      "epoch": 3.1858802502234136,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029679935119158375,
      "loss": 0.8055,
      "num_input_tokens_seen": 12411760,
      "step": 21390
    },
    {
      "epoch": 3.186624962764373,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029679534390015208,
      "loss": 0.8135,
      "num_input_tokens_seen": 12414224,
      "step": 21395
    },
    {
      "epoch": 3.187369675305332,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.029679133412876226,
      "loss": 0.815,
      "num_input_tokens_seen": 12417296,
      "step": 21400
    },
    {
      "epoch": 3.188114387846291,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02967873218774821,
      "loss": 0.7924,
      "num_input_tokens_seen": 12420240,
      "step": 21405
    },
    {
      "epoch": 3.1888591003872504,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029678330714637928,
      "loss": 0.7989,
      "num_input_tokens_seen": 12422928,
      "step": 21410
    },
    {
      "epoch": 3.1896038129282096,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.029677928993552167,
      "loss": 0.8001,
      "num_input_tokens_seen": 12425936,
      "step": 21415
    },
    {
      "epoch": 3.190348525469169,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02967752702449771,
      "loss": 0.8159,
      "num_input_tokens_seen": 12429040,
      "step": 21420
    },
    {
      "epoch": 3.191093238010128,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02967712480748136,
      "loss": 0.8129,
      "num_input_tokens_seen": 12431952,
      "step": 21425
    },
    {
      "epoch": 3.191837950551087,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029676722342509895,
      "loss": 0.8242,
      "num_input_tokens_seen": 12434800,
      "step": 21430
    },
    {
      "epoch": 3.1925826630920464,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02967631962959013,
      "loss": 0.8067,
      "num_input_tokens_seen": 12437744,
      "step": 21435
    },
    {
      "epoch": 3.1933273756330056,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029675916668728857,
      "loss": 0.7893,
      "num_input_tokens_seen": 12440496,
      "step": 21440
    },
    {
      "epoch": 3.194072088173965,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029675513459932888,
      "loss": 0.7854,
      "num_input_tokens_seen": 12443504,
      "step": 21445
    },
    {
      "epoch": 3.194816800714924,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029675110003209035,
      "loss": 0.8123,
      "num_input_tokens_seen": 12446352,
      "step": 21450
    },
    {
      "epoch": 3.1955615132558832,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029674706298564113,
      "loss": 0.7985,
      "num_input_tokens_seen": 12449392,
      "step": 21455
    },
    {
      "epoch": 3.1963062257968424,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.029674302346004943,
      "loss": 0.8149,
      "num_input_tokens_seen": 12452336,
      "step": 21460
    },
    {
      "epoch": 3.1970509383378016,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02967389814553835,
      "loss": 0.824,
      "num_input_tokens_seen": 12455280,
      "step": 21465
    },
    {
      "epoch": 3.197795650878761,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02967349369717116,
      "loss": 0.7976,
      "num_input_tokens_seen": 12458384,
      "step": 21470
    },
    {
      "epoch": 3.19854036341972,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02967308900091021,
      "loss": 0.8134,
      "num_input_tokens_seen": 12461136,
      "step": 21475
    },
    {
      "epoch": 3.1992850759606792,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.029672684056762332,
      "loss": 0.7986,
      "num_input_tokens_seen": 12463856,
      "step": 21480
    },
    {
      "epoch": 3.2000297885016384,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.029672278864734374,
      "loss": 0.801,
      "num_input_tokens_seen": 12466704,
      "step": 21485
    },
    {
      "epoch": 3.2007745010425976,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02967187342483317,
      "loss": 0.8019,
      "num_input_tokens_seen": 12469744,
      "step": 21490
    },
    {
      "epoch": 3.201519213583557,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029671467737065582,
      "loss": 0.8013,
      "num_input_tokens_seen": 12472432,
      "step": 21495
    },
    {
      "epoch": 3.202263926124516,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029671061801438456,
      "loss": 0.8106,
      "num_input_tokens_seen": 12475312,
      "step": 21500
    },
    {
      "epoch": 3.2030086386654752,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.02967065561795865,
      "loss": 0.8024,
      "num_input_tokens_seen": 12478448,
      "step": 21505
    },
    {
      "epoch": 3.2037533512064345,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02967024918663303,
      "loss": 0.7906,
      "num_input_tokens_seen": 12481520,
      "step": 21510
    },
    {
      "epoch": 3.2044980637473937,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029669842507468458,
      "loss": 0.8216,
      "num_input_tokens_seen": 12484656,
      "step": 21515
    },
    {
      "epoch": 3.205242776288353,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02966943558047181,
      "loss": 0.8073,
      "num_input_tokens_seen": 12487376,
      "step": 21520
    },
    {
      "epoch": 3.205987488829312,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029669028405649955,
      "loss": 0.7942,
      "num_input_tokens_seen": 12490480,
      "step": 21525
    },
    {
      "epoch": 3.2067322013702713,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.029668620983009777,
      "loss": 0.8112,
      "num_input_tokens_seen": 12493360,
      "step": 21530
    },
    {
      "epoch": 3.2074769139112305,
      "grad_norm": 0.025390625,
      "learning_rate": 0.029668213312558152,
      "loss": 0.7892,
      "num_input_tokens_seen": 12496272,
      "step": 21535
    },
    {
      "epoch": 3.2082216264521897,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02966780539430197,
      "loss": 0.8089,
      "num_input_tokens_seen": 12499152,
      "step": 21540
    },
    {
      "epoch": 3.2089663389931484,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.02966739722824813,
      "loss": 0.8046,
      "num_input_tokens_seen": 12502320,
      "step": 21545
    },
    {
      "epoch": 3.2097110515341076,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029666988814403515,
      "loss": 0.8022,
      "num_input_tokens_seen": 12505008,
      "step": 21550
    },
    {
      "epoch": 3.210455764075067,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.029666580152775037,
      "loss": 0.7985,
      "num_input_tokens_seen": 12508016,
      "step": 21555
    },
    {
      "epoch": 3.211200476616026,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.029666171243369586,
      "loss": 0.7977,
      "num_input_tokens_seen": 12510896,
      "step": 21560
    },
    {
      "epoch": 3.2119451891569852,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029665762086194085,
      "loss": 0.7737,
      "num_input_tokens_seen": 12513872,
      "step": 21565
    },
    {
      "epoch": 3.2126899016979444,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.029665352681255436,
      "loss": 0.7865,
      "num_input_tokens_seen": 12516816,
      "step": 21570
    },
    {
      "epoch": 3.2134346142389036,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029664943028560556,
      "loss": 0.8033,
      "num_input_tokens_seen": 12519632,
      "step": 21575
    },
    {
      "epoch": 3.214179326779863,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02966453312811637,
      "loss": 0.8058,
      "num_input_tokens_seen": 12522352,
      "step": 21580
    },
    {
      "epoch": 3.214924039320822,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029664122979929805,
      "loss": 0.793,
      "num_input_tokens_seen": 12525104,
      "step": 21585
    },
    {
      "epoch": 3.2156687518617812,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02966371258400778,
      "loss": 0.7711,
      "num_input_tokens_seen": 12527920,
      "step": 21590
    },
    {
      "epoch": 3.2164134644027405,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02966330194035724,
      "loss": 0.7951,
      "num_input_tokens_seen": 12530800,
      "step": 21595
    },
    {
      "epoch": 3.2171581769436997,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029662891048985115,
      "loss": 0.8237,
      "num_input_tokens_seen": 12533936,
      "step": 21600
    },
    {
      "epoch": 3.217902889484659,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02966247990989835,
      "loss": 0.815,
      "num_input_tokens_seen": 12536656,
      "step": 21605
    },
    {
      "epoch": 3.218647602025618,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029662068523103886,
      "loss": 0.7835,
      "num_input_tokens_seen": 12539600,
      "step": 21610
    },
    {
      "epoch": 3.2193923145665773,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029661656888608678,
      "loss": 0.8154,
      "num_input_tokens_seen": 12542352,
      "step": 21615
    },
    {
      "epoch": 3.2201370271075365,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02966124500641968,
      "loss": 0.8243,
      "num_input_tokens_seen": 12545296,
      "step": 21620
    },
    {
      "epoch": 3.2208817396484957,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029660832876543843,
      "loss": 0.8201,
      "num_input_tokens_seen": 12548240,
      "step": 21625
    },
    {
      "epoch": 3.221626452189455,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02966042049898814,
      "loss": 0.8154,
      "num_input_tokens_seen": 12551152,
      "step": 21630
    },
    {
      "epoch": 3.222371164730414,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.029660007873759536,
      "loss": 0.7994,
      "num_input_tokens_seen": 12553840,
      "step": 21635
    },
    {
      "epoch": 3.2231158772713733,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02965959500086499,
      "loss": 0.8069,
      "num_input_tokens_seen": 12556976,
      "step": 21640
    },
    {
      "epoch": 3.2238605898123325,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029659181880311487,
      "loss": 0.8012,
      "num_input_tokens_seen": 12559664,
      "step": 21645
    },
    {
      "epoch": 3.2246053023532917,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02965876851210601,
      "loss": 0.802,
      "num_input_tokens_seen": 12562704,
      "step": 21650
    },
    {
      "epoch": 3.225350014894251,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029658354896255536,
      "loss": 0.7952,
      "num_input_tokens_seen": 12565392,
      "step": 21655
    },
    {
      "epoch": 3.22609472743521,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.029657941032767053,
      "loss": 0.8031,
      "num_input_tokens_seen": 12568208,
      "step": 21660
    },
    {
      "epoch": 3.2268394399761693,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029657526921647557,
      "loss": 0.8142,
      "num_input_tokens_seen": 12571088,
      "step": 21665
    },
    {
      "epoch": 3.2275841525171285,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.02965711256290404,
      "loss": 0.7976,
      "num_input_tokens_seen": 12573552,
      "step": 21670
    },
    {
      "epoch": 3.2283288650580877,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029656697956543496,
      "loss": 0.8067,
      "num_input_tokens_seen": 12576752,
      "step": 21675
    },
    {
      "epoch": 3.229073577599047,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029656283102572945,
      "loss": 0.7962,
      "num_input_tokens_seen": 12579536,
      "step": 21680
    },
    {
      "epoch": 3.229818290140006,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029655868000999383,
      "loss": 0.8002,
      "num_input_tokens_seen": 12582384,
      "step": 21685
    },
    {
      "epoch": 3.2305630026809653,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.02965545265182983,
      "loss": 0.8069,
      "num_input_tokens_seen": 12585616,
      "step": 21690
    },
    {
      "epoch": 3.2313077152219245,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.029655037055071293,
      "loss": 0.7832,
      "num_input_tokens_seen": 12588304,
      "step": 21695
    },
    {
      "epoch": 3.2320524277628837,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.029654621210730807,
      "loss": 0.7994,
      "num_input_tokens_seen": 12591120,
      "step": 21700
    },
    {
      "epoch": 3.232797140303843,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.029654205118815388,
      "loss": 0.8325,
      "num_input_tokens_seen": 12593648,
      "step": 21705
    },
    {
      "epoch": 3.233541852844802,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029653788779332065,
      "loss": 0.8066,
      "num_input_tokens_seen": 12596496,
      "step": 21710
    },
    {
      "epoch": 3.2342865653857613,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.029653372192287874,
      "loss": 0.8063,
      "num_input_tokens_seen": 12599248,
      "step": 21715
    },
    {
      "epoch": 3.23503127792672,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.029652955357689855,
      "loss": 0.8098,
      "num_input_tokens_seen": 12601904,
      "step": 21720
    },
    {
      "epoch": 3.2357759904676793,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029652538275545048,
      "loss": 0.798,
      "num_input_tokens_seen": 12604880,
      "step": 21725
    },
    {
      "epoch": 3.2365207030086385,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029652120945860498,
      "loss": 0.8068,
      "num_input_tokens_seen": 12607824,
      "step": 21730
    },
    {
      "epoch": 3.2372654155495977,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029651703368643253,
      "loss": 0.8078,
      "num_input_tokens_seen": 12610800,
      "step": 21735
    },
    {
      "epoch": 3.238010128090557,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029651285543900376,
      "loss": 0.7906,
      "num_input_tokens_seen": 12613616,
      "step": 21740
    },
    {
      "epoch": 3.238754840631516,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02965086747163892,
      "loss": 0.7907,
      "num_input_tokens_seen": 12616880,
      "step": 21745
    },
    {
      "epoch": 3.2394995531724753,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.029650449151865946,
      "loss": 0.8176,
      "num_input_tokens_seen": 12619792,
      "step": 21750
    },
    {
      "epoch": 3.2402442657134345,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.029650030584588524,
      "loss": 0.7776,
      "num_input_tokens_seen": 12622576,
      "step": 21755
    },
    {
      "epoch": 3.2409889782543937,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029649611769813726,
      "loss": 0.7804,
      "num_input_tokens_seen": 12625456,
      "step": 21760
    },
    {
      "epoch": 3.241733690795353,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02964919270754863,
      "loss": 0.7791,
      "num_input_tokens_seen": 12628400,
      "step": 21765
    },
    {
      "epoch": 3.242478403336312,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029648773397800306,
      "loss": 0.7998,
      "num_input_tokens_seen": 12631248,
      "step": 21770
    },
    {
      "epoch": 3.2432231158772713,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.029648353840575844,
      "loss": 0.7885,
      "num_input_tokens_seen": 12634192,
      "step": 21775
    },
    {
      "epoch": 3.2439678284182305,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.029647934035882336,
      "loss": 0.8044,
      "num_input_tokens_seen": 12636880,
      "step": 21780
    },
    {
      "epoch": 3.2447125409591897,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029647513983726868,
      "loss": 0.8009,
      "num_input_tokens_seen": 12640016,
      "step": 21785
    },
    {
      "epoch": 3.245457253500149,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.029647093684116537,
      "loss": 0.7945,
      "num_input_tokens_seen": 12643120,
      "step": 21790
    },
    {
      "epoch": 3.246201966041108,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029646673137058446,
      "loss": 0.8067,
      "num_input_tokens_seen": 12646000,
      "step": 21795
    },
    {
      "epoch": 3.2469466785820673,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.0296462523425597,
      "loss": 0.7817,
      "num_input_tokens_seen": 12648784,
      "step": 21800
    },
    {
      "epoch": 3.2476913911230265,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.029645831300627402,
      "loss": 0.817,
      "num_input_tokens_seen": 12651632,
      "step": 21805
    },
    {
      "epoch": 3.2484361036639857,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029645410011268674,
      "loss": 0.7973,
      "num_input_tokens_seen": 12654256,
      "step": 21810
    },
    {
      "epoch": 3.249180816204945,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.029644988474490628,
      "loss": 0.8217,
      "num_input_tokens_seen": 12657104,
      "step": 21815
    },
    {
      "epoch": 3.249925528745904,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029644566690300383,
      "loss": 0.7899,
      "num_input_tokens_seen": 12659728,
      "step": 21820
    },
    {
      "epoch": 3.2506702412868633,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.029644144658705074,
      "loss": 0.7822,
      "num_input_tokens_seen": 12662768,
      "step": 21825
    },
    {
      "epoch": 3.2514149538278225,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02964372237971182,
      "loss": 0.8036,
      "num_input_tokens_seen": 12665776,
      "step": 21830
    },
    {
      "epoch": 3.2521596663687817,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02964329985332776,
      "loss": 0.7991,
      "num_input_tokens_seen": 12668592,
      "step": 21835
    },
    {
      "epoch": 3.252904378909741,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02964287707956003,
      "loss": 0.82,
      "num_input_tokens_seen": 12671664,
      "step": 21840
    },
    {
      "epoch": 3.2536490914507,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.029642454058415778,
      "loss": 0.7987,
      "num_input_tokens_seen": 12674512,
      "step": 21845
    },
    {
      "epoch": 3.2543938039916593,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.029642030789902147,
      "loss": 0.7706,
      "num_input_tokens_seen": 12677520,
      "step": 21850
    },
    {
      "epoch": 3.2551385165326185,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029641607274026285,
      "loss": 0.8081,
      "num_input_tokens_seen": 12680304,
      "step": 21855
    },
    {
      "epoch": 3.2558832290735777,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02964118351079535,
      "loss": 0.8143,
      "num_input_tokens_seen": 12683280,
      "step": 21860
    },
    {
      "epoch": 3.256627941614537,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.029640759500216496,
      "loss": 0.8283,
      "num_input_tokens_seen": 12686352,
      "step": 21865
    },
    {
      "epoch": 3.257372654155496,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029640335242296895,
      "loss": 0.7921,
      "num_input_tokens_seen": 12689424,
      "step": 21870
    },
    {
      "epoch": 3.2581173666964554,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029639910737043713,
      "loss": 0.8247,
      "num_input_tokens_seen": 12692304,
      "step": 21875
    },
    {
      "epoch": 3.2588620792374146,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02963948598446411,
      "loss": 0.7873,
      "num_input_tokens_seen": 12695184,
      "step": 21880
    },
    {
      "epoch": 3.2596067917783733,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029639060984565275,
      "loss": 0.8113,
      "num_input_tokens_seen": 12697808,
      "step": 21885
    },
    {
      "epoch": 3.260351504319333,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029638635737354386,
      "loss": 0.7801,
      "num_input_tokens_seen": 12700720,
      "step": 21890
    },
    {
      "epoch": 3.2610962168602917,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.02963821024283862,
      "loss": 0.8095,
      "num_input_tokens_seen": 12703472,
      "step": 21895
    },
    {
      "epoch": 3.2618409294012514,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02963778450102517,
      "loss": 0.8075,
      "num_input_tokens_seen": 12707152,
      "step": 21900
    },
    {
      "epoch": 3.26258564194221,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02963735851192123,
      "loss": 0.8044,
      "num_input_tokens_seen": 12710192,
      "step": 21905
    },
    {
      "epoch": 3.2633303544831693,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02963693227553399,
      "loss": 0.814,
      "num_input_tokens_seen": 12713264,
      "step": 21910
    },
    {
      "epoch": 3.2640750670241285,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029636505791870663,
      "loss": 0.7795,
      "num_input_tokens_seen": 12716112,
      "step": 21915
    },
    {
      "epoch": 3.2648197795650877,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02963607906093844,
      "loss": 0.8099,
      "num_input_tokens_seen": 12718736,
      "step": 21920
    },
    {
      "epoch": 3.265564492106047,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02963565208274454,
      "loss": 0.8225,
      "num_input_tokens_seen": 12721680,
      "step": 21925
    },
    {
      "epoch": 3.266309204647006,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029635224857296173,
      "loss": 0.7968,
      "num_input_tokens_seen": 12724496,
      "step": 21930
    },
    {
      "epoch": 3.2670539171879653,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.029634797384600554,
      "loss": 0.8065,
      "num_input_tokens_seen": 12727440,
      "step": 21935
    },
    {
      "epoch": 3.2677986297289245,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.02963436966466491,
      "loss": 0.8278,
      "num_input_tokens_seen": 12730320,
      "step": 21940
    },
    {
      "epoch": 3.2685433422698837,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029633941697496464,
      "loss": 0.8235,
      "num_input_tokens_seen": 12733264,
      "step": 21945
    },
    {
      "epoch": 3.269288054810843,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029633513483102444,
      "loss": 0.7802,
      "num_input_tokens_seen": 12736144,
      "step": 21950
    },
    {
      "epoch": 3.270032767351802,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029633085021490087,
      "loss": 0.7983,
      "num_input_tokens_seen": 12739184,
      "step": 21955
    },
    {
      "epoch": 3.2707774798927614,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.029632656312666632,
      "loss": 0.7993,
      "num_input_tokens_seen": 12742224,
      "step": 21960
    },
    {
      "epoch": 3.2715221924337206,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029632227356639323,
      "loss": 0.8111,
      "num_input_tokens_seen": 12745232,
      "step": 21965
    },
    {
      "epoch": 3.2722669049746798,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.029631798153415403,
      "loss": 0.8163,
      "num_input_tokens_seen": 12747984,
      "step": 21970
    },
    {
      "epoch": 3.273011617515639,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029631368703002123,
      "loss": 0.8033,
      "num_input_tokens_seen": 12751184,
      "step": 21975
    },
    {
      "epoch": 3.273756330056598,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02963093900540674,
      "loss": 0.7936,
      "num_input_tokens_seen": 12753744,
      "step": 21980
    },
    {
      "epoch": 3.2745010425975574,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.02963050906063651,
      "loss": 0.8066,
      "num_input_tokens_seen": 12756848,
      "step": 21985
    },
    {
      "epoch": 3.2752457551385166,
      "grad_norm": 0.01190185546875,
      "learning_rate": 0.029630078868698704,
      "loss": 0.821,
      "num_input_tokens_seen": 12759728,
      "step": 21990
    },
    {
      "epoch": 3.2759904676794758,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02962964842960058,
      "loss": 0.8073,
      "num_input_tokens_seen": 12762576,
      "step": 21995
    },
    {
      "epoch": 3.276735180220435,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029629217743349416,
      "loss": 0.8035,
      "num_input_tokens_seen": 12765520,
      "step": 22000
    },
    {
      "epoch": 3.277479892761394,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02962878680995249,
      "loss": 0.8135,
      "num_input_tokens_seen": 12768336,
      "step": 22005
    },
    {
      "epoch": 3.2782246053023534,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02962835562941708,
      "loss": 0.7898,
      "num_input_tokens_seen": 12771184,
      "step": 22010
    },
    {
      "epoch": 3.2789693178433126,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029627924201750463,
      "loss": 0.7988,
      "num_input_tokens_seen": 12774320,
      "step": 22015
    },
    {
      "epoch": 3.279714030384272,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029627492526959936,
      "loss": 0.796,
      "num_input_tokens_seen": 12777008,
      "step": 22020
    },
    {
      "epoch": 3.280458742925231,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.029627060605052793,
      "loss": 0.7879,
      "num_input_tokens_seen": 12780240,
      "step": 22025
    },
    {
      "epoch": 3.28120345546619,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029626628436036324,
      "loss": 0.8076,
      "num_input_tokens_seen": 12783056,
      "step": 22030
    },
    {
      "epoch": 3.2819481680071494,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029626196019917835,
      "loss": 0.7818,
      "num_input_tokens_seen": 12785712,
      "step": 22035
    },
    {
      "epoch": 3.2826928805481086,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.029625763356704627,
      "loss": 0.8194,
      "num_input_tokens_seen": 12788496,
      "step": 22040
    },
    {
      "epoch": 3.283437593089068,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029625330446404013,
      "loss": 0.8671,
      "num_input_tokens_seen": 12791088,
      "step": 22045
    },
    {
      "epoch": 3.284182305630027,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02962489728902331,
      "loss": 0.8054,
      "num_input_tokens_seen": 12794064,
      "step": 22050
    },
    {
      "epoch": 3.284927018170986,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.029624463884569824,
      "loss": 0.821,
      "num_input_tokens_seen": 12796848,
      "step": 22055
    },
    {
      "epoch": 3.285671730711945,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02962403023305089,
      "loss": 0.8041,
      "num_input_tokens_seen": 12799824,
      "step": 22060
    },
    {
      "epoch": 3.2864164432529046,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02962359633447383,
      "loss": 0.8181,
      "num_input_tokens_seen": 12802928,
      "step": 22065
    },
    {
      "epoch": 3.2871611557938634,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.029623162188845968,
      "loss": 0.7962,
      "num_input_tokens_seen": 12805456,
      "step": 22070
    },
    {
      "epoch": 3.2879058683348226,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02962272779617464,
      "loss": 0.8073,
      "num_input_tokens_seen": 12808112,
      "step": 22075
    },
    {
      "epoch": 3.2886505808757818,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029622293156467194,
      "loss": 0.8079,
      "num_input_tokens_seen": 12811088,
      "step": 22080
    },
    {
      "epoch": 3.289395293416741,
      "grad_norm": 0.041015625,
      "learning_rate": 0.029621858269730968,
      "loss": 0.812,
      "num_input_tokens_seen": 12813872,
      "step": 22085
    },
    {
      "epoch": 3.2901400059577,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.0296214231359733,
      "loss": 0.8101,
      "num_input_tokens_seen": 12816688,
      "step": 22090
    },
    {
      "epoch": 3.2908847184986594,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.029620987755201552,
      "loss": 0.8147,
      "num_input_tokens_seen": 12819760,
      "step": 22095
    },
    {
      "epoch": 3.2916294310396186,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02962055212742308,
      "loss": 0.7993,
      "num_input_tokens_seen": 12822736,
      "step": 22100
    },
    {
      "epoch": 3.292374143580578,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029620116252645236,
      "loss": 0.7916,
      "num_input_tokens_seen": 12825360,
      "step": 22105
    },
    {
      "epoch": 3.293118856121537,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029619680130875386,
      "loss": 0.805,
      "num_input_tokens_seen": 12828400,
      "step": 22110
    },
    {
      "epoch": 3.293863568662496,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.0296192437621209,
      "loss": 0.8006,
      "num_input_tokens_seen": 12831312,
      "step": 22115
    },
    {
      "epoch": 3.2946082812034554,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029618807146389155,
      "loss": 0.7896,
      "num_input_tokens_seen": 12834128,
      "step": 22120
    },
    {
      "epoch": 3.2953529937444146,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029618370283687512,
      "loss": 0.8136,
      "num_input_tokens_seen": 12837360,
      "step": 22125
    },
    {
      "epoch": 3.296097706285374,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.02961793317402336,
      "loss": 0.8021,
      "num_input_tokens_seen": 12840496,
      "step": 22130
    },
    {
      "epoch": 3.296842418826333,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029617495817404093,
      "loss": 0.8023,
      "num_input_tokens_seen": 12843472,
      "step": 22135
    },
    {
      "epoch": 3.297587131367292,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.029617058213837085,
      "loss": 0.7941,
      "num_input_tokens_seen": 12846576,
      "step": 22140
    },
    {
      "epoch": 3.2983318439082514,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029616620363329734,
      "loss": 0.8129,
      "num_input_tokens_seen": 12849264,
      "step": 22145
    },
    {
      "epoch": 3.2990765564492106,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.029616182265889444,
      "loss": 0.8095,
      "num_input_tokens_seen": 12852176,
      "step": 22150
    },
    {
      "epoch": 3.29982126899017,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.0296157439215236,
      "loss": 0.8113,
      "num_input_tokens_seen": 12854992,
      "step": 22155
    },
    {
      "epoch": 3.300565981531129,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.029615305330239622,
      "loss": 0.8022,
      "num_input_tokens_seen": 12857872,
      "step": 22160
    },
    {
      "epoch": 3.301310694072088,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029614866492044915,
      "loss": 0.7964,
      "num_input_tokens_seen": 12860752,
      "step": 22165
    },
    {
      "epoch": 3.3020554066130474,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.02961442740694689,
      "loss": 0.8053,
      "num_input_tokens_seen": 12863440,
      "step": 22170
    },
    {
      "epoch": 3.3028001191540066,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02961398807495297,
      "loss": 0.7871,
      "num_input_tokens_seen": 12866224,
      "step": 22175
    },
    {
      "epoch": 3.303544831694966,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029613548496070577,
      "loss": 0.7953,
      "num_input_tokens_seen": 12869040,
      "step": 22180
    },
    {
      "epoch": 3.304289544235925,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029613108670307135,
      "loss": 0.816,
      "num_input_tokens_seen": 12871888,
      "step": 22185
    },
    {
      "epoch": 3.3050342567768842,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02961266859767007,
      "loss": 0.7864,
      "num_input_tokens_seen": 12874576,
      "step": 22190
    },
    {
      "epoch": 3.3057789693178434,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029612228278166817,
      "loss": 0.8195,
      "num_input_tokens_seen": 12877456,
      "step": 22195
    },
    {
      "epoch": 3.3065236818588026,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.029611787711804824,
      "loss": 0.8056,
      "num_input_tokens_seen": 12880592,
      "step": 22200
    },
    {
      "epoch": 3.307268394399762,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02961134689859153,
      "loss": 0.8096,
      "num_input_tokens_seen": 12883472,
      "step": 22205
    },
    {
      "epoch": 3.308013106940721,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.029610905838534377,
      "loss": 0.807,
      "num_input_tokens_seen": 12886032,
      "step": 22210
    },
    {
      "epoch": 3.3087578194816802,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.02961046453164082,
      "loss": 0.7944,
      "num_input_tokens_seen": 12889360,
      "step": 22215
    },
    {
      "epoch": 3.3095025320226394,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029610022977918316,
      "loss": 0.7991,
      "num_input_tokens_seen": 12892048,
      "step": 22220
    },
    {
      "epoch": 3.310247244563598,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.02960958117737432,
      "loss": 0.8006,
      "num_input_tokens_seen": 12894896,
      "step": 22225
    },
    {
      "epoch": 3.310991957104558,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.029609139130016303,
      "loss": 0.787,
      "num_input_tokens_seen": 12897648,
      "step": 22230
    },
    {
      "epoch": 3.3117366696455166,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029608696835851725,
      "loss": 0.7696,
      "num_input_tokens_seen": 12900272,
      "step": 22235
    },
    {
      "epoch": 3.3124813821864763,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029608254294888062,
      "loss": 0.8079,
      "num_input_tokens_seen": 12903344,
      "step": 22240
    },
    {
      "epoch": 3.313226094727435,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.029607811507132792,
      "loss": 0.7818,
      "num_input_tokens_seen": 12905904,
      "step": 22245
    },
    {
      "epoch": 3.313970807268394,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02960736847259339,
      "loss": 0.7968,
      "num_input_tokens_seen": 12908752,
      "step": 22250
    },
    {
      "epoch": 3.3147155198093534,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.029606925191277347,
      "loss": 0.8064,
      "num_input_tokens_seen": 12911664,
      "step": 22255
    },
    {
      "epoch": 3.3154602323503126,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029606481663192147,
      "loss": 0.7883,
      "num_input_tokens_seen": 12914512,
      "step": 22260
    },
    {
      "epoch": 3.316204944891272,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029606037888345284,
      "loss": 0.8021,
      "num_input_tokens_seen": 12917232,
      "step": 22265
    },
    {
      "epoch": 3.316949657432231,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029605593866744256,
      "loss": 0.8132,
      "num_input_tokens_seen": 12920016,
      "step": 22270
    },
    {
      "epoch": 3.3176943699731902,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029605149598396566,
      "loss": 0.7917,
      "num_input_tokens_seen": 12923024,
      "step": 22275
    },
    {
      "epoch": 3.3184390825141494,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029604705083309715,
      "loss": 0.8085,
      "num_input_tokens_seen": 12925872,
      "step": 22280
    },
    {
      "epoch": 3.3191837950551086,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029604260321491216,
      "loss": 0.8204,
      "num_input_tokens_seen": 12928720,
      "step": 22285
    },
    {
      "epoch": 3.319928507596068,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.029603815312948582,
      "loss": 0.7887,
      "num_input_tokens_seen": 12931600,
      "step": 22290
    },
    {
      "epoch": 3.320673220137027,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02960337005768933,
      "loss": 0.793,
      "num_input_tokens_seen": 12934224,
      "step": 22295
    },
    {
      "epoch": 3.3214179326779862,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029602924555720986,
      "loss": 0.8032,
      "num_input_tokens_seen": 12937296,
      "step": 22300
    },
    {
      "epoch": 3.3221626452189454,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.02960247880705107,
      "loss": 0.8323,
      "num_input_tokens_seen": 12940112,
      "step": 22305
    },
    {
      "epoch": 3.3229073577599046,
      "grad_norm": 0.041015625,
      "learning_rate": 0.029602032811687117,
      "loss": 0.7974,
      "num_input_tokens_seen": 12943248,
      "step": 22310
    },
    {
      "epoch": 3.323652070300864,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029601586569636663,
      "loss": 0.7942,
      "num_input_tokens_seen": 12945968,
      "step": 22315
    },
    {
      "epoch": 3.324396782841823,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029601140080907244,
      "loss": 0.7784,
      "num_input_tokens_seen": 12949040,
      "step": 22320
    },
    {
      "epoch": 3.3251414953827823,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.0296006933455064,
      "loss": 0.789,
      "num_input_tokens_seen": 12951504,
      "step": 22325
    },
    {
      "epoch": 3.3258862079237415,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.029600246363441683,
      "loss": 0.8215,
      "num_input_tokens_seen": 12954608,
      "step": 22330
    },
    {
      "epoch": 3.3266309204647007,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.02959979913472064,
      "loss": 0.8048,
      "num_input_tokens_seen": 12957232,
      "step": 22335
    },
    {
      "epoch": 3.32737563300566,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029599351659350834,
      "loss": 0.7949,
      "num_input_tokens_seen": 12959856,
      "step": 22340
    },
    {
      "epoch": 3.328120345546619,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029598903937339816,
      "loss": 0.8366,
      "num_input_tokens_seen": 12962928,
      "step": 22345
    },
    {
      "epoch": 3.3288650580875783,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029598455968695158,
      "loss": 0.8045,
      "num_input_tokens_seen": 12965648,
      "step": 22350
    },
    {
      "epoch": 3.3296097706285375,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02959800775342442,
      "loss": 0.797,
      "num_input_tokens_seen": 12968528,
      "step": 22355
    },
    {
      "epoch": 3.3303544831694967,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029597559291535177,
      "loss": 0.8071,
      "num_input_tokens_seen": 12971184,
      "step": 22360
    },
    {
      "epoch": 3.331099195710456,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02959711058303501,
      "loss": 0.8141,
      "num_input_tokens_seen": 12974128,
      "step": 22365
    },
    {
      "epoch": 3.331843908251415,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02959666162793149,
      "loss": 0.7976,
      "num_input_tokens_seen": 12976976,
      "step": 22370
    },
    {
      "epoch": 3.3325886207923743,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.029596212426232213,
      "loss": 0.7949,
      "num_input_tokens_seen": 12980112,
      "step": 22375
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02959576297794476,
      "loss": 0.816,
      "num_input_tokens_seen": 12983184,
      "step": 22380
    },
    {
      "epoch": 3.3340780458742927,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029595313283076723,
      "loss": 0.8027,
      "num_input_tokens_seen": 12985968,
      "step": 22385
    },
    {
      "epoch": 3.334822758415252,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029594863341635705,
      "loss": 0.7961,
      "num_input_tokens_seen": 12989200,
      "step": 22390
    },
    {
      "epoch": 3.335567470956211,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.029594413153629304,
      "loss": 0.7953,
      "num_input_tokens_seen": 12991952,
      "step": 22395
    },
    {
      "epoch": 3.33631218349717,
      "grad_norm": 0.017578125,
      "learning_rate": 0.02959396271906513,
      "loss": 0.8035,
      "num_input_tokens_seen": 12994832,
      "step": 22400
    },
    {
      "epoch": 3.3370568960381295,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.029593512037950784,
      "loss": 0.7962,
      "num_input_tokens_seen": 12997616,
      "step": 22405
    },
    {
      "epoch": 3.3378016085790883,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029593061110293882,
      "loss": 0.807,
      "num_input_tokens_seen": 13000304,
      "step": 22410
    },
    {
      "epoch": 3.338546321120048,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.02959260993610205,
      "loss": 0.804,
      "num_input_tokens_seen": 13003056,
      "step": 22415
    },
    {
      "epoch": 3.3392910336610067,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029592158515382896,
      "loss": 0.8368,
      "num_input_tokens_seen": 13006160,
      "step": 22420
    },
    {
      "epoch": 3.340035746201966,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.029591706848144064,
      "loss": 0.7916,
      "num_input_tokens_seen": 13008848,
      "step": 22425
    },
    {
      "epoch": 3.340780458742925,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02959125493439317,
      "loss": 0.8118,
      "num_input_tokens_seen": 13011536,
      "step": 22430
    },
    {
      "epoch": 3.3415251712838843,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.029590802774137854,
      "loss": 0.7999,
      "num_input_tokens_seen": 13014384,
      "step": 22435
    },
    {
      "epoch": 3.3422698838248435,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029590350367385755,
      "loss": 0.791,
      "num_input_tokens_seen": 13017456,
      "step": 22440
    },
    {
      "epoch": 3.3430145963658027,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029589897714144518,
      "loss": 0.7944,
      "num_input_tokens_seen": 13020432,
      "step": 22445
    },
    {
      "epoch": 3.343759308906762,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.029589444814421786,
      "loss": 0.8062,
      "num_input_tokens_seen": 13023312,
      "step": 22450
    },
    {
      "epoch": 3.344504021447721,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.02958899166822521,
      "loss": 0.7836,
      "num_input_tokens_seen": 13026096,
      "step": 22455
    },
    {
      "epoch": 3.3452487339886803,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02958853827556245,
      "loss": 0.8101,
      "num_input_tokens_seen": 13028592,
      "step": 22460
    },
    {
      "epoch": 3.3459934465296395,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02958808463644116,
      "loss": 0.7927,
      "num_input_tokens_seen": 13031888,
      "step": 22465
    },
    {
      "epoch": 3.3467381590705987,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02958763075086901,
      "loss": 0.8235,
      "num_input_tokens_seen": 13034672,
      "step": 22470
    },
    {
      "epoch": 3.347482871611558,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029587176618853663,
      "loss": 0.8008,
      "num_input_tokens_seen": 13037520,
      "step": 22475
    },
    {
      "epoch": 3.348227584152517,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029586722240402794,
      "loss": 0.7863,
      "num_input_tokens_seen": 13040688,
      "step": 22480
    },
    {
      "epoch": 3.3489722966934763,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.029586267615524076,
      "loss": 0.8116,
      "num_input_tokens_seen": 13043504,
      "step": 22485
    },
    {
      "epoch": 3.3497170092344355,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02958581274422519,
      "loss": 0.7874,
      "num_input_tokens_seen": 13046512,
      "step": 22490
    },
    {
      "epoch": 3.3504617217753947,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029585357626513827,
      "loss": 0.7647,
      "num_input_tokens_seen": 13049392,
      "step": 22495
    },
    {
      "epoch": 3.351206434316354,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029584902262397666,
      "loss": 0.7753,
      "num_input_tokens_seen": 13052496,
      "step": 22500
    },
    {
      "epoch": 3.351951146857313,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029584446651884408,
      "loss": 0.8109,
      "num_input_tokens_seen": 13055312,
      "step": 22505
    },
    {
      "epoch": 3.3526958593982723,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029583990794981748,
      "loss": 0.7866,
      "num_input_tokens_seen": 13058288,
      "step": 22510
    },
    {
      "epoch": 3.3534405719392315,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02958353469169738,
      "loss": 0.7781,
      "num_input_tokens_seen": 13061168,
      "step": 22515
    },
    {
      "epoch": 3.3541852844801907,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02958307834203902,
      "loss": 0.7994,
      "num_input_tokens_seen": 13064112,
      "step": 22520
    },
    {
      "epoch": 3.35492999702115,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029582621746014375,
      "loss": 0.7889,
      "num_input_tokens_seen": 13067088,
      "step": 22525
    },
    {
      "epoch": 3.355674709562109,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029582164903631154,
      "loss": 0.7681,
      "num_input_tokens_seen": 13070096,
      "step": 22530
    },
    {
      "epoch": 3.3564194221030683,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02958170781489708,
      "loss": 0.7951,
      "num_input_tokens_seen": 13073040,
      "step": 22535
    },
    {
      "epoch": 3.3571641346440275,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029581250479819868,
      "loss": 0.804,
      "num_input_tokens_seen": 13075792,
      "step": 22540
    },
    {
      "epoch": 3.3579088471849867,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02958079289840725,
      "loss": 0.8117,
      "num_input_tokens_seen": 13078736,
      "step": 22545
    },
    {
      "epoch": 3.358653559725946,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.029580335070666958,
      "loss": 0.7838,
      "num_input_tokens_seen": 13081488,
      "step": 22550
    },
    {
      "epoch": 3.359398272266905,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.029579876996606722,
      "loss": 0.7909,
      "num_input_tokens_seen": 13084208,
      "step": 22555
    },
    {
      "epoch": 3.3601429848078643,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02957941867623428,
      "loss": 0.8233,
      "num_input_tokens_seen": 13087152,
      "step": 22560
    },
    {
      "epoch": 3.3608876973488235,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.029578960109557382,
      "loss": 0.8348,
      "num_input_tokens_seen": 13089872,
      "step": 22565
    },
    {
      "epoch": 3.3616324098897827,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029578501296583768,
      "loss": 0.8056,
      "num_input_tokens_seen": 13092944,
      "step": 22570
    },
    {
      "epoch": 3.3623771224307415,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029578042237321193,
      "loss": 0.8062,
      "num_input_tokens_seen": 13095760,
      "step": 22575
    },
    {
      "epoch": 3.363121834971701,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.029577582931777407,
      "loss": 0.756,
      "num_input_tokens_seen": 13098864,
      "step": 22580
    },
    {
      "epoch": 3.36386654751266,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.029577123379960178,
      "loss": 0.8215,
      "num_input_tokens_seen": 13102096,
      "step": 22585
    },
    {
      "epoch": 3.3646112600536195,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.02957666358187726,
      "loss": 0.8522,
      "num_input_tokens_seen": 13104752,
      "step": 22590
    },
    {
      "epoch": 3.3653559725945783,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02957620353753643,
      "loss": 0.7584,
      "num_input_tokens_seen": 13108080,
      "step": 22595
    },
    {
      "epoch": 3.3661006851355375,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029575743246945455,
      "loss": 0.795,
      "num_input_tokens_seen": 13111152,
      "step": 22600
    },
    {
      "epoch": 3.3668453976764967,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029575282710112113,
      "loss": 0.7891,
      "num_input_tokens_seen": 13114160,
      "step": 22605
    },
    {
      "epoch": 3.367590110217456,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029574821927044182,
      "loss": 0.7958,
      "num_input_tokens_seen": 13117520,
      "step": 22610
    },
    {
      "epoch": 3.368334822758415,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.02957436089774945,
      "loss": 0.8249,
      "num_input_tokens_seen": 13120272,
      "step": 22615
    },
    {
      "epoch": 3.3690795352993743,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029573899622235698,
      "loss": 0.7679,
      "num_input_tokens_seen": 13123056,
      "step": 22620
    },
    {
      "epoch": 3.3698242478403335,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.02957343810051073,
      "loss": 0.817,
      "num_input_tokens_seen": 13126064,
      "step": 22625
    },
    {
      "epoch": 3.3705689603812927,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029572976332582334,
      "loss": 0.8235,
      "num_input_tokens_seen": 13128720,
      "step": 22630
    },
    {
      "epoch": 3.371313672922252,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.029572514318458313,
      "loss": 0.8063,
      "num_input_tokens_seen": 13131600,
      "step": 22635
    },
    {
      "epoch": 3.372058385463211,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.02957205205814647,
      "loss": 0.8033,
      "num_input_tokens_seen": 13134576,
      "step": 22640
    },
    {
      "epoch": 3.3728030980041703,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029571589551654627,
      "loss": 0.8345,
      "num_input_tokens_seen": 13137488,
      "step": 22645
    },
    {
      "epoch": 3.3735478105451295,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.029571126798990585,
      "loss": 0.7883,
      "num_input_tokens_seen": 13140240,
      "step": 22650
    },
    {
      "epoch": 3.3742925230860887,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.029570663800162163,
      "loss": 0.8144,
      "num_input_tokens_seen": 13143344,
      "step": 22655
    },
    {
      "epoch": 3.375037235627048,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029570200555177187,
      "loss": 0.83,
      "num_input_tokens_seen": 13146640,
      "step": 22660
    },
    {
      "epoch": 3.375781948168007,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029569737064043478,
      "loss": 0.8,
      "num_input_tokens_seen": 13149616,
      "step": 22665
    },
    {
      "epoch": 3.3765266607089663,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029569273326768872,
      "loss": 0.8095,
      "num_input_tokens_seen": 13152496,
      "step": 22670
    },
    {
      "epoch": 3.3772713732499255,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.029568809343361202,
      "loss": 0.8202,
      "num_input_tokens_seen": 13155408,
      "step": 22675
    },
    {
      "epoch": 3.3780160857908847,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.0295683451138283,
      "loss": 0.8069,
      "num_input_tokens_seen": 13158128,
      "step": 22680
    },
    {
      "epoch": 3.378760798331844,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02956788063817802,
      "loss": 0.8077,
      "num_input_tokens_seen": 13161296,
      "step": 22685
    },
    {
      "epoch": 3.379505510872803,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.029567415916418198,
      "loss": 0.8171,
      "num_input_tokens_seen": 13164176,
      "step": 22690
    },
    {
      "epoch": 3.3802502234137624,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029566950948556697,
      "loss": 0.7923,
      "num_input_tokens_seen": 13166768,
      "step": 22695
    },
    {
      "epoch": 3.3809949359547216,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02956648573460136,
      "loss": 0.7982,
      "num_input_tokens_seen": 13169648,
      "step": 22700
    },
    {
      "epoch": 3.3817396484956808,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02956602027456005,
      "loss": 0.7888,
      "num_input_tokens_seen": 13172240,
      "step": 22705
    },
    {
      "epoch": 3.38248436103664,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029565554568440632,
      "loss": 0.81,
      "num_input_tokens_seen": 13175824,
      "step": 22710
    },
    {
      "epoch": 3.383229073577599,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.029565088616250974,
      "loss": 0.8258,
      "num_input_tokens_seen": 13178608,
      "step": 22715
    },
    {
      "epoch": 3.3839737861185584,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029564622417998947,
      "loss": 0.8201,
      "num_input_tokens_seen": 13181424,
      "step": 22720
    },
    {
      "epoch": 3.3847184986595176,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029564155973692427,
      "loss": 0.8369,
      "num_input_tokens_seen": 13184048,
      "step": 22725
    },
    {
      "epoch": 3.3854632112004768,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.029563689283339297,
      "loss": 0.809,
      "num_input_tokens_seen": 13187120,
      "step": 22730
    },
    {
      "epoch": 3.386207923741436,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.029563222346947433,
      "loss": 0.8129,
      "num_input_tokens_seen": 13190256,
      "step": 22735
    },
    {
      "epoch": 3.386952636282395,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029562755164524734,
      "loss": 0.8122,
      "num_input_tokens_seen": 13193200,
      "step": 22740
    },
    {
      "epoch": 3.3876973488233544,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.029562287736079083,
      "loss": 0.8064,
      "num_input_tokens_seen": 13196016,
      "step": 22745
    },
    {
      "epoch": 3.388442061364313,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.029561820061618387,
      "loss": 0.7903,
      "num_input_tokens_seen": 13198960,
      "step": 22750
    },
    {
      "epoch": 3.389186773905273,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.029561352141150535,
      "loss": 0.8089,
      "num_input_tokens_seen": 13201872,
      "step": 22755
    },
    {
      "epoch": 3.3899314864462315,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02956088397468344,
      "loss": 0.8083,
      "num_input_tokens_seen": 13205072,
      "step": 22760
    },
    {
      "epoch": 3.390676198987191,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.029560415562225013,
      "loss": 0.8108,
      "num_input_tokens_seen": 13208048,
      "step": 22765
    },
    {
      "epoch": 3.39142091152815,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029559946903783156,
      "loss": 0.7941,
      "num_input_tokens_seen": 13211216,
      "step": 22770
    },
    {
      "epoch": 3.392165624069109,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029559477999365803,
      "loss": 0.7838,
      "num_input_tokens_seen": 13213936,
      "step": 22775
    },
    {
      "epoch": 3.3929103366100684,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.029559008848980863,
      "loss": 0.8247,
      "num_input_tokens_seen": 13216752,
      "step": 22780
    },
    {
      "epoch": 3.3936550491510276,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029558539452636264,
      "loss": 0.8217,
      "num_input_tokens_seen": 13219472,
      "step": 22785
    },
    {
      "epoch": 3.3943997616919868,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029558069810339937,
      "loss": 0.7955,
      "num_input_tokens_seen": 13222416,
      "step": 22790
    },
    {
      "epoch": 3.395144474232946,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029557599922099817,
      "loss": 0.7981,
      "num_input_tokens_seen": 13225072,
      "step": 22795
    },
    {
      "epoch": 3.395889186773905,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.029557129787923846,
      "loss": 0.8091,
      "num_input_tokens_seen": 13227984,
      "step": 22800
    },
    {
      "epoch": 3.3966338993148644,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.029556659407819957,
      "loss": 0.8,
      "num_input_tokens_seen": 13230992,
      "step": 22805
    },
    {
      "epoch": 3.3973786118558236,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029556188781796108,
      "loss": 0.8199,
      "num_input_tokens_seen": 13233968,
      "step": 22810
    },
    {
      "epoch": 3.3981233243967828,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02955571790986024,
      "loss": 0.8128,
      "num_input_tokens_seen": 13237200,
      "step": 22815
    },
    {
      "epoch": 3.398868036937742,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.02955524679202031,
      "loss": 0.7985,
      "num_input_tokens_seen": 13239984,
      "step": 22820
    },
    {
      "epoch": 3.399612749478701,
      "grad_norm": 0.02734375,
      "learning_rate": 0.029554775428284285,
      "loss": 0.8113,
      "num_input_tokens_seen": 13243024,
      "step": 22825
    },
    {
      "epoch": 3.4003574620196604,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029554303818660116,
      "loss": 0.7999,
      "num_input_tokens_seen": 13246000,
      "step": 22830
    },
    {
      "epoch": 3.4011021745606196,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.029553831963155777,
      "loss": 0.7952,
      "num_input_tokens_seen": 13248784,
      "step": 22835
    },
    {
      "epoch": 3.401846887101579,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.02955335986177924,
      "loss": 0.7947,
      "num_input_tokens_seen": 13251600,
      "step": 22840
    },
    {
      "epoch": 3.402591599642538,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02955288751453848,
      "loss": 0.8185,
      "num_input_tokens_seen": 13254512,
      "step": 22845
    },
    {
      "epoch": 3.403336312183497,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.029552414921441478,
      "loss": 0.7827,
      "num_input_tokens_seen": 13257648,
      "step": 22850
    },
    {
      "epoch": 3.4040810247244564,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029551942082496212,
      "loss": 0.8041,
      "num_input_tokens_seen": 13260624,
      "step": 22855
    },
    {
      "epoch": 3.4048257372654156,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029551468997710675,
      "loss": 0.7856,
      "num_input_tokens_seen": 13263760,
      "step": 22860
    },
    {
      "epoch": 3.405570449806375,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029550995667092864,
      "loss": 0.7945,
      "num_input_tokens_seen": 13266384,
      "step": 22865
    },
    {
      "epoch": 3.406315162347334,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.029550522090650767,
      "loss": 0.8081,
      "num_input_tokens_seen": 13269040,
      "step": 22870
    },
    {
      "epoch": 3.407059874888293,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.029550048268392384,
      "loss": 0.8072,
      "num_input_tokens_seen": 13272208,
      "step": 22875
    },
    {
      "epoch": 3.4078045874292524,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02954957420032573,
      "loss": 0.8031,
      "num_input_tokens_seen": 13275088,
      "step": 22880
    },
    {
      "epoch": 3.4085492999702116,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029549099886458804,
      "loss": 0.8091,
      "num_input_tokens_seen": 13278448,
      "step": 22885
    },
    {
      "epoch": 3.409294012511171,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029548625326799624,
      "loss": 0.7737,
      "num_input_tokens_seen": 13281136,
      "step": 22890
    },
    {
      "epoch": 3.41003872505213,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.0295481505213562,
      "loss": 0.8017,
      "num_input_tokens_seen": 13284176,
      "step": 22895
    },
    {
      "epoch": 3.410783437593089,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029547675470136568,
      "loss": 0.7983,
      "num_input_tokens_seen": 13287024,
      "step": 22900
    },
    {
      "epoch": 3.4115281501340484,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.029547200173148738,
      "loss": 0.77,
      "num_input_tokens_seen": 13290096,
      "step": 22905
    },
    {
      "epoch": 3.4122728626750076,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029546724630400745,
      "loss": 0.7924,
      "num_input_tokens_seen": 13292880,
      "step": 22910
    },
    {
      "epoch": 3.413017575215967,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029546248841900626,
      "loss": 0.7985,
      "num_input_tokens_seen": 13295696,
      "step": 22915
    },
    {
      "epoch": 3.413762287756926,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029545772807656422,
      "loss": 0.8143,
      "num_input_tokens_seen": 13298448,
      "step": 22920
    },
    {
      "epoch": 3.414507000297885,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029545296527676165,
      "loss": 0.8076,
      "num_input_tokens_seen": 13301040,
      "step": 22925
    },
    {
      "epoch": 3.4152517128388444,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02954482000196791,
      "loss": 0.7795,
      "num_input_tokens_seen": 13304112,
      "step": 22930
    },
    {
      "epoch": 3.415996425379803,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.0295443432305397,
      "loss": 0.8015,
      "num_input_tokens_seen": 13306960,
      "step": 22935
    },
    {
      "epoch": 3.4167411379207624,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.029543866213399595,
      "loss": 0.7976,
      "num_input_tokens_seen": 13309520,
      "step": 22940
    },
    {
      "epoch": 3.4174858504617216,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029543388950555652,
      "loss": 0.7818,
      "num_input_tokens_seen": 13312368,
      "step": 22945
    },
    {
      "epoch": 3.418230563002681,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02954291144201593,
      "loss": 0.7599,
      "num_input_tokens_seen": 13315280,
      "step": 22950
    },
    {
      "epoch": 3.41897527554364,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.029542433687788503,
      "loss": 0.8185,
      "num_input_tokens_seen": 13318000,
      "step": 22955
    },
    {
      "epoch": 3.419719988084599,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02954195568788144,
      "loss": 0.8147,
      "num_input_tokens_seen": 13320688,
      "step": 22960
    },
    {
      "epoch": 3.4204647006255584,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.029541477442302816,
      "loss": 0.7904,
      "num_input_tokens_seen": 13323600,
      "step": 22965
    },
    {
      "epoch": 3.4212094131665176,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02954099895106071,
      "loss": 0.8019,
      "num_input_tokens_seen": 13326416,
      "step": 22970
    },
    {
      "epoch": 3.421954125707477,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.0295405202141632,
      "loss": 0.7719,
      "num_input_tokens_seen": 13329712,
      "step": 22975
    },
    {
      "epoch": 3.422698838248436,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02954004123161839,
      "loss": 0.7813,
      "num_input_tokens_seen": 13332528,
      "step": 22980
    },
    {
      "epoch": 3.423443550789395,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029539562003434352,
      "loss": 0.7595,
      "num_input_tokens_seen": 13335760,
      "step": 22985
    },
    {
      "epoch": 3.4241882633303544,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029539082529619193,
      "loss": 0.8266,
      "num_input_tokens_seen": 13338992,
      "step": 22990
    },
    {
      "epoch": 3.4249329758713136,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.029538602810181017,
      "loss": 0.8005,
      "num_input_tokens_seen": 13341712,
      "step": 22995
    },
    {
      "epoch": 3.425677688412273,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02953812284512792,
      "loss": 0.8068,
      "num_input_tokens_seen": 13344528,
      "step": 23000
    },
    {
      "epoch": 3.426422400953232,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02953764263446801,
      "loss": 0.8225,
      "num_input_tokens_seen": 13347280,
      "step": 23005
    },
    {
      "epoch": 3.4271671134941912,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029537162178209404,
      "loss": 0.8082,
      "num_input_tokens_seen": 13350352,
      "step": 23010
    },
    {
      "epoch": 3.4279118260351504,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02953668147636022,
      "loss": 0.7978,
      "num_input_tokens_seen": 13353456,
      "step": 23015
    },
    {
      "epoch": 3.4286565385761096,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02953620052892858,
      "loss": 0.7813,
      "num_input_tokens_seen": 13356368,
      "step": 23020
    },
    {
      "epoch": 3.429401251117069,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029535719335922597,
      "loss": 0.8335,
      "num_input_tokens_seen": 13359408,
      "step": 23025
    },
    {
      "epoch": 3.430145963658028,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.029535237897350415,
      "loss": 0.7891,
      "num_input_tokens_seen": 13362064,
      "step": 23030
    },
    {
      "epoch": 3.4308906761989872,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.02953475621322016,
      "loss": 0.7775,
      "num_input_tokens_seen": 13365136,
      "step": 23035
    },
    {
      "epoch": 3.4316353887399464,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.029534274283539975,
      "loss": 0.7957,
      "num_input_tokens_seen": 13368112,
      "step": 23040
    },
    {
      "epoch": 3.4323801012809056,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029533792108317993,
      "loss": 0.8227,
      "num_input_tokens_seen": 13371152,
      "step": 23045
    },
    {
      "epoch": 3.433124813821865,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029533309687562372,
      "loss": 0.8059,
      "num_input_tokens_seen": 13373840,
      "step": 23050
    },
    {
      "epoch": 3.433869526362824,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.02953282702128125,
      "loss": 0.8331,
      "num_input_tokens_seen": 13376944,
      "step": 23055
    },
    {
      "epoch": 3.4346142389037833,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029532344109482783,
      "loss": 0.8069,
      "num_input_tokens_seen": 13379600,
      "step": 23060
    },
    {
      "epoch": 3.4353589514447425,
      "grad_norm": 0.0111083984375,
      "learning_rate": 0.029531860952175137,
      "loss": 0.7877,
      "num_input_tokens_seen": 13382256,
      "step": 23065
    },
    {
      "epoch": 3.4361036639857017,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029531377549366467,
      "loss": 0.8228,
      "num_input_tokens_seen": 13385168,
      "step": 23070
    },
    {
      "epoch": 3.436848376526661,
      "grad_norm": 0.0108642578125,
      "learning_rate": 0.029530893901064944,
      "loss": 0.8113,
      "num_input_tokens_seen": 13387984,
      "step": 23075
    },
    {
      "epoch": 3.43759308906762,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029530410007278733,
      "loss": 0.8138,
      "num_input_tokens_seen": 13390832,
      "step": 23080
    },
    {
      "epoch": 3.4383378016085793,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.02952992586801602,
      "loss": 0.8221,
      "num_input_tokens_seen": 13393872,
      "step": 23085
    },
    {
      "epoch": 3.4390825141495385,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02952944148328497,
      "loss": 0.7972,
      "num_input_tokens_seen": 13397008,
      "step": 23090
    },
    {
      "epoch": 3.4398272266904977,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029528956853093775,
      "loss": 0.8216,
      "num_input_tokens_seen": 13399824,
      "step": 23095
    },
    {
      "epoch": 3.4405719392314564,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.029528471977450628,
      "loss": 0.8148,
      "num_input_tokens_seen": 13402896,
      "step": 23100
    },
    {
      "epoch": 3.441316651772416,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.029527986856363705,
      "loss": 0.8011,
      "num_input_tokens_seen": 13406096,
      "step": 23105
    },
    {
      "epoch": 3.442061364313375,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02952750148984121,
      "loss": 0.8044,
      "num_input_tokens_seen": 13409200,
      "step": 23110
    },
    {
      "epoch": 3.442806076854334,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.029527015877891347,
      "loss": 0.7914,
      "num_input_tokens_seen": 13412304,
      "step": 23115
    },
    {
      "epoch": 3.4435507893952932,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02952653002052231,
      "loss": 0.8243,
      "num_input_tokens_seen": 13414960,
      "step": 23120
    },
    {
      "epoch": 3.4442955019362524,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.029526043917742315,
      "loss": 0.7843,
      "num_input_tokens_seen": 13417680,
      "step": 23125
    },
    {
      "epoch": 3.4450402144772116,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02952555756955957,
      "loss": 0.7921,
      "num_input_tokens_seen": 13420240,
      "step": 23130
    },
    {
      "epoch": 3.445784927018171,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029525070975982295,
      "loss": 0.8118,
      "num_input_tokens_seen": 13423216,
      "step": 23135
    },
    {
      "epoch": 3.44652963955913,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02952458413701871,
      "loss": 0.8218,
      "num_input_tokens_seen": 13426192,
      "step": 23140
    },
    {
      "epoch": 3.4472743521000893,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029524097052677032,
      "loss": 0.806,
      "num_input_tokens_seen": 13429232,
      "step": 23145
    },
    {
      "epoch": 3.4480190646410485,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029523609722965503,
      "loss": 0.809,
      "num_input_tokens_seen": 13432048,
      "step": 23150
    },
    {
      "epoch": 3.4487637771820077,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029523122147892347,
      "loss": 0.7781,
      "num_input_tokens_seen": 13434960,
      "step": 23155
    },
    {
      "epoch": 3.449508489722967,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.0295226343274658,
      "loss": 0.8071,
      "num_input_tokens_seen": 13437712,
      "step": 23160
    },
    {
      "epoch": 3.450253202263926,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.029522146261694106,
      "loss": 0.8179,
      "num_input_tokens_seen": 13440496,
      "step": 23165
    },
    {
      "epoch": 3.4509979148048853,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.029521657950585516,
      "loss": 0.8272,
      "num_input_tokens_seen": 13443632,
      "step": 23170
    },
    {
      "epoch": 3.4517426273458445,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.02952116939414827,
      "loss": 0.8068,
      "num_input_tokens_seen": 13446608,
      "step": 23175
    },
    {
      "epoch": 3.4524873398868037,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.029520680592390627,
      "loss": 0.8175,
      "num_input_tokens_seen": 13449520,
      "step": 23180
    },
    {
      "epoch": 3.453232052427763,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029520191545320842,
      "loss": 0.803,
      "num_input_tokens_seen": 13452240,
      "step": 23185
    },
    {
      "epoch": 3.453976764968722,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02951970225294718,
      "loss": 0.8026,
      "num_input_tokens_seen": 13455248,
      "step": 23190
    },
    {
      "epoch": 3.4547214775096813,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029519212715277902,
      "loss": 0.8175,
      "num_input_tokens_seen": 13458320,
      "step": 23195
    },
    {
      "epoch": 3.4554661900506405,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.029518722932321287,
      "loss": 0.8023,
      "num_input_tokens_seen": 13460912,
      "step": 23200
    },
    {
      "epoch": 3.4562109025915997,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029518232904085603,
      "loss": 0.7939,
      "num_input_tokens_seen": 13463568,
      "step": 23205
    },
    {
      "epoch": 3.456955615132559,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.02951774263057913,
      "loss": 0.8063,
      "num_input_tokens_seen": 13466448,
      "step": 23210
    },
    {
      "epoch": 3.457700327673518,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02951725211181015,
      "loss": 0.7946,
      "num_input_tokens_seen": 13469392,
      "step": 23215
    },
    {
      "epoch": 3.4584450402144773,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.029516761347786948,
      "loss": 0.8062,
      "num_input_tokens_seen": 13472464,
      "step": 23220
    },
    {
      "epoch": 3.4591897527554365,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.02951627033851782,
      "loss": 0.8118,
      "num_input_tokens_seen": 13475408,
      "step": 23225
    },
    {
      "epoch": 3.4599344652963957,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029515779084011057,
      "loss": 0.8063,
      "num_input_tokens_seen": 13478288,
      "step": 23230
    },
    {
      "epoch": 3.460679177837355,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.02951528758427496,
      "loss": 0.8039,
      "num_input_tokens_seen": 13481232,
      "step": 23235
    },
    {
      "epoch": 3.461423890378314,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029514795839317834,
      "loss": 0.8143,
      "num_input_tokens_seen": 13484048,
      "step": 23240
    },
    {
      "epoch": 3.4621686029192733,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.02951430384914798,
      "loss": 0.8242,
      "num_input_tokens_seen": 13486832,
      "step": 23245
    },
    {
      "epoch": 3.4629133154602325,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029513811613773713,
      "loss": 0.7942,
      "num_input_tokens_seen": 13489456,
      "step": 23250
    },
    {
      "epoch": 3.4636580280011917,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.029513319133203358,
      "loss": 0.8106,
      "num_input_tokens_seen": 13492272,
      "step": 23255
    },
    {
      "epoch": 3.464402740542151,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02951282640744522,
      "loss": 0.7922,
      "num_input_tokens_seen": 13495088,
      "step": 23260
    },
    {
      "epoch": 3.4651474530831097,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02951233343650763,
      "loss": 0.7903,
      "num_input_tokens_seen": 13497904,
      "step": 23265
    },
    {
      "epoch": 3.4658921656240693,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02951184022039892,
      "loss": 0.7952,
      "num_input_tokens_seen": 13500880,
      "step": 23270
    },
    {
      "epoch": 3.466636878165028,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.029511346759127418,
      "loss": 0.824,
      "num_input_tokens_seen": 13503952,
      "step": 23275
    },
    {
      "epoch": 3.4673815907059877,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029510853052701458,
      "loss": 0.7916,
      "num_input_tokens_seen": 13507056,
      "step": 23280
    },
    {
      "epoch": 3.4681263032469465,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.029510359101129386,
      "loss": 0.8238,
      "num_input_tokens_seen": 13509776,
      "step": 23285
    },
    {
      "epoch": 3.4688710157879057,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029509864904419543,
      "loss": 0.8029,
      "num_input_tokens_seen": 13512848,
      "step": 23290
    },
    {
      "epoch": 3.469615728328865,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029509370462580287,
      "loss": 0.7914,
      "num_input_tokens_seen": 13515888,
      "step": 23295
    },
    {
      "epoch": 3.470360440869824,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029508875775619954,
      "loss": 0.8048,
      "num_input_tokens_seen": 13518352,
      "step": 23300
    },
    {
      "epoch": 3.4711051534107833,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029508380843546914,
      "loss": 0.7968,
      "num_input_tokens_seen": 13521200,
      "step": 23305
    },
    {
      "epoch": 3.4718498659517425,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02950788566636953,
      "loss": 0.8093,
      "num_input_tokens_seen": 13523984,
      "step": 23310
    },
    {
      "epoch": 3.4725945784927017,
      "grad_norm": 0.015625,
      "learning_rate": 0.029507390244096158,
      "loss": 0.8007,
      "num_input_tokens_seen": 13526608,
      "step": 23315
    },
    {
      "epoch": 3.473339291033661,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.029506894576735172,
      "loss": 0.8007,
      "num_input_tokens_seen": 13529392,
      "step": 23320
    },
    {
      "epoch": 3.47408400357462,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.02950639866429495,
      "loss": 0.8191,
      "num_input_tokens_seen": 13532208,
      "step": 23325
    },
    {
      "epoch": 3.4748287161155793,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.02950590250678386,
      "loss": 0.7958,
      "num_input_tokens_seen": 13535024,
      "step": 23330
    },
    {
      "epoch": 3.4755734286565385,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029505406104210298,
      "loss": 0.8139,
      "num_input_tokens_seen": 13538160,
      "step": 23335
    },
    {
      "epoch": 3.4763181411974977,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.029504909456582636,
      "loss": 0.806,
      "num_input_tokens_seen": 13541360,
      "step": 23340
    },
    {
      "epoch": 3.477062853738457,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.029504412563909274,
      "loss": 0.8108,
      "num_input_tokens_seen": 13544208,
      "step": 23345
    },
    {
      "epoch": 3.477807566279416,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029503915426198606,
      "loss": 0.7945,
      "num_input_tokens_seen": 13547216,
      "step": 23350
    },
    {
      "epoch": 3.4785522788203753,
      "grad_norm": 0.02734375,
      "learning_rate": 0.029503418043459025,
      "loss": 0.8204,
      "num_input_tokens_seen": 13549840,
      "step": 23355
    },
    {
      "epoch": 3.4792969913613345,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.029502920415698936,
      "loss": 0.8234,
      "num_input_tokens_seen": 13552592,
      "step": 23360
    },
    {
      "epoch": 3.4800417039022937,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.02950242254292675,
      "loss": 0.8144,
      "num_input_tokens_seen": 13555408,
      "step": 23365
    },
    {
      "epoch": 3.480786416443253,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029501924425150868,
      "loss": 0.809,
      "num_input_tokens_seen": 13558384,
      "step": 23370
    },
    {
      "epoch": 3.481531128984212,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029501426062379717,
      "loss": 0.8031,
      "num_input_tokens_seen": 13561488,
      "step": 23375
    },
    {
      "epoch": 3.4822758415251713,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029500927454621708,
      "loss": 0.8048,
      "num_input_tokens_seen": 13564432,
      "step": 23380
    },
    {
      "epoch": 3.4830205540661305,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.029500428601885273,
      "loss": 0.8047,
      "num_input_tokens_seen": 13567472,
      "step": 23385
    },
    {
      "epoch": 3.4837652666070897,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.029499929504178832,
      "loss": 0.7967,
      "num_input_tokens_seen": 13570448,
      "step": 23390
    },
    {
      "epoch": 3.484509979148049,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02949943016151082,
      "loss": 0.8016,
      "num_input_tokens_seen": 13573584,
      "step": 23395
    },
    {
      "epoch": 3.485254691689008,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.029498930573889668,
      "loss": 0.798,
      "num_input_tokens_seen": 13576624,
      "step": 23400
    },
    {
      "epoch": 3.4859994042299673,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.029498430741323824,
      "loss": 0.8136,
      "num_input_tokens_seen": 13579664,
      "step": 23405
    },
    {
      "epoch": 3.4867441167709265,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029497930663821725,
      "loss": 0.8126,
      "num_input_tokens_seen": 13582736,
      "step": 23410
    },
    {
      "epoch": 3.4874888293118858,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029497430341391826,
      "loss": 0.7942,
      "num_input_tokens_seen": 13585328,
      "step": 23415
    },
    {
      "epoch": 3.488233541852845,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02949692977404257,
      "loss": 0.8122,
      "num_input_tokens_seen": 13588112,
      "step": 23420
    },
    {
      "epoch": 3.488978254393804,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02949642896178243,
      "loss": 0.7991,
      "num_input_tokens_seen": 13590960,
      "step": 23425
    },
    {
      "epoch": 3.4897229669347634,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02949592790461985,
      "loss": 0.798,
      "num_input_tokens_seen": 13593776,
      "step": 23430
    },
    {
      "epoch": 3.4904676794757226,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.029495426602563296,
      "loss": 0.8036,
      "num_input_tokens_seen": 13596688,
      "step": 23435
    },
    {
      "epoch": 3.4912123920166813,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029494925055621252,
      "loss": 0.8137,
      "num_input_tokens_seen": 13599792,
      "step": 23440
    },
    {
      "epoch": 3.491957104557641,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029494423263802177,
      "loss": 0.8135,
      "num_input_tokens_seen": 13602864,
      "step": 23445
    },
    {
      "epoch": 3.4927018170985997,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.029493921227114545,
      "loss": 0.8013,
      "num_input_tokens_seen": 13605584,
      "step": 23450
    },
    {
      "epoch": 3.4934465296395594,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.029493418945566854,
      "loss": 0.7996,
      "num_input_tokens_seen": 13608464,
      "step": 23455
    },
    {
      "epoch": 3.494191242180518,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.02949291641916758,
      "loss": 0.7974,
      "num_input_tokens_seen": 13611472,
      "step": 23460
    },
    {
      "epoch": 3.4949359547214773,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02949241364792521,
      "loss": 0.7939,
      "num_input_tokens_seen": 13614608,
      "step": 23465
    },
    {
      "epoch": 3.4956806672624365,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.029491910631848244,
      "loss": 0.7909,
      "num_input_tokens_seen": 13617520,
      "step": 23470
    },
    {
      "epoch": 3.4964253798033957,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02949140737094517,
      "loss": 0.7937,
      "num_input_tokens_seen": 13620784,
      "step": 23475
    },
    {
      "epoch": 3.497170092344355,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029490903865224502,
      "loss": 0.8057,
      "num_input_tokens_seen": 13623696,
      "step": 23480
    },
    {
      "epoch": 3.497914804885314,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02949040011469474,
      "loss": 0.7919,
      "num_input_tokens_seen": 13626640,
      "step": 23485
    },
    {
      "epoch": 3.4986595174262733,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029489896119364395,
      "loss": 0.7987,
      "num_input_tokens_seen": 13629584,
      "step": 23490
    },
    {
      "epoch": 3.4994042299672325,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029489391879241985,
      "loss": 0.8026,
      "num_input_tokens_seen": 13632560,
      "step": 23495
    },
    {
      "epoch": 3.5001489425081918,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.02948888739433602,
      "loss": 0.8093,
      "num_input_tokens_seen": 13635376,
      "step": 23500
    },
    {
      "epoch": 3.500893655049151,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.029488382664655036,
      "loss": 0.7888,
      "num_input_tokens_seen": 13638128,
      "step": 23505
    },
    {
      "epoch": 3.50163836759011,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.029487877690207547,
      "loss": 0.814,
      "num_input_tokens_seen": 13640848,
      "step": 23510
    },
    {
      "epoch": 3.5023830801310694,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02948737247100209,
      "loss": 0.7856,
      "num_input_tokens_seen": 13643888,
      "step": 23515
    },
    {
      "epoch": 3.5031277926720286,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.0294868670070472,
      "loss": 0.8023,
      "num_input_tokens_seen": 13646864,
      "step": 23520
    },
    {
      "epoch": 3.5038725052129878,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.029486361298351416,
      "loss": 0.8072,
      "num_input_tokens_seen": 13649872,
      "step": 23525
    },
    {
      "epoch": 3.504617217753947,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.02948585534492328,
      "loss": 0.8098,
      "num_input_tokens_seen": 13652560,
      "step": 23530
    },
    {
      "epoch": 3.505361930294906,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02948534914677134,
      "loss": 0.7986,
      "num_input_tokens_seen": 13656048,
      "step": 23535
    },
    {
      "epoch": 3.5061066428358654,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.02948484270390415,
      "loss": 0.786,
      "num_input_tokens_seen": 13658800,
      "step": 23540
    },
    {
      "epoch": 3.5068513553768246,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029484336016330263,
      "loss": 0.7893,
      "num_input_tokens_seen": 13661584,
      "step": 23545
    },
    {
      "epoch": 3.5075960679177838,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.029483829084058243,
      "loss": 0.8172,
      "num_input_tokens_seen": 13664464,
      "step": 23550
    },
    {
      "epoch": 3.508340780458743,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.029483321907096648,
      "loss": 0.8267,
      "num_input_tokens_seen": 13667312,
      "step": 23555
    },
    {
      "epoch": 3.509085492999702,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02948281448545405,
      "loss": 0.8078,
      "num_input_tokens_seen": 13670224,
      "step": 23560
    },
    {
      "epoch": 3.5098302055406614,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.029482306819139024,
      "loss": 0.7798,
      "num_input_tokens_seen": 13673008,
      "step": 23565
    },
    {
      "epoch": 3.5105749180816206,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.029481798908160144,
      "loss": 0.7956,
      "num_input_tokens_seen": 13676048,
      "step": 23570
    },
    {
      "epoch": 3.51131963062258,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.029481290752525986,
      "loss": 0.8062,
      "num_input_tokens_seen": 13678960,
      "step": 23575
    },
    {
      "epoch": 3.512064343163539,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029480782352245136,
      "loss": 0.8229,
      "num_input_tokens_seen": 13682160,
      "step": 23580
    },
    {
      "epoch": 3.512809055704498,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.029480273707326192,
      "loss": 0.7987,
      "num_input_tokens_seen": 13685264,
      "step": 23585
    },
    {
      "epoch": 3.5135537682454574,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02947976481777774,
      "loss": 0.7894,
      "num_input_tokens_seen": 13688176,
      "step": 23590
    },
    {
      "epoch": 3.5142984807864166,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029479255683608374,
      "loss": 0.787,
      "num_input_tokens_seen": 13691088,
      "step": 23595
    },
    {
      "epoch": 3.515043193327376,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029478746304826706,
      "loss": 0.8082,
      "num_input_tokens_seen": 13693936,
      "step": 23600
    },
    {
      "epoch": 3.5157879058683346,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02947823668144133,
      "loss": 0.7978,
      "num_input_tokens_seen": 13697008,
      "step": 23605
    },
    {
      "epoch": 3.516532618409294,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02947772681346086,
      "loss": 0.7897,
      "num_input_tokens_seen": 13699984,
      "step": 23610
    },
    {
      "epoch": 3.517277330950253,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.029477216700893913,
      "loss": 0.7993,
      "num_input_tokens_seen": 13702768,
      "step": 23615
    },
    {
      "epoch": 3.5180220434912126,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0294767063437491,
      "loss": 0.7944,
      "num_input_tokens_seen": 13705744,
      "step": 23620
    },
    {
      "epoch": 3.5187667560321714,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02947619574203505,
      "loss": 0.7679,
      "num_input_tokens_seen": 13708464,
      "step": 23625
    },
    {
      "epoch": 3.519511468573131,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02947568489576038,
      "loss": 0.7847,
      "num_input_tokens_seen": 13711536,
      "step": 23630
    },
    {
      "epoch": 3.5202561811140898,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.029475173804933735,
      "loss": 0.7996,
      "num_input_tokens_seen": 13714512,
      "step": 23635
    },
    {
      "epoch": 3.5210008936550494,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.029474662469563737,
      "loss": 0.826,
      "num_input_tokens_seen": 13717360,
      "step": 23640
    },
    {
      "epoch": 3.521745606196008,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.029474150889659025,
      "loss": 0.8373,
      "num_input_tokens_seen": 13720432,
      "step": 23645
    },
    {
      "epoch": 3.5224903187369674,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029473639065228246,
      "loss": 0.8327,
      "num_input_tokens_seen": 13723536,
      "step": 23650
    },
    {
      "epoch": 3.5232350312779266,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02947312699628005,
      "loss": 0.7899,
      "num_input_tokens_seen": 13726320,
      "step": 23655
    },
    {
      "epoch": 3.523979743818886,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029472614682823077,
      "loss": 0.8161,
      "num_input_tokens_seen": 13728944,
      "step": 23660
    },
    {
      "epoch": 3.524724456359845,
      "grad_norm": 0.015625,
      "learning_rate": 0.029472102124865993,
      "loss": 0.8015,
      "num_input_tokens_seen": 13731888,
      "step": 23665
    },
    {
      "epoch": 3.525469168900804,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.02947158932241745,
      "loss": 0.8035,
      "num_input_tokens_seen": 13735024,
      "step": 23670
    },
    {
      "epoch": 3.5262138814417634,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029471076275486112,
      "loss": 0.7996,
      "num_input_tokens_seen": 13737840,
      "step": 23675
    },
    {
      "epoch": 3.5269585939827226,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.029470562984080653,
      "loss": 0.777,
      "num_input_tokens_seen": 13740368,
      "step": 23680
    },
    {
      "epoch": 3.527703306523682,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.02947004944820973,
      "loss": 0.8198,
      "num_input_tokens_seen": 13743536,
      "step": 23685
    },
    {
      "epoch": 3.528448019064641,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.029469535667882036,
      "loss": 0.798,
      "num_input_tokens_seen": 13746448,
      "step": 23690
    },
    {
      "epoch": 3.5291927316056,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029469021643106243,
      "loss": 0.7996,
      "num_input_tokens_seen": 13749232,
      "step": 23695
    },
    {
      "epoch": 3.5299374441465594,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02946850737389103,
      "loss": 0.8135,
      "num_input_tokens_seen": 13752048,
      "step": 23700
    },
    {
      "epoch": 3.5306821566875186,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029467992860245096,
      "loss": 0.7957,
      "num_input_tokens_seen": 13754832,
      "step": 23705
    },
    {
      "epoch": 3.531426869228478,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029467478102177126,
      "loss": 0.8092,
      "num_input_tokens_seen": 13757744,
      "step": 23710
    },
    {
      "epoch": 3.532171581769437,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.02946696309969582,
      "loss": 0.7948,
      "num_input_tokens_seen": 13760976,
      "step": 23715
    },
    {
      "epoch": 3.532916294310396,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029466447852809868,
      "loss": 0.7879,
      "num_input_tokens_seen": 13763920,
      "step": 23720
    },
    {
      "epoch": 3.5336610068513554,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02946593236152799,
      "loss": 0.8024,
      "num_input_tokens_seen": 13766448,
      "step": 23725
    },
    {
      "epoch": 3.5344057193923146,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029465416625858883,
      "loss": 0.7988,
      "num_input_tokens_seen": 13769808,
      "step": 23730
    },
    {
      "epoch": 3.535150431933274,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.029464900645811264,
      "loss": 0.809,
      "num_input_tokens_seen": 13772752,
      "step": 23735
    },
    {
      "epoch": 3.535895144474233,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.029464384421393854,
      "loss": 0.7914,
      "num_input_tokens_seen": 13775728,
      "step": 23740
    },
    {
      "epoch": 3.5366398570151922,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.029463867952615366,
      "loss": 0.7891,
      "num_input_tokens_seen": 13778992,
      "step": 23745
    },
    {
      "epoch": 3.5373845695561514,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.029463351239484528,
      "loss": 0.8167,
      "num_input_tokens_seen": 13782128,
      "step": 23750
    },
    {
      "epoch": 3.5381292820971106,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02946283428201008,
      "loss": 0.8199,
      "num_input_tokens_seen": 13784976,
      "step": 23755
    },
    {
      "epoch": 3.53887399463807,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029462317080200737,
      "loss": 0.8208,
      "num_input_tokens_seen": 13787824,
      "step": 23760
    },
    {
      "epoch": 3.539618707179029,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02946179963406525,
      "loss": 0.8106,
      "num_input_tokens_seen": 13790704,
      "step": 23765
    },
    {
      "epoch": 3.5403634197199882,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029461281943612354,
      "loss": 0.7919,
      "num_input_tokens_seen": 13793744,
      "step": 23770
    },
    {
      "epoch": 3.5411081322609474,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029460764008850795,
      "loss": 0.7937,
      "num_input_tokens_seen": 13796688,
      "step": 23775
    },
    {
      "epoch": 3.541852844801906,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.02946024582978933,
      "loss": 0.8211,
      "num_input_tokens_seen": 13799664,
      "step": 23780
    },
    {
      "epoch": 3.542597557342866,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02945972740643671,
      "loss": 0.7926,
      "num_input_tokens_seen": 13802352,
      "step": 23785
    },
    {
      "epoch": 3.5433422698838246,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.029459208738801687,
      "loss": 0.7913,
      "num_input_tokens_seen": 13804976,
      "step": 23790
    },
    {
      "epoch": 3.5440869824247843,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029458689826893033,
      "loss": 0.7813,
      "num_input_tokens_seen": 13808048,
      "step": 23795
    },
    {
      "epoch": 3.544831694965743,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.029458170670719503,
      "loss": 0.7858,
      "num_input_tokens_seen": 13810928,
      "step": 23800
    },
    {
      "epoch": 3.5455764075067027,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02945765127028988,
      "loss": 0.8008,
      "num_input_tokens_seen": 13813808,
      "step": 23805
    },
    {
      "epoch": 3.5463211200476614,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029457131625612927,
      "loss": 0.7885,
      "num_input_tokens_seen": 13816720,
      "step": 23810
    },
    {
      "epoch": 3.5470658325886206,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029456611736697436,
      "loss": 0.8072,
      "num_input_tokens_seen": 13819504,
      "step": 23815
    },
    {
      "epoch": 3.54781054512958,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02945609160355218,
      "loss": 0.8057,
      "num_input_tokens_seen": 13822672,
      "step": 23820
    },
    {
      "epoch": 3.548555257670539,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.029455571226185945,
      "loss": 0.7748,
      "num_input_tokens_seen": 13825520,
      "step": 23825
    },
    {
      "epoch": 3.5492999702114982,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02945505060460753,
      "loss": 0.777,
      "num_input_tokens_seen": 13828432,
      "step": 23830
    },
    {
      "epoch": 3.5500446827524574,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.029454529738825723,
      "loss": 0.7931,
      "num_input_tokens_seen": 13831696,
      "step": 23835
    },
    {
      "epoch": 3.5507893952934166,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029454008628849334,
      "loss": 0.7935,
      "num_input_tokens_seen": 13834640,
      "step": 23840
    },
    {
      "epoch": 3.551534107834376,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029453487274687152,
      "loss": 0.7839,
      "num_input_tokens_seen": 13837808,
      "step": 23845
    },
    {
      "epoch": 3.552278820375335,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.029452965676348,
      "loss": 0.8098,
      "num_input_tokens_seen": 13840816,
      "step": 23850
    },
    {
      "epoch": 3.5530235329162942,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.029452443833840675,
      "loss": 0.8027,
      "num_input_tokens_seen": 13843632,
      "step": 23855
    },
    {
      "epoch": 3.5537682454572534,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.029451921747174002,
      "loss": 0.8315,
      "num_input_tokens_seen": 13846640,
      "step": 23860
    },
    {
      "epoch": 3.5545129579982127,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.029451399416356803,
      "loss": 0.767,
      "num_input_tokens_seen": 13849424,
      "step": 23865
    },
    {
      "epoch": 3.555257670539172,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.029450876841397897,
      "loss": 0.7894,
      "num_input_tokens_seen": 13852176,
      "step": 23870
    },
    {
      "epoch": 3.556002383080131,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02945035402230611,
      "loss": 0.8129,
      "num_input_tokens_seen": 13855216,
      "step": 23875
    },
    {
      "epoch": 3.5567470956210903,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.029449830959090283,
      "loss": 0.8197,
      "num_input_tokens_seen": 13858096,
      "step": 23880
    },
    {
      "epoch": 3.5574918081620495,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029449307651759245,
      "loss": 0.7664,
      "num_input_tokens_seen": 13861328,
      "step": 23885
    },
    {
      "epoch": 3.5582365207030087,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02944878410032184,
      "loss": 0.8244,
      "num_input_tokens_seen": 13864240,
      "step": 23890
    },
    {
      "epoch": 3.558981233243968,
      "grad_norm": 0.011474609375,
      "learning_rate": 0.029448260304786915,
      "loss": 0.7934,
      "num_input_tokens_seen": 13866992,
      "step": 23895
    },
    {
      "epoch": 3.559725945784927,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.029447736265163317,
      "loss": 0.7876,
      "num_input_tokens_seen": 13869712,
      "step": 23900
    },
    {
      "epoch": 3.5604706583258863,
      "grad_norm": 0.021484375,
      "learning_rate": 0.0294472119814599,
      "loss": 0.7935,
      "num_input_tokens_seen": 13872528,
      "step": 23905
    },
    {
      "epoch": 3.5612153708668455,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02944668745368552,
      "loss": 0.8275,
      "num_input_tokens_seen": 13875184,
      "step": 23910
    },
    {
      "epoch": 3.5619600834078047,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029446162681849034,
      "loss": 0.7768,
      "num_input_tokens_seen": 13878096,
      "step": 23915
    },
    {
      "epoch": 3.562704795948764,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02944563766595931,
      "loss": 0.8169,
      "num_input_tokens_seen": 13881008,
      "step": 23920
    },
    {
      "epoch": 3.563449508489723,
      "grad_norm": 0.03125,
      "learning_rate": 0.029445112406025223,
      "loss": 0.8075,
      "num_input_tokens_seen": 13883824,
      "step": 23925
    },
    {
      "epoch": 3.5641942210306823,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.029444586902055644,
      "loss": 0.8,
      "num_input_tokens_seen": 13886384,
      "step": 23930
    },
    {
      "epoch": 3.5649389335716415,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02944406115405945,
      "loss": 0.787,
      "num_input_tokens_seen": 13889360,
      "step": 23935
    },
    {
      "epoch": 3.5656836461126007,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.029443535162045526,
      "loss": 0.8098,
      "num_input_tokens_seen": 13892592,
      "step": 23940
    },
    {
      "epoch": 3.5664283586535594,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.029443008926022748,
      "loss": 0.8104,
      "num_input_tokens_seen": 13895632,
      "step": 23945
    },
    {
      "epoch": 3.567173071194519,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02944248244600002,
      "loss": 0.814,
      "num_input_tokens_seen": 13898800,
      "step": 23950
    },
    {
      "epoch": 3.567917783735478,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029441955721986227,
      "loss": 0.8105,
      "num_input_tokens_seen": 13901680,
      "step": 23955
    },
    {
      "epoch": 3.5686624962764375,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02944142875399027,
      "loss": 0.818,
      "num_input_tokens_seen": 13904464,
      "step": 23960
    },
    {
      "epoch": 3.5694072088173963,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.02944090154202105,
      "loss": 0.801,
      "num_input_tokens_seen": 13907280,
      "step": 23965
    },
    {
      "epoch": 3.570151921358356,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02944037408608748,
      "loss": 0.7817,
      "num_input_tokens_seen": 13910064,
      "step": 23970
    },
    {
      "epoch": 3.5708966338993147,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.029439846386198464,
      "loss": 0.8205,
      "num_input_tokens_seen": 13913072,
      "step": 23975
    },
    {
      "epoch": 3.5716413464402743,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02943931844236292,
      "loss": 0.8033,
      "num_input_tokens_seen": 13915824,
      "step": 23980
    },
    {
      "epoch": 3.572386058981233,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02943879025458977,
      "loss": 0.7958,
      "num_input_tokens_seen": 13919152,
      "step": 23985
    },
    {
      "epoch": 3.5731307715221923,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029438261822887925,
      "loss": 0.8069,
      "num_input_tokens_seen": 13921840,
      "step": 23990
    },
    {
      "epoch": 3.5738754840631515,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.029437733147266328,
      "loss": 0.8043,
      "num_input_tokens_seen": 13924496,
      "step": 23995
    },
    {
      "epoch": 3.5746201966041107,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.0294372042277339,
      "loss": 0.8132,
      "num_input_tokens_seen": 13927408,
      "step": 24000
    },
    {
      "epoch": 3.57536490914507,
      "grad_norm": 0.041015625,
      "learning_rate": 0.029436675064299582,
      "loss": 0.8039,
      "num_input_tokens_seen": 13930224,
      "step": 24005
    },
    {
      "epoch": 3.576109621686029,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029436145656972307,
      "loss": 0.7995,
      "num_input_tokens_seen": 13933072,
      "step": 24010
    },
    {
      "epoch": 3.5768543342269883,
      "grad_norm": 0.03125,
      "learning_rate": 0.02943561600576103,
      "loss": 0.7889,
      "num_input_tokens_seen": 13936208,
      "step": 24015
    },
    {
      "epoch": 3.5775990467679475,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029435086110674684,
      "loss": 0.8021,
      "num_input_tokens_seen": 13939216,
      "step": 24020
    },
    {
      "epoch": 3.5783437593089067,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.029434555971722235,
      "loss": 0.8154,
      "num_input_tokens_seen": 13942384,
      "step": 24025
    },
    {
      "epoch": 3.579088471849866,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029434025588912636,
      "loss": 0.7915,
      "num_input_tokens_seen": 13945136,
      "step": 24030
    },
    {
      "epoch": 3.579833184390825,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02943349496225484,
      "loss": 0.7851,
      "num_input_tokens_seen": 13947728,
      "step": 24035
    },
    {
      "epoch": 3.5805778969317843,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02943296409175782,
      "loss": 0.7973,
      "num_input_tokens_seen": 13950672,
      "step": 24040
    },
    {
      "epoch": 3.5813226094727435,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029432432977430537,
      "loss": 0.7948,
      "num_input_tokens_seen": 13953648,
      "step": 24045
    },
    {
      "epoch": 3.5820673220137027,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02943190161928197,
      "loss": 0.8014,
      "num_input_tokens_seen": 13956560,
      "step": 24050
    },
    {
      "epoch": 3.582812034554662,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02943137001732109,
      "loss": 0.8029,
      "num_input_tokens_seen": 13959504,
      "step": 24055
    },
    {
      "epoch": 3.583556747095621,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02943083817155688,
      "loss": 0.8136,
      "num_input_tokens_seen": 13962384,
      "step": 24060
    },
    {
      "epoch": 3.5843014596365803,
      "grad_norm": 0.029296875,
      "learning_rate": 0.029430306081998328,
      "loss": 0.7951,
      "num_input_tokens_seen": 13965200,
      "step": 24065
    },
    {
      "epoch": 3.5850461721775395,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029429773748654423,
      "loss": 0.7869,
      "num_input_tokens_seen": 13967952,
      "step": 24070
    },
    {
      "epoch": 3.5857908847184987,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029429241171534157,
      "loss": 0.8303,
      "num_input_tokens_seen": 13970928,
      "step": 24075
    },
    {
      "epoch": 3.586535597259458,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.029428708350646524,
      "loss": 0.792,
      "num_input_tokens_seen": 13973872,
      "step": 24080
    },
    {
      "epoch": 3.587280309800417,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.02942817528600053,
      "loss": 0.7695,
      "num_input_tokens_seen": 13976592,
      "step": 24085
    },
    {
      "epoch": 3.5880250223413763,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02942764197760518,
      "loss": 0.8002,
      "num_input_tokens_seen": 13979472,
      "step": 24090
    },
    {
      "epoch": 3.5887697348823355,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02942710842546948,
      "loss": 0.8137,
      "num_input_tokens_seen": 13982288,
      "step": 24095
    },
    {
      "epoch": 3.5895144474232947,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.029426574629602444,
      "loss": 0.7903,
      "num_input_tokens_seen": 13985136,
      "step": 24100
    },
    {
      "epoch": 3.590259159964254,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.029426040590013094,
      "loss": 0.7903,
      "num_input_tokens_seen": 13988016,
      "step": 24105
    },
    {
      "epoch": 3.591003872505213,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.029425506306710446,
      "loss": 2.8683,
      "num_input_tokens_seen": 13991120,
      "step": 24110
    },
    {
      "epoch": 3.5917485850461723,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.029424971779703536,
      "loss": 0.7865,
      "num_input_tokens_seen": 13993968,
      "step": 24115
    },
    {
      "epoch": 3.592493297587131,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029424437009001388,
      "loss": 0.8088,
      "num_input_tokens_seen": 13996752,
      "step": 24120
    },
    {
      "epoch": 3.5932380101280907,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.029423901994613037,
      "loss": 0.7727,
      "num_input_tokens_seen": 13999696,
      "step": 24125
    },
    {
      "epoch": 3.5939827226690495,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.02942336673654752,
      "loss": 0.7795,
      "num_input_tokens_seen": 14002832,
      "step": 24130
    },
    {
      "epoch": 3.594727435210009,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.029422831234813884,
      "loss": 0.8163,
      "num_input_tokens_seen": 14005712,
      "step": 24135
    },
    {
      "epoch": 3.595472147750968,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02942229548942117,
      "loss": 1.2877,
      "num_input_tokens_seen": 14009040,
      "step": 24140
    },
    {
      "epoch": 3.5962168602919276,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.029421759500378432,
      "loss": 0.7919,
      "num_input_tokens_seen": 14011600,
      "step": 24145
    },
    {
      "epoch": 3.5969615728328863,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.029421223267694727,
      "loss": 0.8241,
      "num_input_tokens_seen": 14014416,
      "step": 24150
    },
    {
      "epoch": 3.597706285373846,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.02942068679137911,
      "loss": 0.8132,
      "num_input_tokens_seen": 14017104,
      "step": 24155
    },
    {
      "epoch": 3.5984509979148047,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02942015007144065,
      "loss": 0.8336,
      "num_input_tokens_seen": 14019856,
      "step": 24160
    },
    {
      "epoch": 3.599195710455764,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.029419613107888407,
      "loss": 0.8183,
      "num_input_tokens_seen": 14022608,
      "step": 24165
    },
    {
      "epoch": 3.599940422996723,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.029419075900731458,
      "loss": 0.7901,
      "num_input_tokens_seen": 14025392,
      "step": 24170
    },
    {
      "epoch": 3.6006851355376823,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029418538449978875,
      "loss": 0.8087,
      "num_input_tokens_seen": 14028208,
      "step": 24175
    },
    {
      "epoch": 3.6014298480786415,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.029418000755639738,
      "loss": 0.8157,
      "num_input_tokens_seen": 14031024,
      "step": 24180
    },
    {
      "epoch": 3.6021745606196007,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02941746281772313,
      "loss": 0.7826,
      "num_input_tokens_seen": 14033936,
      "step": 24185
    },
    {
      "epoch": 3.60291927316056,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029416924636238146,
      "loss": 0.7982,
      "num_input_tokens_seen": 14036976,
      "step": 24190
    },
    {
      "epoch": 3.603663985701519,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02941638621119387,
      "loss": 0.7597,
      "num_input_tokens_seen": 14039920,
      "step": 24195
    },
    {
      "epoch": 3.6044086982424783,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029415847542599405,
      "loss": 0.821,
      "num_input_tokens_seen": 14042896,
      "step": 24200
    },
    {
      "epoch": 3.6051534107834375,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.029415308630463845,
      "loss": 0.7829,
      "num_input_tokens_seen": 14045616,
      "step": 24205
    },
    {
      "epoch": 3.6058981233243967,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.029414769474796296,
      "loss": 0.79,
      "num_input_tokens_seen": 14048592,
      "step": 24210
    },
    {
      "epoch": 3.606642835865356,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.029414230075605872,
      "loss": 0.8429,
      "num_input_tokens_seen": 14051568,
      "step": 24215
    },
    {
      "epoch": 3.607387548406315,
      "grad_norm": 0.04296875,
      "learning_rate": 0.029413690432901676,
      "loss": 0.8213,
      "num_input_tokens_seen": 14054448,
      "step": 24220
    },
    {
      "epoch": 3.6081322609472744,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.02941315054669283,
      "loss": 0.8068,
      "num_input_tokens_seen": 14057488,
      "step": 24225
    },
    {
      "epoch": 3.6088769734882336,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029412610416988453,
      "loss": 0.8098,
      "num_input_tokens_seen": 14060400,
      "step": 24230
    },
    {
      "epoch": 3.6096216860291928,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029412070043797672,
      "loss": 0.8084,
      "num_input_tokens_seen": 14063152,
      "step": 24235
    },
    {
      "epoch": 3.610366398570152,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.029411529427129615,
      "loss": 0.7852,
      "num_input_tokens_seen": 14066288,
      "step": 24240
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02941098856699342,
      "loss": 0.7801,
      "num_input_tokens_seen": 14069168,
      "step": 24245
    },
    {
      "epoch": 3.6118558236520704,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.029410447463398215,
      "loss": 0.806,
      "num_input_tokens_seen": 14072176,
      "step": 24250
    },
    {
      "epoch": 3.6126005361930296,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02940990611635315,
      "loss": 0.8097,
      "num_input_tokens_seen": 14075056,
      "step": 24255
    },
    {
      "epoch": 3.6133452487339888,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.02940936452586736,
      "loss": 0.7913,
      "num_input_tokens_seen": 14078224,
      "step": 24260
    },
    {
      "epoch": 3.614089961274948,
      "grad_norm": 0.037109375,
      "learning_rate": 0.02940882269195001,
      "loss": 0.8181,
      "num_input_tokens_seen": 14081360,
      "step": 24265
    },
    {
      "epoch": 3.614834673815907,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02940828061461024,
      "loss": 0.7879,
      "num_input_tokens_seen": 14084336,
      "step": 24270
    },
    {
      "epoch": 3.6155793863568664,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.029407738293857216,
      "loss": 0.8022,
      "num_input_tokens_seen": 14087312,
      "step": 24275
    },
    {
      "epoch": 3.6163240988978256,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.029407195729700093,
      "loss": 0.809,
      "num_input_tokens_seen": 14090128,
      "step": 24280
    },
    {
      "epoch": 3.617068811438785,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029406652922148044,
      "loss": 0.8297,
      "num_input_tokens_seen": 14092656,
      "step": 24285
    },
    {
      "epoch": 3.617813523979744,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029406109871210236,
      "loss": 0.7836,
      "num_input_tokens_seen": 14095664,
      "step": 24290
    },
    {
      "epoch": 3.6185582365207027,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029405566576895845,
      "loss": 0.8031,
      "num_input_tokens_seen": 14098800,
      "step": 24295
    },
    {
      "epoch": 3.6193029490616624,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02940502303921404,
      "loss": 0.8316,
      "num_input_tokens_seen": 14101584,
      "step": 24300
    },
    {
      "epoch": 3.620047661602621,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.029404479258174018,
      "loss": 0.7847,
      "num_input_tokens_seen": 14104528,
      "step": 24305
    },
    {
      "epoch": 3.620792374143581,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02940393523378496,
      "loss": 0.8128,
      "num_input_tokens_seen": 14107184,
      "step": 24310
    },
    {
      "epoch": 3.6215370866845396,
      "grad_norm": 0.033203125,
      "learning_rate": 0.029403390966056056,
      "loss": 0.8133,
      "num_input_tokens_seen": 14110128,
      "step": 24315
    },
    {
      "epoch": 3.622281799225499,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02940284645499649,
      "loss": 0.7957,
      "num_input_tokens_seen": 14112688,
      "step": 24320
    },
    {
      "epoch": 3.623026511766458,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.02940230170061548,
      "loss": 0.8281,
      "num_input_tokens_seen": 14115536,
      "step": 24325
    },
    {
      "epoch": 3.6237712243074176,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.029401756702922225,
      "loss": 0.8039,
      "num_input_tokens_seen": 14118256,
      "step": 24330
    },
    {
      "epoch": 3.6245159368483764,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02940121146192592,
      "loss": 0.8159,
      "num_input_tokens_seen": 14121200,
      "step": 24335
    },
    {
      "epoch": 3.6252606493893356,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02940066597763579,
      "loss": 0.8135,
      "num_input_tokens_seen": 14124144,
      "step": 24340
    },
    {
      "epoch": 3.6260053619302948,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02940012025006104,
      "loss": 0.8175,
      "num_input_tokens_seen": 14127152,
      "step": 24345
    },
    {
      "epoch": 3.626750074471254,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0293995742792109,
      "loss": 0.7894,
      "num_input_tokens_seen": 14129872,
      "step": 24350
    },
    {
      "epoch": 3.627494787012213,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02939902806509458,
      "loss": 0.795,
      "num_input_tokens_seen": 14132752,
      "step": 24355
    },
    {
      "epoch": 3.6282394995531724,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02939848160772132,
      "loss": 0.8116,
      "num_input_tokens_seen": 14135696,
      "step": 24360
    },
    {
      "epoch": 3.6289842120941316,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029397934907100343,
      "loss": 0.7977,
      "num_input_tokens_seen": 14138512,
      "step": 24365
    },
    {
      "epoch": 3.629728924635091,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.02939738796324089,
      "loss": 0.8125,
      "num_input_tokens_seen": 14141104,
      "step": 24370
    },
    {
      "epoch": 3.63047363717605,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029396840776152205,
      "loss": 0.8196,
      "num_input_tokens_seen": 14143952,
      "step": 24375
    },
    {
      "epoch": 3.631218349717009,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.029396293345843524,
      "loss": 0.8136,
      "num_input_tokens_seen": 14146576,
      "step": 24380
    },
    {
      "epoch": 3.6319630622579684,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.0293957456723241,
      "loss": 0.8027,
      "num_input_tokens_seen": 14149392,
      "step": 24385
    },
    {
      "epoch": 3.6327077747989276,
      "grad_norm": 0.04296875,
      "learning_rate": 0.029395197755603186,
      "loss": 0.7935,
      "num_input_tokens_seen": 14152496,
      "step": 24390
    },
    {
      "epoch": 3.633452487339887,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02939464959569003,
      "loss": 0.8309,
      "num_input_tokens_seen": 14155312,
      "step": 24395
    },
    {
      "epoch": 3.634197199880846,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.029394101192593905,
      "loss": 0.7977,
      "num_input_tokens_seen": 14158288,
      "step": 24400
    },
    {
      "epoch": 3.634941912421805,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02939355254632407,
      "loss": 0.8069,
      "num_input_tokens_seen": 14161296,
      "step": 24405
    },
    {
      "epoch": 3.6356866249627644,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02939300365688979,
      "loss": 0.7985,
      "num_input_tokens_seen": 14163952,
      "step": 24410
    },
    {
      "epoch": 3.6364313375037236,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029392454524300345,
      "loss": 0.8161,
      "num_input_tokens_seen": 14166768,
      "step": 24415
    },
    {
      "epoch": 3.637176050044683,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029391905148565007,
      "loss": 0.7965,
      "num_input_tokens_seen": 14169648,
      "step": 24420
    },
    {
      "epoch": 3.637920762585642,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.029391355529693054,
      "loss": 0.8321,
      "num_input_tokens_seen": 14172656,
      "step": 24425
    },
    {
      "epoch": 3.638665475126601,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02939080566769378,
      "loss": 0.8106,
      "num_input_tokens_seen": 14175632,
      "step": 24430
    },
    {
      "epoch": 3.6394101876675604,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029390255562576476,
      "loss": 0.8082,
      "num_input_tokens_seen": 14178416,
      "step": 24435
    },
    {
      "epoch": 3.6401549002085196,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02938970521435042,
      "loss": 0.8075,
      "num_input_tokens_seen": 14181424,
      "step": 24440
    },
    {
      "epoch": 3.640899612749479,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.029389154623024927,
      "loss": 0.8236,
      "num_input_tokens_seen": 14184272,
      "step": 24445
    },
    {
      "epoch": 3.641644325290438,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.029388603788609288,
      "loss": 0.7974,
      "num_input_tokens_seen": 14187216,
      "step": 24450
    },
    {
      "epoch": 3.6423890378313972,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029388052711112816,
      "loss": 0.8041,
      "num_input_tokens_seen": 14190256,
      "step": 24455
    },
    {
      "epoch": 3.6431337503723564,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029387501390544808,
      "loss": 0.8029,
      "num_input_tokens_seen": 14193360,
      "step": 24460
    },
    {
      "epoch": 3.6438784629133156,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.029386949826914595,
      "loss": 0.7892,
      "num_input_tokens_seen": 14196240,
      "step": 24465
    },
    {
      "epoch": 3.6446231754542744,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029386398020231484,
      "loss": 0.8063,
      "num_input_tokens_seen": 14199152,
      "step": 24470
    },
    {
      "epoch": 3.645367887995234,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.0293858459705048,
      "loss": 0.8214,
      "num_input_tokens_seen": 14202032,
      "step": 24475
    },
    {
      "epoch": 3.646112600536193,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02938529367774387,
      "loss": 0.8182,
      "num_input_tokens_seen": 14204816,
      "step": 24480
    },
    {
      "epoch": 3.6468573130771524,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02938474114195802,
      "loss": 0.8116,
      "num_input_tokens_seen": 14207792,
      "step": 24485
    },
    {
      "epoch": 3.647602025618111,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029384188363156594,
      "loss": 0.7932,
      "num_input_tokens_seen": 14210992,
      "step": 24490
    },
    {
      "epoch": 3.648346738159071,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.029383635341348923,
      "loss": 0.8059,
      "num_input_tokens_seen": 14213552,
      "step": 24495
    },
    {
      "epoch": 3.6490914507000296,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029383082076544352,
      "loss": 0.7995,
      "num_input_tokens_seen": 14216368,
      "step": 24500
    },
    {
      "epoch": 3.6498361632409893,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.029382528568752227,
      "loss": 0.814,
      "num_input_tokens_seen": 14219216,
      "step": 24505
    },
    {
      "epoch": 3.650580875781948,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.0293819748179819,
      "loss": 0.799,
      "num_input_tokens_seen": 14222416,
      "step": 24510
    },
    {
      "epoch": 3.651325588322907,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029381420824242728,
      "loss": 0.8011,
      "num_input_tokens_seen": 14225104,
      "step": 24515
    },
    {
      "epoch": 3.6520703008638664,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.02938086658754406,
      "loss": 0.8126,
      "num_input_tokens_seen": 14228144,
      "step": 24520
    },
    {
      "epoch": 3.6528150134048256,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029380312107895275,
      "loss": 0.8047,
      "num_input_tokens_seen": 14231056,
      "step": 24525
    },
    {
      "epoch": 3.653559725945785,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.029379757385305728,
      "loss": 0.7957,
      "num_input_tokens_seen": 14234000,
      "step": 24530
    },
    {
      "epoch": 3.654304438486744,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.029379202419784797,
      "loss": 0.8119,
      "num_input_tokens_seen": 14236912,
      "step": 24535
    },
    {
      "epoch": 3.6550491510277032,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.029378647211341855,
      "loss": 0.7997,
      "num_input_tokens_seen": 14239696,
      "step": 24540
    },
    {
      "epoch": 3.6557938635686624,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.029378091759986284,
      "loss": 0.8038,
      "num_input_tokens_seen": 14242576,
      "step": 24545
    },
    {
      "epoch": 3.6565385761096216,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02937753606572746,
      "loss": 0.7944,
      "num_input_tokens_seen": 14245648,
      "step": 24550
    },
    {
      "epoch": 3.657283288650581,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.029376980128574783,
      "loss": 0.7948,
      "num_input_tokens_seen": 14248464,
      "step": 24555
    },
    {
      "epoch": 3.65802800119154,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.029376423948537636,
      "loss": 0.8018,
      "num_input_tokens_seen": 14251568,
      "step": 24560
    },
    {
      "epoch": 3.6587727137324992,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02937586752562542,
      "loss": 0.7938,
      "num_input_tokens_seen": 14254544,
      "step": 24565
    },
    {
      "epoch": 3.6595174262734584,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02937531085984753,
      "loss": 0.796,
      "num_input_tokens_seen": 14257296,
      "step": 24570
    },
    {
      "epoch": 3.6602621388144176,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.02937475395121338,
      "loss": 0.7928,
      "num_input_tokens_seen": 14260368,
      "step": 24575
    },
    {
      "epoch": 3.661006851355377,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.029374196799732365,
      "loss": 0.8218,
      "num_input_tokens_seen": 14263216,
      "step": 24580
    },
    {
      "epoch": 3.661751563896336,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.029373639405413907,
      "loss": 0.7896,
      "num_input_tokens_seen": 14266192,
      "step": 24585
    },
    {
      "epoch": 3.6624962764372953,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.029373081768267417,
      "loss": 0.8124,
      "num_input_tokens_seen": 14268976,
      "step": 24590
    },
    {
      "epoch": 3.6632409889782545,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.029372523888302325,
      "loss": 0.7919,
      "num_input_tokens_seen": 14271664,
      "step": 24595
    },
    {
      "epoch": 3.6639857015192137,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029371965765528044,
      "loss": 0.7994,
      "num_input_tokens_seen": 14274416,
      "step": 24600
    },
    {
      "epoch": 3.664730414060173,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.029371407399954013,
      "loss": 0.7906,
      "num_input_tokens_seen": 14277456,
      "step": 24605
    },
    {
      "epoch": 3.665475126601132,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.029370848791589656,
      "loss": 0.8118,
      "num_input_tokens_seen": 14280624,
      "step": 24610
    },
    {
      "epoch": 3.6662198391420913,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.029370289940444414,
      "loss": 0.8419,
      "num_input_tokens_seen": 14283536,
      "step": 24615
    },
    {
      "epoch": 3.6669645516830505,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.029369730846527733,
      "loss": 0.795,
      "num_input_tokens_seen": 14286640,
      "step": 24620
    },
    {
      "epoch": 3.6677092642240097,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02936917150984905,
      "loss": 0.8172,
      "num_input_tokens_seen": 14289584,
      "step": 24625
    },
    {
      "epoch": 3.668453976764969,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.029368611930417822,
      "loss": 0.8006,
      "num_input_tokens_seen": 14292528,
      "step": 24630
    },
    {
      "epoch": 3.669198689305928,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.0293680521082435,
      "loss": 0.8143,
      "num_input_tokens_seen": 14295440,
      "step": 24635
    },
    {
      "epoch": 3.6699434018468873,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02936749204333554,
      "loss": 0.807,
      "num_input_tokens_seen": 14298672,
      "step": 24640
    },
    {
      "epoch": 3.670688114387846,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.0293669317357034,
      "loss": 0.8048,
      "num_input_tokens_seen": 14301296,
      "step": 24645
    },
    {
      "epoch": 3.6714328269288057,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.029366371185356552,
      "loss": 0.7927,
      "num_input_tokens_seen": 14304208,
      "step": 24650
    },
    {
      "epoch": 3.6721775394697644,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029365810392304466,
      "loss": 0.7917,
      "num_input_tokens_seen": 14307504,
      "step": 24655
    },
    {
      "epoch": 3.672922252010724,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029365249356556614,
      "loss": 0.8006,
      "num_input_tokens_seen": 14310736,
      "step": 24660
    },
    {
      "epoch": 3.673666964551683,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02936468807812247,
      "loss": 0.7823,
      "num_input_tokens_seen": 14313776,
      "step": 24665
    },
    {
      "epoch": 3.6744116770926425,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.029364126557011525,
      "loss": 0.8135,
      "num_input_tokens_seen": 14316784,
      "step": 24670
    },
    {
      "epoch": 3.6751563896336013,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.029363564793233262,
      "loss": 0.8027,
      "num_input_tokens_seen": 14319440,
      "step": 24675
    },
    {
      "epoch": 3.675901102174561,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.029363002786797168,
      "loss": 0.8099,
      "num_input_tokens_seen": 14322384,
      "step": 24680
    },
    {
      "epoch": 3.6766458147155197,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02936244053771274,
      "loss": 0.7721,
      "num_input_tokens_seen": 14325200,
      "step": 24685
    },
    {
      "epoch": 3.677390527256479,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029361878045989477,
      "loss": 0.8283,
      "num_input_tokens_seen": 14328144,
      "step": 24690
    },
    {
      "epoch": 3.678135239797438,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029361315311636875,
      "loss": 0.8138,
      "num_input_tokens_seen": 14330864,
      "step": 24695
    },
    {
      "epoch": 3.6788799523383973,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029360752334664456,
      "loss": 0.8094,
      "num_input_tokens_seen": 14333808,
      "step": 24700
    },
    {
      "epoch": 3.6796246648793565,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029360189115081717,
      "loss": 0.8027,
      "num_input_tokens_seen": 14336816,
      "step": 24705
    },
    {
      "epoch": 3.6803693774203157,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02935962565289818,
      "loss": 0.7653,
      "num_input_tokens_seen": 14340016,
      "step": 24710
    },
    {
      "epoch": 3.681114089961275,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029359061948123363,
      "loss": 0.8016,
      "num_input_tokens_seen": 14343664,
      "step": 24715
    },
    {
      "epoch": 3.681858802502234,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029358498000766786,
      "loss": 0.7921,
      "num_input_tokens_seen": 14346480,
      "step": 24720
    },
    {
      "epoch": 3.6826035150431933,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.02935793381083798,
      "loss": 0.8104,
      "num_input_tokens_seen": 14349136,
      "step": 24725
    },
    {
      "epoch": 3.6833482275841525,
      "grad_norm": 0.02734375,
      "learning_rate": 0.029357369378346473,
      "loss": 0.8109,
      "num_input_tokens_seen": 14351760,
      "step": 24730
    },
    {
      "epoch": 3.6840929401251117,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029356804703301806,
      "loss": 0.8025,
      "num_input_tokens_seen": 14354640,
      "step": 24735
    },
    {
      "epoch": 3.684837652666071,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.029356239785713514,
      "loss": 0.8076,
      "num_input_tokens_seen": 14357680,
      "step": 24740
    },
    {
      "epoch": 3.68558236520703,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02935567462559114,
      "loss": 0.8182,
      "num_input_tokens_seen": 14360464,
      "step": 24745
    },
    {
      "epoch": 3.6863270777479893,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.029355109222944233,
      "loss": 0.8228,
      "num_input_tokens_seen": 14363408,
      "step": 24750
    },
    {
      "epoch": 3.6870717902889485,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02935454357778235,
      "loss": 0.8234,
      "num_input_tokens_seen": 14366512,
      "step": 24755
    },
    {
      "epoch": 3.6878165028299077,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029353977690115038,
      "loss": 0.799,
      "num_input_tokens_seen": 14369424,
      "step": 24760
    },
    {
      "epoch": 3.688561215370867,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.029353411559951862,
      "loss": 0.8044,
      "num_input_tokens_seen": 14372400,
      "step": 24765
    },
    {
      "epoch": 3.689305927911826,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.029352845187302388,
      "loss": 0.8018,
      "num_input_tokens_seen": 14375280,
      "step": 24770
    },
    {
      "epoch": 3.6900506404527853,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029352278572176184,
      "loss": 0.7956,
      "num_input_tokens_seen": 14378288,
      "step": 24775
    },
    {
      "epoch": 3.6907953529937445,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02935171171458282,
      "loss": 0.803,
      "num_input_tokens_seen": 14381200,
      "step": 24780
    },
    {
      "epoch": 3.6915400655347037,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.029351144614531866,
      "loss": 0.8204,
      "num_input_tokens_seen": 14384560,
      "step": 24785
    },
    {
      "epoch": 3.692284778075663,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029350577272032916,
      "loss": 0.7817,
      "num_input_tokens_seen": 14387408,
      "step": 24790
    },
    {
      "epoch": 3.693029490616622,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.029350009687095547,
      "loss": 0.8058,
      "num_input_tokens_seen": 14390512,
      "step": 24795
    },
    {
      "epoch": 3.6937742031575813,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02934944185972935,
      "loss": 0.7783,
      "num_input_tokens_seen": 14393296,
      "step": 24800
    },
    {
      "epoch": 3.6945189156985405,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029348873789943913,
      "loss": 0.7783,
      "num_input_tokens_seen": 14395984,
      "step": 24805
    },
    {
      "epoch": 3.6952636282394993,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029348305477748842,
      "loss": 0.8154,
      "num_input_tokens_seen": 14399152,
      "step": 24810
    },
    {
      "epoch": 3.696008340780459,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02934773692315373,
      "loss": 0.7934,
      "num_input_tokens_seen": 14402064,
      "step": 24815
    },
    {
      "epoch": 3.6967530533214177,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029347168126168183,
      "loss": 0.7985,
      "num_input_tokens_seen": 14404688,
      "step": 24820
    },
    {
      "epoch": 3.6974977658623773,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029346599086801817,
      "loss": 0.7933,
      "num_input_tokens_seen": 14407600,
      "step": 24825
    },
    {
      "epoch": 3.698242478403336,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02934602980506424,
      "loss": 0.7922,
      "num_input_tokens_seen": 14410416,
      "step": 24830
    },
    {
      "epoch": 3.6989871909442957,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029345460280965067,
      "loss": 0.811,
      "num_input_tokens_seen": 14413296,
      "step": 24835
    },
    {
      "epoch": 3.6997319034852545,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029344890514513926,
      "loss": 0.7902,
      "num_input_tokens_seen": 14415792,
      "step": 24840
    },
    {
      "epoch": 3.700476616026214,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029344320505720434,
      "loss": 0.7971,
      "num_input_tokens_seen": 14418896,
      "step": 24845
    },
    {
      "epoch": 3.701221328567173,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.029343750254594233,
      "loss": 0.7795,
      "num_input_tokens_seen": 14421584,
      "step": 24850
    },
    {
      "epoch": 3.701966041108132,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.029343179761144946,
      "loss": 0.8055,
      "num_input_tokens_seen": 14424528,
      "step": 24855
    },
    {
      "epoch": 3.7027107536490913,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.029342609025382212,
      "loss": 0.7826,
      "num_input_tokens_seen": 14427888,
      "step": 24860
    },
    {
      "epoch": 3.7034554661900505,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029342038047315683,
      "loss": 0.7858,
      "num_input_tokens_seen": 14430768,
      "step": 24865
    },
    {
      "epoch": 3.7042001787310097,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02934146682695499,
      "loss": 0.7985,
      "num_input_tokens_seen": 14433872,
      "step": 24870
    },
    {
      "epoch": 3.704944891271969,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.029340895364309796,
      "loss": 0.8313,
      "num_input_tokens_seen": 14436624,
      "step": 24875
    },
    {
      "epoch": 3.705689603812928,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02934032365938975,
      "loss": 0.7622,
      "num_input_tokens_seen": 14439952,
      "step": 24880
    },
    {
      "epoch": 3.7064343163538873,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.029339751712204508,
      "loss": 0.8433,
      "num_input_tokens_seen": 14443056,
      "step": 24885
    },
    {
      "epoch": 3.7071790288948465,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029339179522763735,
      "loss": 0.8395,
      "num_input_tokens_seen": 14445808,
      "step": 24890
    },
    {
      "epoch": 3.7079237414358057,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.0293386070910771,
      "loss": 0.7872,
      "num_input_tokens_seen": 14448624,
      "step": 24895
    },
    {
      "epoch": 3.708668453976765,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.029338034417154272,
      "loss": 0.8167,
      "num_input_tokens_seen": 14451472,
      "step": 24900
    },
    {
      "epoch": 3.709413166517724,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029337461501004924,
      "loss": 0.7728,
      "num_input_tokens_seen": 14454320,
      "step": 24905
    },
    {
      "epoch": 3.7101578790586833,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029336888342638735,
      "loss": 0.7932,
      "num_input_tokens_seen": 14457424,
      "step": 24910
    },
    {
      "epoch": 3.7109025915996425,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029336314942065386,
      "loss": 0.8194,
      "num_input_tokens_seen": 14460336,
      "step": 24915
    },
    {
      "epoch": 3.7116473041406017,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02933574129929457,
      "loss": 0.8157,
      "num_input_tokens_seen": 14463280,
      "step": 24920
    },
    {
      "epoch": 3.712392016681561,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029335167414335978,
      "loss": 0.8148,
      "num_input_tokens_seen": 14466160,
      "step": 24925
    },
    {
      "epoch": 3.71313672922252,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029334593287199298,
      "loss": 0.7947,
      "num_input_tokens_seen": 14468880,
      "step": 24930
    },
    {
      "epoch": 3.7138814417634793,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029334018917894233,
      "loss": 0.7957,
      "num_input_tokens_seen": 14471632,
      "step": 24935
    },
    {
      "epoch": 3.7146261543044385,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02933344430643049,
      "loss": 0.7904,
      "num_input_tokens_seen": 14474608,
      "step": 24940
    },
    {
      "epoch": 3.7153708668453977,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.02933286945281777,
      "loss": 0.8125,
      "num_input_tokens_seen": 14477680,
      "step": 24945
    },
    {
      "epoch": 3.716115579386357,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02933229435706579,
      "loss": 0.7782,
      "num_input_tokens_seen": 14480496,
      "step": 24950
    },
    {
      "epoch": 3.716860291927316,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.029331719019184264,
      "loss": 0.8203,
      "num_input_tokens_seen": 14483472,
      "step": 24955
    },
    {
      "epoch": 3.7176050044682754,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02933114343918291,
      "loss": 0.814,
      "num_input_tokens_seen": 14486224,
      "step": 24960
    },
    {
      "epoch": 3.7183497170092346,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02933056761707145,
      "loss": 0.8495,
      "num_input_tokens_seen": 14489456,
      "step": 24965
    },
    {
      "epoch": 3.7190944295501938,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029329991552859622,
      "loss": 0.7973,
      "num_input_tokens_seen": 14492784,
      "step": 24970
    },
    {
      "epoch": 3.719839142091153,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.029329415246557147,
      "loss": 0.7837,
      "num_input_tokens_seen": 14495664,
      "step": 24975
    },
    {
      "epoch": 3.720583854632112,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.029328838698173765,
      "loss": 0.8257,
      "num_input_tokens_seen": 14498640,
      "step": 24980
    },
    {
      "epoch": 3.721328567173071,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029328261907719217,
      "loss": 0.8173,
      "num_input_tokens_seen": 14501488,
      "step": 24985
    },
    {
      "epoch": 3.7220732797140306,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02932768487520325,
      "loss": 0.7944,
      "num_input_tokens_seen": 14504528,
      "step": 24990
    },
    {
      "epoch": 3.7228179922549893,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029327107600635602,
      "loss": 0.8043,
      "num_input_tokens_seen": 14507472,
      "step": 24995
    },
    {
      "epoch": 3.723562704795949,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02932653008402604,
      "loss": 0.8169,
      "num_input_tokens_seen": 14510256,
      "step": 25000
    },
    {
      "epoch": 3.7243074173369077,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02932595232538431,
      "loss": 0.7947,
      "num_input_tokens_seen": 14513200,
      "step": 25005
    },
    {
      "epoch": 3.7250521298778674,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029325374324720172,
      "loss": 0.8228,
      "num_input_tokens_seen": 14515792,
      "step": 25010
    },
    {
      "epoch": 3.725796842418826,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.0293247960820434,
      "loss": 0.8204,
      "num_input_tokens_seen": 14518640,
      "step": 25015
    },
    {
      "epoch": 3.726541554959786,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.029324217597363755,
      "loss": 0.8022,
      "num_input_tokens_seen": 14521904,
      "step": 25020
    },
    {
      "epoch": 3.7272862675007445,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02932363887069101,
      "loss": 0.7905,
      "num_input_tokens_seen": 14524848,
      "step": 25025
    },
    {
      "epoch": 3.7280309800417037,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.029323059902034945,
      "loss": 0.7833,
      "num_input_tokens_seen": 14527600,
      "step": 25030
    },
    {
      "epoch": 3.728775692582663,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02932248069140534,
      "loss": 0.8158,
      "num_input_tokens_seen": 14530352,
      "step": 25035
    },
    {
      "epoch": 3.729520405123622,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.029321901238811985,
      "loss": 0.7883,
      "num_input_tokens_seen": 14533328,
      "step": 25040
    },
    {
      "epoch": 3.7302651176645814,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029321321544264662,
      "loss": 0.797,
      "num_input_tokens_seen": 14536624,
      "step": 25045
    },
    {
      "epoch": 3.7310098302055406,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029320741607773166,
      "loss": 0.8052,
      "num_input_tokens_seen": 14539280,
      "step": 25050
    },
    {
      "epoch": 3.7317545427464998,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029320161429347298,
      "loss": 0.8096,
      "num_input_tokens_seen": 14542000,
      "step": 25055
    },
    {
      "epoch": 3.732499255287459,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.029319581008996856,
      "loss": 0.7849,
      "num_input_tokens_seen": 14545072,
      "step": 25060
    },
    {
      "epoch": 3.733243967828418,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029319000346731645,
      "loss": 0.8053,
      "num_input_tokens_seen": 14547920,
      "step": 25065
    },
    {
      "epoch": 3.7339886803693774,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029318419442561475,
      "loss": 0.7988,
      "num_input_tokens_seen": 14550800,
      "step": 25070
    },
    {
      "epoch": 3.7347333929103366,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02931783829649616,
      "loss": 0.7987,
      "num_input_tokens_seen": 14553680,
      "step": 25075
    },
    {
      "epoch": 3.7354781054512958,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.029317256908545523,
      "loss": 0.7823,
      "num_input_tokens_seen": 14556624,
      "step": 25080
    },
    {
      "epoch": 3.736222817992255,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.02931667527871938,
      "loss": 0.816,
      "num_input_tokens_seen": 14559568,
      "step": 25085
    },
    {
      "epoch": 3.736967530533214,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02931609340702756,
      "loss": 0.8132,
      "num_input_tokens_seen": 14562608,
      "step": 25090
    },
    {
      "epoch": 3.7377122430741734,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029315511293479896,
      "loss": 0.8232,
      "num_input_tokens_seen": 14565456,
      "step": 25095
    },
    {
      "epoch": 3.7384569556151326,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029314928938086213,
      "loss": 0.8044,
      "num_input_tokens_seen": 14568304,
      "step": 25100
    },
    {
      "epoch": 3.739201668156092,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029314346340856355,
      "loss": 0.7972,
      "num_input_tokens_seen": 14571248,
      "step": 25105
    },
    {
      "epoch": 3.739946380697051,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.029313763501800167,
      "loss": 0.805,
      "num_input_tokens_seen": 14574096,
      "step": 25110
    },
    {
      "epoch": 3.74069109323801,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029313180420927492,
      "loss": 0.7784,
      "num_input_tokens_seen": 14576688,
      "step": 25115
    },
    {
      "epoch": 3.7414358057789694,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02931259709824818,
      "loss": 0.8001,
      "num_input_tokens_seen": 14579504,
      "step": 25120
    },
    {
      "epoch": 3.7421805183199286,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029312013533772083,
      "loss": 0.8088,
      "num_input_tokens_seen": 14582320,
      "step": 25125
    },
    {
      "epoch": 3.742925230860888,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.029311429727509067,
      "loss": 0.822,
      "num_input_tokens_seen": 14585072,
      "step": 25130
    },
    {
      "epoch": 3.743669943401847,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.029310845679468994,
      "loss": 0.822,
      "num_input_tokens_seen": 14588016,
      "step": 25135
    },
    {
      "epoch": 3.744414655942806,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.029310261389661724,
      "loss": 0.8003,
      "num_input_tokens_seen": 14591088,
      "step": 25140
    },
    {
      "epoch": 3.7451593684837654,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02930967685809713,
      "loss": 0.8001,
      "num_input_tokens_seen": 14593712,
      "step": 25145
    },
    {
      "epoch": 3.7459040810247246,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.029309092084785094,
      "loss": 0.7849,
      "num_input_tokens_seen": 14596528,
      "step": 25150
    },
    {
      "epoch": 3.746648793565684,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02930850706973549,
      "loss": 0.8066,
      "num_input_tokens_seen": 14599280,
      "step": 25155
    },
    {
      "epoch": 3.7473935061066426,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.0293079218129582,
      "loss": 0.8139,
      "num_input_tokens_seen": 14602352,
      "step": 25160
    },
    {
      "epoch": 3.748138218647602,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029307336314463115,
      "loss": 0.802,
      "num_input_tokens_seen": 14605264,
      "step": 25165
    },
    {
      "epoch": 3.748882931188561,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029306750574260118,
      "loss": 0.7929,
      "num_input_tokens_seen": 14608208,
      "step": 25170
    },
    {
      "epoch": 3.7496276437295206,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029306164592359116,
      "loss": 0.7878,
      "num_input_tokens_seen": 14611376,
      "step": 25175
    },
    {
      "epoch": 3.7503723562704794,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.029305578368770006,
      "loss": 0.7914,
      "num_input_tokens_seen": 14614512,
      "step": 25180
    },
    {
      "epoch": 3.751117068811439,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.029304991903502685,
      "loss": 0.776,
      "num_input_tokens_seen": 14617424,
      "step": 25185
    },
    {
      "epoch": 3.751861781352398,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029304405196567067,
      "loss": 0.7739,
      "num_input_tokens_seen": 14620400,
      "step": 25190
    },
    {
      "epoch": 3.7526064938933574,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.02930381824797306,
      "loss": 0.8325,
      "num_input_tokens_seen": 14623728,
      "step": 25195
    },
    {
      "epoch": 3.753351206434316,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02930323105773058,
      "loss": 0.7735,
      "num_input_tokens_seen": 14626864,
      "step": 25200
    },
    {
      "epoch": 3.7540959189752754,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029302643625849552,
      "loss": 0.8028,
      "num_input_tokens_seen": 14629712,
      "step": 25205
    },
    {
      "epoch": 3.7548406315162346,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029302055952339895,
      "loss": 0.8089,
      "num_input_tokens_seen": 14632432,
      "step": 25210
    },
    {
      "epoch": 3.755585344057194,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029301468037211537,
      "loss": 0.7719,
      "num_input_tokens_seen": 14635216,
      "step": 25215
    },
    {
      "epoch": 3.756330056598153,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029300879880474414,
      "loss": 0.8385,
      "num_input_tokens_seen": 14638160,
      "step": 25220
    },
    {
      "epoch": 3.757074769139112,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02930029148213846,
      "loss": 0.8246,
      "num_input_tokens_seen": 14641360,
      "step": 25225
    },
    {
      "epoch": 3.7578194816800714,
      "grad_norm": 0.03125,
      "learning_rate": 0.029299702842213612,
      "loss": 0.766,
      "num_input_tokens_seen": 14644432,
      "step": 25230
    },
    {
      "epoch": 3.7585641942210306,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029299113960709822,
      "loss": 0.7909,
      "num_input_tokens_seen": 14647856,
      "step": 25235
    },
    {
      "epoch": 3.75930890676199,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029298524837637032,
      "loss": 0.7946,
      "num_input_tokens_seen": 14650704,
      "step": 25240
    },
    {
      "epoch": 3.760053619302949,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029297935473005198,
      "loss": 0.7919,
      "num_input_tokens_seen": 14653552,
      "step": 25245
    },
    {
      "epoch": 3.760798331843908,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029297345866824278,
      "loss": 0.8039,
      "num_input_tokens_seen": 14656688,
      "step": 25250
    },
    {
      "epoch": 3.7615430443848674,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02929675601910423,
      "loss": 0.8435,
      "num_input_tokens_seen": 14659664,
      "step": 25255
    },
    {
      "epoch": 3.7622877569258266,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02929616592985502,
      "loss": 0.7957,
      "num_input_tokens_seen": 14662608,
      "step": 25260
    },
    {
      "epoch": 3.763032469466786,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.029295575599086614,
      "loss": 0.794,
      "num_input_tokens_seen": 14665584,
      "step": 25265
    },
    {
      "epoch": 3.763777182007745,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.029294985026808988,
      "loss": 0.7909,
      "num_input_tokens_seen": 14668560,
      "step": 25270
    },
    {
      "epoch": 3.7645218945487042,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02929439421303212,
      "loss": 0.8085,
      "num_input_tokens_seen": 14671184,
      "step": 25275
    },
    {
      "epoch": 3.7652666070896634,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.02929380315776599,
      "loss": 0.8132,
      "num_input_tokens_seen": 14674064,
      "step": 25280
    },
    {
      "epoch": 3.7660113196306226,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.029293211861020584,
      "loss": 0.8026,
      "num_input_tokens_seen": 14676848,
      "step": 25285
    },
    {
      "epoch": 3.766756032171582,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029292620322805888,
      "loss": 0.8001,
      "num_input_tokens_seen": 14679664,
      "step": 25290
    },
    {
      "epoch": 3.767500744712541,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029292028543131902,
      "loss": 0.7777,
      "num_input_tokens_seen": 14682704,
      "step": 25295
    },
    {
      "epoch": 3.7682454572535002,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.029291436522008617,
      "loss": 0.806,
      "num_input_tokens_seen": 14685456,
      "step": 25300
    },
    {
      "epoch": 3.7689901697944594,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029290844259446037,
      "loss": 0.8172,
      "num_input_tokens_seen": 14688592,
      "step": 25305
    },
    {
      "epoch": 3.7697348823354186,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029290251755454168,
      "loss": 0.7665,
      "num_input_tokens_seen": 14691760,
      "step": 25310
    },
    {
      "epoch": 3.770479594876378,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.02928965901004302,
      "loss": 0.8023,
      "num_input_tokens_seen": 14694768,
      "step": 25315
    },
    {
      "epoch": 3.771224307417337,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029289066023222607,
      "loss": 0.8167,
      "num_input_tokens_seen": 14698096,
      "step": 25320
    },
    {
      "epoch": 3.7719690199582963,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029288472795002945,
      "loss": 0.8009,
      "num_input_tokens_seen": 14700816,
      "step": 25325
    },
    {
      "epoch": 3.7727137324992555,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029287879325394055,
      "loss": 0.7631,
      "num_input_tokens_seen": 14703664,
      "step": 25330
    },
    {
      "epoch": 3.773458445040214,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02928728561440597,
      "loss": 0.8174,
      "num_input_tokens_seen": 14706672,
      "step": 25335
    },
    {
      "epoch": 3.774203157581174,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.02928669166204871,
      "loss": 0.8257,
      "num_input_tokens_seen": 14709552,
      "step": 25340
    },
    {
      "epoch": 3.7749478701221326,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02928609746833232,
      "loss": 0.7775,
      "num_input_tokens_seen": 14712720,
      "step": 25345
    },
    {
      "epoch": 3.7756925826630923,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02928550303326683,
      "loss": 0.7973,
      "num_input_tokens_seen": 14715280,
      "step": 25350
    },
    {
      "epoch": 3.776437295204051,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029284908356862288,
      "loss": 0.8129,
      "num_input_tokens_seen": 14717968,
      "step": 25355
    },
    {
      "epoch": 3.7771820077450107,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029284313439128736,
      "loss": 0.7804,
      "num_input_tokens_seen": 14720560,
      "step": 25360
    },
    {
      "epoch": 3.7779267202859694,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02928371828007623,
      "loss": 0.8067,
      "num_input_tokens_seen": 14723472,
      "step": 25365
    },
    {
      "epoch": 3.778671432826929,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.029283122879714813,
      "loss": 0.8064,
      "num_input_tokens_seen": 14726256,
      "step": 25370
    },
    {
      "epoch": 3.779416145367888,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02928252723805456,
      "loss": 0.8434,
      "num_input_tokens_seen": 14728944,
      "step": 25375
    },
    {
      "epoch": 3.780160857908847,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029281931355105522,
      "loss": 0.7805,
      "num_input_tokens_seen": 14731664,
      "step": 25380
    },
    {
      "epoch": 3.7809055704498062,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029281335230877767,
      "loss": 0.837,
      "num_input_tokens_seen": 14734736,
      "step": 25385
    },
    {
      "epoch": 3.7816502829907654,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.02928073886538137,
      "loss": 0.8258,
      "num_input_tokens_seen": 14737808,
      "step": 25390
    },
    {
      "epoch": 3.7823949955317246,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.029280142258626407,
      "loss": 0.8225,
      "num_input_tokens_seen": 14740688,
      "step": 25395
    },
    {
      "epoch": 3.783139708072684,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02927954541062295,
      "loss": 0.804,
      "num_input_tokens_seen": 14743344,
      "step": 25400
    },
    {
      "epoch": 3.783884420613643,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02927894832138109,
      "loss": 0.7913,
      "num_input_tokens_seen": 14746096,
      "step": 25405
    },
    {
      "epoch": 3.7846291331546023,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.029278350990910907,
      "loss": 0.8062,
      "num_input_tokens_seen": 14749328,
      "step": 25410
    },
    {
      "epoch": 3.7853738456955615,
      "grad_norm": 0.029296875,
      "learning_rate": 0.029277753419222497,
      "loss": 0.8166,
      "num_input_tokens_seen": 14752144,
      "step": 25415
    },
    {
      "epoch": 3.7861185582365207,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029277155606325957,
      "loss": 0.8119,
      "num_input_tokens_seen": 14754992,
      "step": 25420
    },
    {
      "epoch": 3.78686327077748,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.029276557552231384,
      "loss": 0.7998,
      "num_input_tokens_seen": 14757712,
      "step": 25425
    },
    {
      "epoch": 3.787607983318439,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029275959256948876,
      "loss": 0.7938,
      "num_input_tokens_seen": 14760720,
      "step": 25430
    },
    {
      "epoch": 3.7883526958593983,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02927536072048855,
      "loss": 0.7948,
      "num_input_tokens_seen": 14763472,
      "step": 25435
    },
    {
      "epoch": 3.7890974084003575,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02927476194286051,
      "loss": 0.7871,
      "num_input_tokens_seen": 14766192,
      "step": 25440
    },
    {
      "epoch": 3.7898421209413167,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.02927416292407488,
      "loss": 0.8068,
      "num_input_tokens_seen": 14769008,
      "step": 25445
    },
    {
      "epoch": 3.790586833482276,
      "grad_norm": 0.01544189453125,
      "learning_rate": 0.029273563664141772,
      "loss": 0.816,
      "num_input_tokens_seen": 14771984,
      "step": 25450
    },
    {
      "epoch": 3.791331546023235,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029272964163071312,
      "loss": 0.8083,
      "num_input_tokens_seen": 14775152,
      "step": 25455
    },
    {
      "epoch": 3.7920762585641943,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02927236442087363,
      "loss": 0.8143,
      "num_input_tokens_seen": 14778000,
      "step": 25460
    },
    {
      "epoch": 3.7928209711051535,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029271764437558856,
      "loss": 0.8009,
      "num_input_tokens_seen": 14780912,
      "step": 25465
    },
    {
      "epoch": 3.7935656836461127,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02927116421313713,
      "loss": 0.792,
      "num_input_tokens_seen": 14783888,
      "step": 25470
    },
    {
      "epoch": 3.794310396187072,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029270563747618588,
      "loss": 0.7905,
      "num_input_tokens_seen": 14786896,
      "step": 25475
    },
    {
      "epoch": 3.795055108728031,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029269963041013373,
      "loss": 0.7992,
      "num_input_tokens_seen": 14789968,
      "step": 25480
    },
    {
      "epoch": 3.7957998212689903,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.029269362093331638,
      "loss": 0.8143,
      "num_input_tokens_seen": 14792816,
      "step": 25485
    },
    {
      "epoch": 3.7965445338099495,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.029268760904583527,
      "loss": 0.789,
      "num_input_tokens_seen": 14795760,
      "step": 25490
    },
    {
      "epoch": 3.7972892463509087,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.029268159474779207,
      "loss": 0.7934,
      "num_input_tokens_seen": 14798416,
      "step": 25495
    },
    {
      "epoch": 3.798033958891868,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02926755780392884,
      "loss": 0.817,
      "num_input_tokens_seen": 14801264,
      "step": 25500
    },
    {
      "epoch": 3.798778671432827,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.02926695589204258,
      "loss": 0.8071,
      "num_input_tokens_seen": 14803888,
      "step": 25505
    },
    {
      "epoch": 3.799523383973786,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.0292663537391306,
      "loss": 0.7667,
      "num_input_tokens_seen": 14806640,
      "step": 25510
    },
    {
      "epoch": 3.8002680965147455,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029265751345203074,
      "loss": 0.8097,
      "num_input_tokens_seen": 14809552,
      "step": 25515
    },
    {
      "epoch": 3.8010128090557043,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.029265148710270177,
      "loss": 0.826,
      "num_input_tokens_seen": 14812592,
      "step": 25520
    },
    {
      "epoch": 3.801757521596664,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02926454583434209,
      "loss": 0.7996,
      "num_input_tokens_seen": 14815632,
      "step": 25525
    },
    {
      "epoch": 3.8025022341376227,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.029263942717429,
      "loss": 0.8125,
      "num_input_tokens_seen": 14818544,
      "step": 25530
    },
    {
      "epoch": 3.8032469466785823,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.029263339359541098,
      "loss": 0.8004,
      "num_input_tokens_seen": 14821200,
      "step": 25535
    },
    {
      "epoch": 3.803991659219541,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029262735760688573,
      "loss": 0.8027,
      "num_input_tokens_seen": 14824080,
      "step": 25540
    },
    {
      "epoch": 3.8047363717605007,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029262131920881618,
      "loss": 0.7823,
      "num_input_tokens_seen": 14826864,
      "step": 25545
    },
    {
      "epoch": 3.8054810843014595,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.029261527840130447,
      "loss": 0.8099,
      "num_input_tokens_seen": 14830096,
      "step": 25550
    },
    {
      "epoch": 3.8062257968424187,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029260923518445253,
      "loss": 0.7999,
      "num_input_tokens_seen": 14833104,
      "step": 25555
    },
    {
      "epoch": 3.806970509383378,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.029260318955836254,
      "loss": 0.7838,
      "num_input_tokens_seen": 14836432,
      "step": 25560
    },
    {
      "epoch": 3.807715221924337,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02925971415231366,
      "loss": 0.8037,
      "num_input_tokens_seen": 14839184,
      "step": 25565
    },
    {
      "epoch": 3.8084599344652963,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029259109107887687,
      "loss": 0.8136,
      "num_input_tokens_seen": 14842064,
      "step": 25570
    },
    {
      "epoch": 3.8092046470062555,
      "grad_norm": 0.025390625,
      "learning_rate": 0.029258503822568553,
      "loss": 0.8073,
      "num_input_tokens_seen": 14844784,
      "step": 25575
    },
    {
      "epoch": 3.8099493595472147,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02925789829636649,
      "loss": 0.8006,
      "num_input_tokens_seen": 14847728,
      "step": 25580
    },
    {
      "epoch": 3.810694072088174,
      "grad_norm": 0.0126953125,
      "learning_rate": 0.02925729252929173,
      "loss": 0.809,
      "num_input_tokens_seen": 14850768,
      "step": 25585
    },
    {
      "epoch": 3.811438784629133,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.029256686521354505,
      "loss": 0.8114,
      "num_input_tokens_seen": 14853360,
      "step": 25590
    },
    {
      "epoch": 3.8121834971700923,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02925608027256505,
      "loss": 0.7998,
      "num_input_tokens_seen": 14856400,
      "step": 25595
    },
    {
      "epoch": 3.8129282097110515,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029255473782933607,
      "loss": 0.8054,
      "num_input_tokens_seen": 14859344,
      "step": 25600
    },
    {
      "epoch": 3.8136729222520107,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02925486705247042,
      "loss": 0.7826,
      "num_input_tokens_seen": 14862288,
      "step": 25605
    },
    {
      "epoch": 3.81441763479297,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029254260081185743,
      "loss": 0.7906,
      "num_input_tokens_seen": 14865360,
      "step": 25610
    },
    {
      "epoch": 3.815162347333929,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029253652869089833,
      "loss": 0.7867,
      "num_input_tokens_seen": 14868112,
      "step": 25615
    },
    {
      "epoch": 3.8159070598748883,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02925304541619294,
      "loss": 0.8131,
      "num_input_tokens_seen": 14871152,
      "step": 25620
    },
    {
      "epoch": 3.8166517724158475,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029252437722505336,
      "loss": 0.8332,
      "num_input_tokens_seen": 14874096,
      "step": 25625
    },
    {
      "epoch": 3.8173964849568067,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.029251829788037278,
      "loss": 0.8162,
      "num_input_tokens_seen": 14877136,
      "step": 25630
    },
    {
      "epoch": 3.818141197497766,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.02925122161279904,
      "loss": 0.8046,
      "num_input_tokens_seen": 14879952,
      "step": 25635
    },
    {
      "epoch": 3.818885910038725,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.0292506131968009,
      "loss": 0.8054,
      "num_input_tokens_seen": 14882736,
      "step": 25640
    },
    {
      "epoch": 3.8196306225796843,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.029250004540053134,
      "loss": 0.8272,
      "num_input_tokens_seen": 14885968,
      "step": 25645
    },
    {
      "epoch": 3.8203753351206435,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.029249395642566026,
      "loss": 0.8021,
      "num_input_tokens_seen": 14889008,
      "step": 25650
    },
    {
      "epoch": 3.8211200476616027,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.029248786504349856,
      "loss": 0.7977,
      "num_input_tokens_seen": 14891920,
      "step": 25655
    },
    {
      "epoch": 3.821864760202562,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.029248177125414925,
      "loss": 0.7964,
      "num_input_tokens_seen": 14894448,
      "step": 25660
    },
    {
      "epoch": 3.822609472743521,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.029247567505771517,
      "loss": 0.8029,
      "num_input_tokens_seen": 14897424,
      "step": 25665
    },
    {
      "epoch": 3.8233541852844803,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029246957645429936,
      "loss": 0.8039,
      "num_input_tokens_seen": 14900240,
      "step": 25670
    },
    {
      "epoch": 3.824098897825439,
      "grad_norm": 0.025390625,
      "learning_rate": 0.029246347544400492,
      "loss": 0.7883,
      "num_input_tokens_seen": 14903248,
      "step": 25675
    },
    {
      "epoch": 3.8248436103663987,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02924573720269348,
      "loss": 0.8038,
      "num_input_tokens_seen": 14906256,
      "step": 25680
    },
    {
      "epoch": 3.8255883229073575,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029245126620319223,
      "loss": 0.8043,
      "num_input_tokens_seen": 14909264,
      "step": 25685
    },
    {
      "epoch": 3.826333035448317,
      "grad_norm": 0.0123291015625,
      "learning_rate": 0.029244515797288023,
      "loss": 0.8054,
      "num_input_tokens_seen": 14911888,
      "step": 25690
    },
    {
      "epoch": 3.827077747989276,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029243904733610212,
      "loss": 0.8136,
      "num_input_tokens_seen": 14914672,
      "step": 25695
    },
    {
      "epoch": 3.8278224605302356,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.029243293429296102,
      "loss": 0.8124,
      "num_input_tokens_seen": 14917808,
      "step": 25700
    },
    {
      "epoch": 3.8285671730711943,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.02924268188435603,
      "loss": 0.7775,
      "num_input_tokens_seen": 14920848,
      "step": 25705
    },
    {
      "epoch": 3.829311885612154,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.02924207009880032,
      "loss": 0.7765,
      "num_input_tokens_seen": 14923664,
      "step": 25710
    },
    {
      "epoch": 3.8300565981531127,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029241458072639316,
      "loss": 0.8032,
      "num_input_tokens_seen": 14926320,
      "step": 25715
    },
    {
      "epoch": 3.830801310694072,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029240845805883348,
      "loss": 0.7922,
      "num_input_tokens_seen": 14929328,
      "step": 25720
    },
    {
      "epoch": 3.831546023235031,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02924023329854276,
      "loss": 0.7944,
      "num_input_tokens_seen": 14932304,
      "step": 25725
    },
    {
      "epoch": 3.8322907357759903,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02923962055062791,
      "loss": 0.8118,
      "num_input_tokens_seen": 14935216,
      "step": 25730
    },
    {
      "epoch": 3.8330354483169495,
      "grad_norm": 0.015625,
      "learning_rate": 0.029239007562149144,
      "loss": 0.7952,
      "num_input_tokens_seen": 14938256,
      "step": 25735
    },
    {
      "epoch": 3.8337801608579087,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02923839433311681,
      "loss": 0.7912,
      "num_input_tokens_seen": 14941296,
      "step": 25740
    },
    {
      "epoch": 3.834524873398868,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02923778086354128,
      "loss": 0.7825,
      "num_input_tokens_seen": 14944080,
      "step": 25745
    },
    {
      "epoch": 3.835269585939827,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.02923716715343291,
      "loss": 0.816,
      "num_input_tokens_seen": 14947056,
      "step": 25750
    },
    {
      "epoch": 3.8360142984807863,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.029236553202802076,
      "loss": 0.7899,
      "num_input_tokens_seen": 14949968,
      "step": 25755
    },
    {
      "epoch": 3.8367590110217455,
      "grad_norm": 0.01190185546875,
      "learning_rate": 0.02923593901165914,
      "loss": 0.8019,
      "num_input_tokens_seen": 14952752,
      "step": 25760
    },
    {
      "epoch": 3.8375037235627047,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.029235324580014488,
      "loss": 0.8001,
      "num_input_tokens_seen": 14955920,
      "step": 25765
    },
    {
      "epoch": 3.838248436103664,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029234709907878492,
      "loss": 0.799,
      "num_input_tokens_seen": 14958832,
      "step": 25770
    },
    {
      "epoch": 3.838993148644623,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.029234094995261542,
      "loss": 0.793,
      "num_input_tokens_seen": 14961968,
      "step": 25775
    },
    {
      "epoch": 3.8397378611855824,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029233479842174018,
      "loss": 0.8162,
      "num_input_tokens_seen": 14964592,
      "step": 25780
    },
    {
      "epoch": 3.8404825737265416,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029232864448626324,
      "loss": 0.8099,
      "num_input_tokens_seen": 14967376,
      "step": 25785
    },
    {
      "epoch": 3.8412272862675008,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.02923224881462885,
      "loss": 0.7792,
      "num_input_tokens_seen": 14970352,
      "step": 25790
    },
    {
      "epoch": 3.84197199880846,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.029231632940191994,
      "loss": 0.8158,
      "num_input_tokens_seen": 14972880,
      "step": 25795
    },
    {
      "epoch": 3.842716711349419,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02923101682532617,
      "loss": 0.8037,
      "num_input_tokens_seen": 14975632,
      "step": 25800
    },
    {
      "epoch": 3.8434614238903784,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029230400470041776,
      "loss": 0.7762,
      "num_input_tokens_seen": 14978864,
      "step": 25805
    },
    {
      "epoch": 3.8442061364313376,
      "grad_norm": 0.03515625,
      "learning_rate": 0.029229783874349227,
      "loss": 0.8124,
      "num_input_tokens_seen": 14981584,
      "step": 25810
    },
    {
      "epoch": 3.8449508489722968,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02922916703825895,
      "loss": 0.8326,
      "num_input_tokens_seen": 14984240,
      "step": 25815
    },
    {
      "epoch": 3.845695561513256,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029228549961781352,
      "loss": 0.7808,
      "num_input_tokens_seen": 14987248,
      "step": 25820
    },
    {
      "epoch": 3.846440274054215,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.029227932644926864,
      "loss": 0.8073,
      "num_input_tokens_seen": 14990256,
      "step": 25825
    },
    {
      "epoch": 3.8471849865951744,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.029227315087705916,
      "loss": 0.8081,
      "num_input_tokens_seen": 14993264,
      "step": 25830
    },
    {
      "epoch": 3.8479296991361336,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029226697290128938,
      "loss": 0.7901,
      "num_input_tokens_seen": 14995952,
      "step": 25835
    },
    {
      "epoch": 3.848674411677093,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02922607925220637,
      "loss": 0.8105,
      "num_input_tokens_seen": 14999216,
      "step": 25840
    },
    {
      "epoch": 3.849419124218052,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.02922546097394865,
      "loss": 0.7878,
      "num_input_tokens_seen": 15002032,
      "step": 25845
    },
    {
      "epoch": 3.8501638367590107,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029224842455366225,
      "loss": 0.8221,
      "num_input_tokens_seen": 15004816,
      "step": 25850
    },
    {
      "epoch": 3.8509085492999704,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.029224223696469546,
      "loss": 0.8173,
      "num_input_tokens_seen": 15007856,
      "step": 25855
    },
    {
      "epoch": 3.851653261840929,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029223604697269062,
      "loss": 0.8188,
      "num_input_tokens_seen": 15010736,
      "step": 25860
    },
    {
      "epoch": 3.852397974381889,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.029222985457775238,
      "loss": 0.7944,
      "num_input_tokens_seen": 15013680,
      "step": 25865
    },
    {
      "epoch": 3.8531426869228476,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.029222365977998526,
      "loss": 0.8102,
      "num_input_tokens_seen": 15016528,
      "step": 25870
    },
    {
      "epoch": 3.853887399463807,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.029221746257949398,
      "loss": 0.8049,
      "num_input_tokens_seen": 15019728,
      "step": 25875
    },
    {
      "epoch": 3.854632112004766,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02922112629763832,
      "loss": 0.8159,
      "num_input_tokens_seen": 15022608,
      "step": 25880
    },
    {
      "epoch": 3.8553768245457256,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029220506097075764,
      "loss": 0.786,
      "num_input_tokens_seen": 15025584,
      "step": 25885
    },
    {
      "epoch": 3.8561215370866844,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029219885656272215,
      "loss": 0.8025,
      "num_input_tokens_seen": 15028560,
      "step": 25890
    },
    {
      "epoch": 3.8568662496276436,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02921926497523815,
      "loss": 0.8081,
      "num_input_tokens_seen": 15031376,
      "step": 25895
    },
    {
      "epoch": 3.8576109621686028,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.029218644053984052,
      "loss": 0.8001,
      "num_input_tokens_seen": 15034448,
      "step": 25900
    },
    {
      "epoch": 3.858355674709562,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.02921802289252042,
      "loss": 0.8046,
      "num_input_tokens_seen": 15037424,
      "step": 25905
    },
    {
      "epoch": 3.859100387250521,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.029217401490857733,
      "loss": 0.8126,
      "num_input_tokens_seen": 15040368,
      "step": 25910
    },
    {
      "epoch": 3.8598450997914804,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029216779849006502,
      "loss": 0.7754,
      "num_input_tokens_seen": 15042864,
      "step": 25915
    },
    {
      "epoch": 3.8605898123324396,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.029216157966977224,
      "loss": 0.7956,
      "num_input_tokens_seen": 15045584,
      "step": 25920
    },
    {
      "epoch": 3.861334524873399,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029215535844780406,
      "loss": 0.7944,
      "num_input_tokens_seen": 15048304,
      "step": 25925
    },
    {
      "epoch": 3.862079237414358,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02921491348242656,
      "loss": 0.7852,
      "num_input_tokens_seen": 15051504,
      "step": 25930
    },
    {
      "epoch": 3.862823949955317,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.0292142908799262,
      "loss": 0.8225,
      "num_input_tokens_seen": 15054384,
      "step": 25935
    },
    {
      "epoch": 3.8635686624962764,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02921366803728984,
      "loss": 0.7874,
      "num_input_tokens_seen": 15057488,
      "step": 25940
    },
    {
      "epoch": 3.8643133750372356,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.029213044954528,
      "loss": 0.8093,
      "num_input_tokens_seen": 15060240,
      "step": 25945
    },
    {
      "epoch": 3.865058087578195,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.029212421631651216,
      "loss": 0.7988,
      "num_input_tokens_seen": 15062992,
      "step": 25950
    },
    {
      "epoch": 3.865802800119154,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02921179806867001,
      "loss": 0.8112,
      "num_input_tokens_seen": 15065808,
      "step": 25955
    },
    {
      "epoch": 3.866547512660113,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.029211174265594924,
      "loss": 0.7961,
      "num_input_tokens_seen": 15068816,
      "step": 25960
    },
    {
      "epoch": 3.8672922252010724,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02921055022243649,
      "loss": 0.785,
      "num_input_tokens_seen": 15071696,
      "step": 25965
    },
    {
      "epoch": 3.8680369377420316,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029209925939205256,
      "loss": 0.7991,
      "num_input_tokens_seen": 15074512,
      "step": 25970
    },
    {
      "epoch": 3.868781650282991,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.029209301415911763,
      "loss": 0.8042,
      "num_input_tokens_seen": 15077680,
      "step": 25975
    },
    {
      "epoch": 3.86952636282395,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.02920867665256656,
      "loss": 0.7814,
      "num_input_tokens_seen": 15080464,
      "step": 25980
    },
    {
      "epoch": 3.870271075364909,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.029208051649180212,
      "loss": 0.817,
      "num_input_tokens_seen": 15083248,
      "step": 25985
    },
    {
      "epoch": 3.8710157879058684,
      "grad_norm": 0.015625,
      "learning_rate": 0.029207426405763272,
      "loss": 0.7972,
      "num_input_tokens_seen": 15086192,
      "step": 25990
    },
    {
      "epoch": 3.8717605004468276,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.0292068009223263,
      "loss": 0.7858,
      "num_input_tokens_seen": 15088944,
      "step": 25995
    },
    {
      "epoch": 3.872505212987787,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.029206175198879865,
      "loss": 0.8538,
      "num_input_tokens_seen": 15091824,
      "step": 26000
    },
    {
      "epoch": 3.873249925528746,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02920554923543454,
      "loss": 0.8288,
      "num_input_tokens_seen": 15094992,
      "step": 26005
    },
    {
      "epoch": 3.8739946380697052,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.0292049230320009,
      "loss": 0.8337,
      "num_input_tokens_seen": 15097904,
      "step": 26010
    },
    {
      "epoch": 3.8747393506106644,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029204296588589522,
      "loss": 0.8078,
      "num_input_tokens_seen": 15100848,
      "step": 26015
    },
    {
      "epoch": 3.8754840631516236,
      "grad_norm": 0.01129150390625,
      "learning_rate": 0.029203669905210984,
      "loss": 0.7955,
      "num_input_tokens_seen": 15103760,
      "step": 26020
    },
    {
      "epoch": 3.8762287756925824,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.029203042981875885,
      "loss": 0.8211,
      "num_input_tokens_seen": 15106448,
      "step": 26025
    },
    {
      "epoch": 3.876973488233542,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02920241581859481,
      "loss": 0.7784,
      "num_input_tokens_seen": 15109360,
      "step": 26030
    },
    {
      "epoch": 3.877718200774501,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02920178841537835,
      "loss": 0.807,
      "num_input_tokens_seen": 15112624,
      "step": 26035
    },
    {
      "epoch": 3.8784629133154604,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029201160772237115,
      "loss": 0.7832,
      "num_input_tokens_seen": 15115280,
      "step": 26040
    },
    {
      "epoch": 3.879207625856419,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0292005328891817,
      "loss": 0.8454,
      "num_input_tokens_seen": 15118000,
      "step": 26045
    },
    {
      "epoch": 3.879952338397379,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02919990476622271,
      "loss": 0.8052,
      "num_input_tokens_seen": 15120656,
      "step": 26050
    },
    {
      "epoch": 3.8806970509383376,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029199276403370764,
      "loss": 0.817,
      "num_input_tokens_seen": 15123728,
      "step": 26055
    },
    {
      "epoch": 3.8814417634792973,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.029198647800636482,
      "loss": 0.8173,
      "num_input_tokens_seen": 15126480,
      "step": 26060
    },
    {
      "epoch": 3.882186476020256,
      "grad_norm": 0.02734375,
      "learning_rate": 0.029198018958030467,
      "loss": 0.8087,
      "num_input_tokens_seen": 15129136,
      "step": 26065
    },
    {
      "epoch": 3.882931188561215,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.029197389875563355,
      "loss": 0.811,
      "num_input_tokens_seen": 15132208,
      "step": 26070
    },
    {
      "epoch": 3.8836759011021744,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02919676055324577,
      "loss": 0.7978,
      "num_input_tokens_seen": 15135056,
      "step": 26075
    },
    {
      "epoch": 3.8844206136431336,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.029196130991088344,
      "loss": 0.8051,
      "num_input_tokens_seen": 15137904,
      "step": 26080
    },
    {
      "epoch": 3.885165326184093,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029195501189101716,
      "loss": 0.8144,
      "num_input_tokens_seen": 15140656,
      "step": 26085
    },
    {
      "epoch": 3.885910038725052,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02919487114729652,
      "loss": 0.8014,
      "num_input_tokens_seen": 15143600,
      "step": 26090
    },
    {
      "epoch": 3.8866547512660112,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.029194240865683403,
      "loss": 0.8045,
      "num_input_tokens_seen": 15146768,
      "step": 26095
    },
    {
      "epoch": 3.8873994638069704,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.029193610344273015,
      "loss": 0.8152,
      "num_input_tokens_seen": 15149776,
      "step": 26100
    },
    {
      "epoch": 3.8881441763479296,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.029192979583076,
      "loss": 0.8039,
      "num_input_tokens_seen": 15152592,
      "step": 26105
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.02919234858210303,
      "loss": 0.8028,
      "num_input_tokens_seen": 15155216,
      "step": 26110
    },
    {
      "epoch": 3.889633601429848,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.02919171734136475,
      "loss": 0.8028,
      "num_input_tokens_seen": 15158352,
      "step": 26115
    },
    {
      "epoch": 3.8903783139708072,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.029191085860871827,
      "loss": 0.806,
      "num_input_tokens_seen": 15161328,
      "step": 26120
    },
    {
      "epoch": 3.8911230265117664,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029190454140634932,
      "loss": 0.8026,
      "num_input_tokens_seen": 15164208,
      "step": 26125
    },
    {
      "epoch": 3.8918677390527256,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02918982218066474,
      "loss": 0.8049,
      "num_input_tokens_seen": 15166992,
      "step": 26130
    },
    {
      "epoch": 3.892612451593685,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029189189980971923,
      "loss": 0.8055,
      "num_input_tokens_seen": 15171056,
      "step": 26135
    },
    {
      "epoch": 3.893357164134644,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029188557541567163,
      "loss": 0.8165,
      "num_input_tokens_seen": 15173840,
      "step": 26140
    },
    {
      "epoch": 3.8941018766756033,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029187924862461145,
      "loss": 0.8029,
      "num_input_tokens_seen": 15176752,
      "step": 26145
    },
    {
      "epoch": 3.8948465892165625,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029187291943664553,
      "loss": 0.8074,
      "num_input_tokens_seen": 15179440,
      "step": 26150
    },
    {
      "epoch": 3.8955913017575217,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.029186658785188085,
      "loss": 0.8102,
      "num_input_tokens_seen": 15182384,
      "step": 26155
    },
    {
      "epoch": 3.896336014298481,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029186025387042437,
      "loss": 0.8067,
      "num_input_tokens_seen": 15185200,
      "step": 26160
    },
    {
      "epoch": 3.89708072683944,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029185391749238305,
      "loss": 0.798,
      "num_input_tokens_seen": 15188368,
      "step": 26165
    },
    {
      "epoch": 3.8978254393803993,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.0291847578717864,
      "loss": 0.8101,
      "num_input_tokens_seen": 15191152,
      "step": 26170
    },
    {
      "epoch": 3.8985701519213585,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.029184123754697425,
      "loss": 0.807,
      "num_input_tokens_seen": 15194096,
      "step": 26175
    },
    {
      "epoch": 3.8993148644623177,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029183489397982095,
      "loss": 0.8164,
      "num_input_tokens_seen": 15196976,
      "step": 26180
    },
    {
      "epoch": 3.900059577003277,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029182854801651132,
      "loss": 0.8037,
      "num_input_tokens_seen": 15199536,
      "step": 26185
    },
    {
      "epoch": 3.900804289544236,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.029182219965715246,
      "loss": 0.7937,
      "num_input_tokens_seen": 15202288,
      "step": 26190
    },
    {
      "epoch": 3.9015490020851953,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02918158489018517,
      "loss": 0.7961,
      "num_input_tokens_seen": 15205424,
      "step": 26195
    },
    {
      "epoch": 3.902293714626154,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.02918094957507163,
      "loss": 0.8142,
      "num_input_tokens_seen": 15208272,
      "step": 26200
    },
    {
      "epoch": 3.9030384271671137,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.02918031402038536,
      "loss": 0.8089,
      "num_input_tokens_seen": 15211120,
      "step": 26205
    },
    {
      "epoch": 3.9037831397080724,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.029179678226137096,
      "loss": 0.7968,
      "num_input_tokens_seen": 15213936,
      "step": 26210
    },
    {
      "epoch": 3.904527852249032,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.029179042192337582,
      "loss": 0.7959,
      "num_input_tokens_seen": 15216560,
      "step": 26215
    },
    {
      "epoch": 3.905272564789991,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.02917840591899756,
      "loss": 0.8108,
      "num_input_tokens_seen": 15219472,
      "step": 26220
    },
    {
      "epoch": 3.9060172773309505,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.02917776940612778,
      "loss": 0.8211,
      "num_input_tokens_seen": 15222832,
      "step": 26225
    },
    {
      "epoch": 3.9067619898719093,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029177132653738995,
      "loss": 0.7988,
      "num_input_tokens_seen": 15225808,
      "step": 26230
    },
    {
      "epoch": 3.907506702412869,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.029176495661841967,
      "loss": 0.8016,
      "num_input_tokens_seen": 15228688,
      "step": 26235
    },
    {
      "epoch": 3.9082514149538277,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029175858430447447,
      "loss": 0.8128,
      "num_input_tokens_seen": 15231888,
      "step": 26240
    },
    {
      "epoch": 3.908996127494787,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029175220959566213,
      "loss": 0.7973,
      "num_input_tokens_seen": 15234960,
      "step": 26245
    },
    {
      "epoch": 3.909740840035746,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029174583249209022,
      "loss": 0.7831,
      "num_input_tokens_seen": 15237744,
      "step": 26250
    },
    {
      "epoch": 3.9104855525767053,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.029173945299386658,
      "loss": 0.7847,
      "num_input_tokens_seen": 15240624,
      "step": 26255
    },
    {
      "epoch": 3.9112302651176645,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.029173307110109892,
      "loss": 0.7968,
      "num_input_tokens_seen": 15243536,
      "step": 26260
    },
    {
      "epoch": 3.9119749776586237,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029172668681389506,
      "loss": 0.7816,
      "num_input_tokens_seen": 15246640,
      "step": 26265
    },
    {
      "epoch": 3.912719690199583,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02917203001323629,
      "loss": 0.8215,
      "num_input_tokens_seen": 15249456,
      "step": 26270
    },
    {
      "epoch": 3.913464402740542,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.029171391105661026,
      "loss": 0.7878,
      "num_input_tokens_seen": 15252592,
      "step": 26275
    },
    {
      "epoch": 3.9142091152815013,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.029170751958674516,
      "loss": 0.8124,
      "num_input_tokens_seen": 15255728,
      "step": 26280
    },
    {
      "epoch": 3.9149538278224605,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029170112572287556,
      "loss": 0.7908,
      "num_input_tokens_seen": 15258384,
      "step": 26285
    },
    {
      "epoch": 3.9156985403634197,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.029169472946510944,
      "loss": 0.8142,
      "num_input_tokens_seen": 15261104,
      "step": 26290
    },
    {
      "epoch": 3.916443252904379,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029168833081355486,
      "loss": 0.8156,
      "num_input_tokens_seen": 15263824,
      "step": 26295
    },
    {
      "epoch": 3.917187965445338,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.029168192976832,
      "loss": 0.7999,
      "num_input_tokens_seen": 15266640,
      "step": 26300
    },
    {
      "epoch": 3.9179326779862973,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029167552632951285,
      "loss": 0.7907,
      "num_input_tokens_seen": 15269552,
      "step": 26305
    },
    {
      "epoch": 3.9186773905272565,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02916691204972417,
      "loss": 0.8236,
      "num_input_tokens_seen": 15272560,
      "step": 26310
    },
    {
      "epoch": 3.9194221030682157,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029166271227161474,
      "loss": 0.7908,
      "num_input_tokens_seen": 15275312,
      "step": 26315
    },
    {
      "epoch": 3.920166815609175,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.02916563016527403,
      "loss": 0.8214,
      "num_input_tokens_seen": 15278256,
      "step": 26320
    },
    {
      "epoch": 3.920911528150134,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.029164988864072657,
      "loss": 0.7996,
      "num_input_tokens_seen": 15280944,
      "step": 26325
    },
    {
      "epoch": 3.9216562406910933,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02916434732356819,
      "loss": 0.8003,
      "num_input_tokens_seen": 15283568,
      "step": 26330
    },
    {
      "epoch": 3.9224009532320525,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.029163705543771477,
      "loss": 0.827,
      "num_input_tokens_seen": 15286320,
      "step": 26335
    },
    {
      "epoch": 3.9231456657730117,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02916306352469335,
      "loss": 0.8109,
      "num_input_tokens_seen": 15288976,
      "step": 26340
    },
    {
      "epoch": 3.923890378313971,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.029162421266344662,
      "loss": 0.8028,
      "num_input_tokens_seen": 15291728,
      "step": 26345
    },
    {
      "epoch": 3.92463509085493,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02916177876873626,
      "loss": 0.8116,
      "num_input_tokens_seen": 15294416,
      "step": 26350
    },
    {
      "epoch": 3.9253798033958893,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029161136031879,
      "loss": 0.8114,
      "num_input_tokens_seen": 15297264,
      "step": 26355
    },
    {
      "epoch": 3.9261245159368485,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.029160493055783738,
      "loss": 0.7871,
      "num_input_tokens_seen": 15300368,
      "step": 26360
    },
    {
      "epoch": 3.9268692284778077,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02915984984046134,
      "loss": 0.8143,
      "num_input_tokens_seen": 15302960,
      "step": 26365
    },
    {
      "epoch": 3.927613941018767,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02915920638592267,
      "loss": 0.802,
      "num_input_tokens_seen": 15305648,
      "step": 26370
    },
    {
      "epoch": 3.9283586535597257,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.029158562692178598,
      "loss": 0.8102,
      "num_input_tokens_seen": 15308624,
      "step": 26375
    },
    {
      "epoch": 3.9291033661006853,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02915791875924,
      "loss": 0.818,
      "num_input_tokens_seen": 15311440,
      "step": 26380
    },
    {
      "epoch": 3.929848078641644,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.029157274587117747,
      "loss": 0.8017,
      "num_input_tokens_seen": 15314224,
      "step": 26385
    },
    {
      "epoch": 3.9305927911826037,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029156630175822738,
      "loss": 0.812,
      "num_input_tokens_seen": 15317296,
      "step": 26390
    },
    {
      "epoch": 3.9313375037235625,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.029155985525365847,
      "loss": 0.8199,
      "num_input_tokens_seen": 15320080,
      "step": 26395
    },
    {
      "epoch": 3.932082216264522,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02915534063575797,
      "loss": 0.7971,
      "num_input_tokens_seen": 15322768,
      "step": 26400
    },
    {
      "epoch": 3.932826928805481,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029154695507009998,
      "loss": 0.8171,
      "num_input_tokens_seen": 15325776,
      "step": 26405
    },
    {
      "epoch": 3.9335716413464406,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02915405013913283,
      "loss": 0.8094,
      "num_input_tokens_seen": 15328720,
      "step": 26410
    },
    {
      "epoch": 3.9343163538873993,
      "grad_norm": 0.021484375,
      "learning_rate": 0.029153404532137373,
      "loss": 0.8035,
      "num_input_tokens_seen": 15331664,
      "step": 26415
    },
    {
      "epoch": 3.9350610664283585,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.02915275868603453,
      "loss": 0.798,
      "num_input_tokens_seen": 15334928,
      "step": 26420
    },
    {
      "epoch": 3.9358057789693177,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.029152112600835216,
      "loss": 0.8124,
      "num_input_tokens_seen": 15337680,
      "step": 26425
    },
    {
      "epoch": 3.936550491510277,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02915146627655034,
      "loss": 0.8035,
      "num_input_tokens_seen": 15340592,
      "step": 26430
    },
    {
      "epoch": 3.937295204051236,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029150819713190824,
      "loss": 0.8167,
      "num_input_tokens_seen": 15343920,
      "step": 26435
    },
    {
      "epoch": 3.9380399165921953,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.029150172910767592,
      "loss": 0.8042,
      "num_input_tokens_seen": 15346544,
      "step": 26440
    },
    {
      "epoch": 3.9387846291331545,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.029149525869291572,
      "loss": 0.8028,
      "num_input_tokens_seen": 15349520,
      "step": 26445
    },
    {
      "epoch": 3.9395293416741137,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.029148878588773693,
      "loss": 0.8003,
      "num_input_tokens_seen": 15352496,
      "step": 26450
    },
    {
      "epoch": 3.940274054215073,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.029148231069224886,
      "loss": 0.8091,
      "num_input_tokens_seen": 15355088,
      "step": 26455
    },
    {
      "epoch": 3.941018766756032,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.0291475833106561,
      "loss": 0.8003,
      "num_input_tokens_seen": 15357936,
      "step": 26460
    },
    {
      "epoch": 3.9417634792969913,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.029146935313078274,
      "loss": 0.8039,
      "num_input_tokens_seen": 15360592,
      "step": 26465
    },
    {
      "epoch": 3.9425081918379505,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02914628707650235,
      "loss": 0.8047,
      "num_input_tokens_seen": 15363664,
      "step": 26470
    },
    {
      "epoch": 3.9432529043789097,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.029145638600939282,
      "loss": 0.8071,
      "num_input_tokens_seen": 15366608,
      "step": 26475
    },
    {
      "epoch": 3.943997616919869,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02914498988640003,
      "loss": 0.796,
      "num_input_tokens_seen": 15369744,
      "step": 26480
    },
    {
      "epoch": 3.944742329460828,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.029144340932895547,
      "loss": 0.8091,
      "num_input_tokens_seen": 15372624,
      "step": 26485
    },
    {
      "epoch": 3.9454870420017873,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029143691740436805,
      "loss": 0.8066,
      "num_input_tokens_seen": 15375440,
      "step": 26490
    },
    {
      "epoch": 3.9462317545427466,
      "grad_norm": 0.0123291015625,
      "learning_rate": 0.029143042309034764,
      "loss": 0.8016,
      "num_input_tokens_seen": 15378128,
      "step": 26495
    },
    {
      "epoch": 3.9469764670837058,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.0291423926387004,
      "loss": 0.8068,
      "num_input_tokens_seen": 15380848,
      "step": 26500
    },
    {
      "epoch": 3.947721179624665,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02914174272944468,
      "loss": 0.8068,
      "num_input_tokens_seen": 15383600,
      "step": 26505
    },
    {
      "epoch": 3.948465892165624,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.029141092581278595,
      "loss": 0.8038,
      "num_input_tokens_seen": 15386288,
      "step": 26510
    },
    {
      "epoch": 3.9492106047065834,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02914044219421312,
      "loss": 0.8074,
      "num_input_tokens_seen": 15389040,
      "step": 26515
    },
    {
      "epoch": 3.9499553172475426,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.029139791568259247,
      "loss": 0.8117,
      "num_input_tokens_seen": 15392080,
      "step": 26520
    },
    {
      "epoch": 3.9507000297885018,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.029139140703427963,
      "loss": 0.8063,
      "num_input_tokens_seen": 15394768,
      "step": 26525
    },
    {
      "epoch": 3.951444742329461,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029138489599730272,
      "loss": 0.8007,
      "num_input_tokens_seen": 15397648,
      "step": 26530
    },
    {
      "epoch": 3.95218945487042,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029137838257177164,
      "loss": 0.7971,
      "num_input_tokens_seen": 15400400,
      "step": 26535
    },
    {
      "epoch": 3.9529341674113794,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.029137186675779653,
      "loss": 0.806,
      "num_input_tokens_seen": 15403152,
      "step": 26540
    },
    {
      "epoch": 3.9536788799523386,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02913653485554874,
      "loss": 0.808,
      "num_input_tokens_seen": 15405936,
      "step": 26545
    },
    {
      "epoch": 3.9544235924932973,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02913588279649544,
      "loss": 0.7889,
      "num_input_tokens_seen": 15408976,
      "step": 26550
    },
    {
      "epoch": 3.955168305034257,
      "grad_norm": 0.01068115234375,
      "learning_rate": 0.029135230498630763,
      "loss": 0.8115,
      "num_input_tokens_seen": 15411824,
      "step": 26555
    },
    {
      "epoch": 3.9559130175752157,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.029134577961965735,
      "loss": 0.7862,
      "num_input_tokens_seen": 15414576,
      "step": 26560
    },
    {
      "epoch": 3.9566577301161754,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.029133925186511377,
      "loss": 0.7835,
      "num_input_tokens_seen": 15417456,
      "step": 26565
    },
    {
      "epoch": 3.957402442657134,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02913327217227872,
      "loss": 0.8113,
      "num_input_tokens_seen": 15420528,
      "step": 26570
    },
    {
      "epoch": 3.958147155198094,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.029132618919278794,
      "loss": 0.7923,
      "num_input_tokens_seen": 15423536,
      "step": 26575
    },
    {
      "epoch": 3.9588918677390526,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029131965427522635,
      "loss": 0.8176,
      "num_input_tokens_seen": 15426160,
      "step": 26580
    },
    {
      "epoch": 3.9596365802800118,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02913131169702128,
      "loss": 0.8201,
      "num_input_tokens_seen": 15429296,
      "step": 26585
    },
    {
      "epoch": 3.960381292820971,
      "grad_norm": 0.03125,
      "learning_rate": 0.029130657727785784,
      "loss": 0.846,
      "num_input_tokens_seen": 15432112,
      "step": 26590
    },
    {
      "epoch": 3.96112600536193,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029130003519827183,
      "loss": 0.8132,
      "num_input_tokens_seen": 15435088,
      "step": 26595
    },
    {
      "epoch": 3.9618707179028894,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02912934907315653,
      "loss": 0.8014,
      "num_input_tokens_seen": 15437840,
      "step": 26600
    },
    {
      "epoch": 3.9626154304438486,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.029128694387784888,
      "loss": 0.8133,
      "num_input_tokens_seen": 15440720,
      "step": 26605
    },
    {
      "epoch": 3.9633601429848078,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.029128039463723315,
      "loss": 0.8061,
      "num_input_tokens_seen": 15443600,
      "step": 26610
    },
    {
      "epoch": 3.964104855525767,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02912738430098287,
      "loss": 0.7979,
      "num_input_tokens_seen": 15446448,
      "step": 26615
    },
    {
      "epoch": 3.964849568066726,
      "grad_norm": 0.01190185546875,
      "learning_rate": 0.029126728899574635,
      "loss": 0.7974,
      "num_input_tokens_seen": 15449072,
      "step": 26620
    },
    {
      "epoch": 3.9655942806076854,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.029126073259509663,
      "loss": 0.8061,
      "num_input_tokens_seen": 15451856,
      "step": 26625
    },
    {
      "epoch": 3.9663389931486446,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.029125417380799046,
      "loss": 0.8002,
      "num_input_tokens_seen": 15454896,
      "step": 26630
    },
    {
      "epoch": 3.967083705689604,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029124761263453855,
      "loss": 0.8122,
      "num_input_tokens_seen": 15457808,
      "step": 26635
    },
    {
      "epoch": 3.967828418230563,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.029124104907485182,
      "loss": 0.8024,
      "num_input_tokens_seen": 15460624,
      "step": 26640
    },
    {
      "epoch": 3.968573130771522,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02912344831290411,
      "loss": 0.8049,
      "num_input_tokens_seen": 15463472,
      "step": 26645
    },
    {
      "epoch": 3.9693178433124814,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02912279147972173,
      "loss": 0.8151,
      "num_input_tokens_seen": 15466320,
      "step": 26650
    },
    {
      "epoch": 3.9700625558534406,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029122134407949148,
      "loss": 0.8111,
      "num_input_tokens_seen": 15469168,
      "step": 26655
    },
    {
      "epoch": 3.9708072683944,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.029121477097597454,
      "loss": 0.8035,
      "num_input_tokens_seen": 15471888,
      "step": 26660
    },
    {
      "epoch": 3.971551980935359,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02912081954867776,
      "loss": 0.7973,
      "num_input_tokens_seen": 15475088,
      "step": 26665
    },
    {
      "epoch": 3.972296693476318,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02912016176120117,
      "loss": 0.792,
      "num_input_tokens_seen": 15478064,
      "step": 26670
    },
    {
      "epoch": 3.9730414060172774,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029119503735178798,
      "loss": 0.7968,
      "num_input_tokens_seen": 15480912,
      "step": 26675
    },
    {
      "epoch": 3.9737861185582366,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029118845470621758,
      "loss": 0.8213,
      "num_input_tokens_seen": 15483952,
      "step": 26680
    },
    {
      "epoch": 3.974530831099196,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02911818696754118,
      "loss": 0.7974,
      "num_input_tokens_seen": 15487088,
      "step": 26685
    },
    {
      "epoch": 3.975275543640155,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.029117528225948174,
      "loss": 0.7862,
      "num_input_tokens_seen": 15489872,
      "step": 26690
    },
    {
      "epoch": 3.976020256181114,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.029116869245853882,
      "loss": 0.7967,
      "num_input_tokens_seen": 15492784,
      "step": 26695
    },
    {
      "epoch": 3.9767649687220734,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.029116210027269433,
      "loss": 0.8222,
      "num_input_tokens_seen": 15495440,
      "step": 26700
    },
    {
      "epoch": 3.9775096812630326,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029115550570205963,
      "loss": 0.8123,
      "num_input_tokens_seen": 15498480,
      "step": 26705
    },
    {
      "epoch": 3.978254393803992,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029114890874674608,
      "loss": 0.8079,
      "num_input_tokens_seen": 15501424,
      "step": 26710
    },
    {
      "epoch": 3.9789991063449506,
      "grad_norm": 0.017578125,
      "learning_rate": 0.029114230940686516,
      "loss": 0.8048,
      "num_input_tokens_seen": 15504304,
      "step": 26715
    },
    {
      "epoch": 3.9797438188859102,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.029113570768252845,
      "loss": 0.808,
      "num_input_tokens_seen": 15507440,
      "step": 26720
    },
    {
      "epoch": 3.980488531426869,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029112910357384735,
      "loss": 0.7859,
      "num_input_tokens_seen": 15510256,
      "step": 26725
    },
    {
      "epoch": 3.9812332439678286,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.029112249708093355,
      "loss": 0.8015,
      "num_input_tokens_seen": 15513168,
      "step": 26730
    },
    {
      "epoch": 3.9819779565087874,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.02911158882038985,
      "loss": 0.8095,
      "num_input_tokens_seen": 15516368,
      "step": 26735
    },
    {
      "epoch": 3.982722669049747,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029110927694285398,
      "loss": 0.7918,
      "num_input_tokens_seen": 15518928,
      "step": 26740
    },
    {
      "epoch": 3.983467381590706,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.029110266329791166,
      "loss": 0.8148,
      "num_input_tokens_seen": 15521808,
      "step": 26745
    },
    {
      "epoch": 3.9842120941316654,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02910960472691832,
      "loss": 0.7949,
      "num_input_tokens_seen": 15524624,
      "step": 26750
    },
    {
      "epoch": 3.984956806672624,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029108942885678047,
      "loss": 0.8009,
      "num_input_tokens_seen": 15527184,
      "step": 26755
    },
    {
      "epoch": 3.9857015192135834,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.029108280806081525,
      "loss": 0.827,
      "num_input_tokens_seen": 15530064,
      "step": 26760
    },
    {
      "epoch": 3.9864462317545426,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.029107618488139933,
      "loss": 0.7846,
      "num_input_tokens_seen": 15533168,
      "step": 26765
    },
    {
      "epoch": 3.987190944295502,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02910695593186447,
      "loss": 0.7893,
      "num_input_tokens_seen": 15536240,
      "step": 26770
    },
    {
      "epoch": 3.987935656836461,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.02910629313726632,
      "loss": 0.8029,
      "num_input_tokens_seen": 15539120,
      "step": 26775
    },
    {
      "epoch": 3.98868036937742,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02910563010435669,
      "loss": 0.8082,
      "num_input_tokens_seen": 15542096,
      "step": 26780
    },
    {
      "epoch": 3.9894250819183794,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02910496683314677,
      "loss": 0.8049,
      "num_input_tokens_seen": 15544976,
      "step": 26785
    },
    {
      "epoch": 3.9901697944593386,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.029104303323647773,
      "loss": 0.7858,
      "num_input_tokens_seen": 15547888,
      "step": 26790
    },
    {
      "epoch": 3.990914507000298,
      "grad_norm": 0.017578125,
      "learning_rate": 0.029103639575870907,
      "loss": 0.7851,
      "num_input_tokens_seen": 15550960,
      "step": 26795
    },
    {
      "epoch": 3.991659219541257,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02910297558982738,
      "loss": 0.8027,
      "num_input_tokens_seen": 15553552,
      "step": 26800
    },
    {
      "epoch": 3.9924039320822162,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.029102311365528422,
      "loss": 0.8132,
      "num_input_tokens_seen": 15556272,
      "step": 26805
    },
    {
      "epoch": 3.9931486446231754,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.029101646902985244,
      "loss": 0.798,
      "num_input_tokens_seen": 15559408,
      "step": 26810
    },
    {
      "epoch": 3.9938933571641346,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.02910098220220907,
      "loss": 0.7947,
      "num_input_tokens_seen": 15562352,
      "step": 26815
    },
    {
      "epoch": 3.994638069705094,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.029100317263211137,
      "loss": 0.8443,
      "num_input_tokens_seen": 15565360,
      "step": 26820
    },
    {
      "epoch": 3.995382782246053,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.029099652086002674,
      "loss": 0.8007,
      "num_input_tokens_seen": 15568176,
      "step": 26825
    },
    {
      "epoch": 3.9961274947870122,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02909898667059492,
      "loss": 0.8155,
      "num_input_tokens_seen": 15571152,
      "step": 26830
    },
    {
      "epoch": 3.9968722073279714,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02909832101699911,
      "loss": 0.8175,
      "num_input_tokens_seen": 15574032,
      "step": 26835
    },
    {
      "epoch": 3.9976169198689306,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.029097655125226503,
      "loss": 0.81,
      "num_input_tokens_seen": 15577072,
      "step": 26840
    },
    {
      "epoch": 3.99836163240989,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.029096988995288335,
      "loss": 0.7796,
      "num_input_tokens_seen": 15579824,
      "step": 26845
    },
    {
      "epoch": 3.999106344950849,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.029096322627195872,
      "loss": 0.804,
      "num_input_tokens_seen": 15582480,
      "step": 26850
    },
    {
      "epoch": 3.9998510574918082,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02909565602096036,
      "loss": 0.7935,
      "num_input_tokens_seen": 15585744,
      "step": 26855
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.8017039895057678,
      "eval_runtime": 70.7018,
      "eval_samples_per_second": 42.205,
      "eval_steps_per_second": 10.551,
      "num_input_tokens_seen": 15585872,
      "step": 26856
    },
    {
      "epoch": 4.000595770032767,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.029094989176593068,
      "loss": 0.8007,
      "num_input_tokens_seen": 15588400,
      "step": 26860
    },
    {
      "epoch": 4.001340482573727,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.029094322094105257,
      "loss": 0.8109,
      "num_input_tokens_seen": 15591440,
      "step": 26865
    },
    {
      "epoch": 4.002085195114685,
      "grad_norm": 0.029296875,
      "learning_rate": 0.0290936547735082,
      "loss": 0.7808,
      "num_input_tokens_seen": 15594608,
      "step": 26870
    },
    {
      "epoch": 4.002829907655645,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.029092987214813168,
      "loss": 0.7978,
      "num_input_tokens_seen": 15597488,
      "step": 26875
    },
    {
      "epoch": 4.003574620196604,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.029092319418031445,
      "loss": 0.8012,
      "num_input_tokens_seen": 15600784,
      "step": 26880
    },
    {
      "epoch": 4.0043193327375635,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.029091651383174304,
      "loss": 0.8057,
      "num_input_tokens_seen": 15603664,
      "step": 26885
    },
    {
      "epoch": 4.005064045278522,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.029090983110253037,
      "loss": 0.8137,
      "num_input_tokens_seen": 15606448,
      "step": 26890
    },
    {
      "epoch": 4.005808757819482,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029090314599278935,
      "loss": 0.803,
      "num_input_tokens_seen": 15609232,
      "step": 26895
    },
    {
      "epoch": 4.006553470360441,
      "grad_norm": 0.017578125,
      "learning_rate": 0.029089645850263284,
      "loss": 0.7899,
      "num_input_tokens_seen": 15612080,
      "step": 26900
    },
    {
      "epoch": 4.0072981829014,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029088976863217385,
      "loss": 0.7868,
      "num_input_tokens_seen": 15614896,
      "step": 26905
    },
    {
      "epoch": 4.008042895442359,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.029088307638152545,
      "loss": 0.8008,
      "num_input_tokens_seen": 15617584,
      "step": 26910
    },
    {
      "epoch": 4.008787607983319,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.029087638175080067,
      "loss": 0.7929,
      "num_input_tokens_seen": 15620592,
      "step": 26915
    },
    {
      "epoch": 4.009532320524277,
      "grad_norm": 0.01953125,
      "learning_rate": 0.029086968474011255,
      "loss": 0.802,
      "num_input_tokens_seen": 15623344,
      "step": 26920
    },
    {
      "epoch": 4.010277033065237,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.029086298534957436,
      "loss": 0.7923,
      "num_input_tokens_seen": 15626320,
      "step": 26925
    },
    {
      "epoch": 4.011021745606196,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02908562835792991,
      "loss": 0.8072,
      "num_input_tokens_seen": 15629136,
      "step": 26930
    },
    {
      "epoch": 4.0117664581471555,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.029084957942940016,
      "loss": 0.7827,
      "num_input_tokens_seen": 15631952,
      "step": 26935
    },
    {
      "epoch": 4.012511170688114,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02908428728999907,
      "loss": 0.821,
      "num_input_tokens_seen": 15634896,
      "step": 26940
    },
    {
      "epoch": 4.013255883229074,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.029083616399118408,
      "loss": 0.8221,
      "num_input_tokens_seen": 15637712,
      "step": 26945
    },
    {
      "epoch": 4.014000595770033,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.029082945270309357,
      "loss": 0.8,
      "num_input_tokens_seen": 15640720,
      "step": 26950
    },
    {
      "epoch": 4.014745308310992,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.029082273903583263,
      "loss": 0.7864,
      "num_input_tokens_seen": 15643696,
      "step": 26955
    },
    {
      "epoch": 4.015490020851951,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.02908160229895146,
      "loss": 0.7711,
      "num_input_tokens_seen": 15646896,
      "step": 26960
    },
    {
      "epoch": 4.016234733392911,
      "grad_norm": 0.0234375,
      "learning_rate": 0.029080930456425298,
      "loss": 0.8371,
      "num_input_tokens_seen": 15650064,
      "step": 26965
    },
    {
      "epoch": 4.0169794459338695,
      "grad_norm": 0.010498046875,
      "learning_rate": 0.02908025837601613,
      "loss": 0.8058,
      "num_input_tokens_seen": 15652880,
      "step": 26970
    },
    {
      "epoch": 4.017724158474829,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.029079586057735304,
      "loss": 0.8299,
      "num_input_tokens_seen": 15655824,
      "step": 26975
    },
    {
      "epoch": 4.018468871015788,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02907891350159418,
      "loss": 0.7999,
      "num_input_tokens_seen": 15658384,
      "step": 26980
    },
    {
      "epoch": 4.0192135835567475,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02907824070760413,
      "loss": 0.8042,
      "num_input_tokens_seen": 15661232,
      "step": 26985
    },
    {
      "epoch": 4.019958296097706,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029077567675776506,
      "loss": 0.78,
      "num_input_tokens_seen": 15664304,
      "step": 26990
    },
    {
      "epoch": 4.020703008638666,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029076894406122687,
      "loss": 0.7979,
      "num_input_tokens_seen": 15667344,
      "step": 26995
    },
    {
      "epoch": 4.021447721179625,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02907622089865404,
      "loss": 0.8253,
      "num_input_tokens_seen": 15670128,
      "step": 27000
    },
    {
      "epoch": 4.022192433720583,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02907554715338195,
      "loss": 0.808,
      "num_input_tokens_seen": 15673328,
      "step": 27005
    },
    {
      "epoch": 4.022937146261543,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.029074873170317792,
      "loss": 0.7979,
      "num_input_tokens_seen": 15676304,
      "step": 27010
    },
    {
      "epoch": 4.023681858802502,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029074198949472965,
      "loss": 0.8111,
      "num_input_tokens_seen": 15679312,
      "step": 27015
    },
    {
      "epoch": 4.0244265713434615,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.029073524490858847,
      "loss": 0.7918,
      "num_input_tokens_seen": 15682320,
      "step": 27020
    },
    {
      "epoch": 4.02517128388442,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029072849794486835,
      "loss": 0.7956,
      "num_input_tokens_seen": 15685328,
      "step": 27025
    },
    {
      "epoch": 4.02591599642538,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029072174860368324,
      "loss": 0.8265,
      "num_input_tokens_seen": 15688144,
      "step": 27030
    },
    {
      "epoch": 4.026660708966339,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029071499688514732,
      "loss": 0.8052,
      "num_input_tokens_seen": 15690928,
      "step": 27035
    },
    {
      "epoch": 4.027405421507298,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.029070824278937448,
      "loss": 0.7883,
      "num_input_tokens_seen": 15694032,
      "step": 27040
    },
    {
      "epoch": 4.028150134048257,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02907014863164789,
      "loss": 0.8075,
      "num_input_tokens_seen": 15697264,
      "step": 27045
    },
    {
      "epoch": 4.028894846589217,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.029069472746657467,
      "loss": 0.8002,
      "num_input_tokens_seen": 15700368,
      "step": 27050
    },
    {
      "epoch": 4.0296395591301755,
      "grad_norm": 0.017578125,
      "learning_rate": 0.02906879662397761,
      "loss": 0.8115,
      "num_input_tokens_seen": 15703536,
      "step": 27055
    },
    {
      "epoch": 4.030384271671135,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.029068120263619726,
      "loss": 0.8093,
      "num_input_tokens_seen": 15706512,
      "step": 27060
    },
    {
      "epoch": 4.031128984212094,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02906744366559525,
      "loss": 0.8117,
      "num_input_tokens_seen": 15709232,
      "step": 27065
    },
    {
      "epoch": 4.0318736967530535,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.029066766829915616,
      "loss": 0.8327,
      "num_input_tokens_seen": 15712400,
      "step": 27070
    },
    {
      "epoch": 4.032618409294012,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029066089756592247,
      "loss": 0.8208,
      "num_input_tokens_seen": 15715024,
      "step": 27075
    },
    {
      "epoch": 4.033363121834972,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.02906541244563659,
      "loss": 0.8099,
      "num_input_tokens_seen": 15717584,
      "step": 27080
    },
    {
      "epoch": 4.034107834375931,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02906473489706008,
      "loss": 0.8003,
      "num_input_tokens_seen": 15720432,
      "step": 27085
    },
    {
      "epoch": 4.03485254691689,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02906405711087418,
      "loss": 0.8063,
      "num_input_tokens_seen": 15723888,
      "step": 27090
    },
    {
      "epoch": 4.035597259457849,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.02906337908709032,
      "loss": 0.782,
      "num_input_tokens_seen": 15726800,
      "step": 27095
    },
    {
      "epoch": 4.036341971998809,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.029062700825719965,
      "loss": 0.805,
      "num_input_tokens_seen": 15729744,
      "step": 27100
    },
    {
      "epoch": 4.0370866845397675,
      "grad_norm": 0.01123046875,
      "learning_rate": 0.029062022326774576,
      "loss": 0.8037,
      "num_input_tokens_seen": 15732944,
      "step": 27105
    },
    {
      "epoch": 4.037831397080727,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.029061343590265607,
      "loss": 0.8265,
      "num_input_tokens_seen": 15736016,
      "step": 27110
    },
    {
      "epoch": 4.038576109621686,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029060664616204532,
      "loss": 0.7917,
      "num_input_tokens_seen": 15738928,
      "step": 27115
    },
    {
      "epoch": 4.0393208221626455,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02905998540460282,
      "loss": 0.799,
      "num_input_tokens_seen": 15742032,
      "step": 27120
    },
    {
      "epoch": 4.040065534703604,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02905930595547194,
      "loss": 0.8,
      "num_input_tokens_seen": 15744880,
      "step": 27125
    },
    {
      "epoch": 4.040810247244564,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.029058626268823384,
      "loss": 0.8007,
      "num_input_tokens_seen": 15747856,
      "step": 27130
    },
    {
      "epoch": 4.041554959785523,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.029057946344668618,
      "loss": 0.8184,
      "num_input_tokens_seen": 15750832,
      "step": 27135
    },
    {
      "epoch": 4.042299672326482,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.029057266183019136,
      "loss": 0.8109,
      "num_input_tokens_seen": 15753552,
      "step": 27140
    },
    {
      "epoch": 4.043044384867441,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.029056585783886434,
      "loss": 0.8195,
      "num_input_tokens_seen": 15756496,
      "step": 27145
    },
    {
      "epoch": 4.043789097408401,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.029055905147281996,
      "loss": 0.804,
      "num_input_tokens_seen": 15759472,
      "step": 27150
    },
    {
      "epoch": 4.0445338099493595,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02905522427321733,
      "loss": 0.7973,
      "num_input_tokens_seen": 15762480,
      "step": 27155
    },
    {
      "epoch": 4.045278522490319,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.029054543161703937,
      "loss": 0.8023,
      "num_input_tokens_seen": 15765232,
      "step": 27160
    },
    {
      "epoch": 4.046023235031278,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.029053861812753316,
      "loss": 0.8006,
      "num_input_tokens_seen": 15768176,
      "step": 27165
    },
    {
      "epoch": 4.046767947572237,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029053180226376984,
      "loss": 0.7978,
      "num_input_tokens_seen": 15770768,
      "step": 27170
    },
    {
      "epoch": 4.047512660113196,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02905249840258646,
      "loss": 0.7997,
      "num_input_tokens_seen": 15773424,
      "step": 27175
    },
    {
      "epoch": 4.048257372654155,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.02905181634139326,
      "loss": 0.808,
      "num_input_tokens_seen": 15776016,
      "step": 27180
    },
    {
      "epoch": 4.049002085195115,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.029051134042808895,
      "loss": 0.7974,
      "num_input_tokens_seen": 15779088,
      "step": 27185
    },
    {
      "epoch": 4.0497467977360735,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.029050451506844908,
      "loss": 0.8156,
      "num_input_tokens_seen": 15781936,
      "step": 27190
    },
    {
      "epoch": 4.050491510277033,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.029049768733512824,
      "loss": 0.8184,
      "num_input_tokens_seen": 15785392,
      "step": 27195
    },
    {
      "epoch": 4.051236222817992,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02904908572282417,
      "loss": 0.7962,
      "num_input_tokens_seen": 15788144,
      "step": 27200
    },
    {
      "epoch": 4.0519809353589515,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.029048402474790496,
      "loss": 0.8037,
      "num_input_tokens_seen": 15790864,
      "step": 27205
    },
    {
      "epoch": 4.05272564789991,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.029047718989423342,
      "loss": 0.7967,
      "num_input_tokens_seen": 15794256,
      "step": 27210
    },
    {
      "epoch": 4.05347036044087,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02904703526673425,
      "loss": 0.82,
      "num_input_tokens_seen": 15797424,
      "step": 27215
    },
    {
      "epoch": 4.054215072981829,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.029046351306734775,
      "loss": 0.8029,
      "num_input_tokens_seen": 15800112,
      "step": 27220
    },
    {
      "epoch": 4.054959785522788,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.02904566710943647,
      "loss": 0.805,
      "num_input_tokens_seen": 15803248,
      "step": 27225
    },
    {
      "epoch": 4.055704498063747,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029044982674850894,
      "loss": 0.8092,
      "num_input_tokens_seen": 15806224,
      "step": 27230
    },
    {
      "epoch": 4.056449210604707,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02904429800298961,
      "loss": 0.8067,
      "num_input_tokens_seen": 15809072,
      "step": 27235
    },
    {
      "epoch": 4.0571939231456655,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.029043613093864187,
      "loss": 0.7839,
      "num_input_tokens_seen": 15811920,
      "step": 27240
    },
    {
      "epoch": 4.057938635686625,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.029042927947486197,
      "loss": 0.7879,
      "num_input_tokens_seen": 15814608,
      "step": 27245
    },
    {
      "epoch": 4.058683348227584,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02904224256386721,
      "loss": 0.8325,
      "num_input_tokens_seen": 15817392,
      "step": 27250
    },
    {
      "epoch": 4.059428060768544,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.029041556943018805,
      "loss": 0.7868,
      "num_input_tokens_seen": 15820240,
      "step": 27255
    },
    {
      "epoch": 4.060172773309502,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02904087108495257,
      "loss": 0.8048,
      "num_input_tokens_seen": 15823024,
      "step": 27260
    },
    {
      "epoch": 4.060917485850462,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.02904018498968008,
      "loss": 0.7813,
      "num_input_tokens_seen": 15825936,
      "step": 27265
    },
    {
      "epoch": 4.061662198391421,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.029039498657212944,
      "loss": 0.7843,
      "num_input_tokens_seen": 15828752,
      "step": 27270
    },
    {
      "epoch": 4.06240691093238,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.029038812087562742,
      "loss": 0.7828,
      "num_input_tokens_seen": 15831568,
      "step": 27275
    },
    {
      "epoch": 4.063151623473339,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02903812528074108,
      "loss": 0.8044,
      "num_input_tokens_seen": 15834480,
      "step": 27280
    },
    {
      "epoch": 4.063896336014299,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02903743823675956,
      "loss": 0.777,
      "num_input_tokens_seen": 15837360,
      "step": 27285
    },
    {
      "epoch": 4.0646410485552575,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.029036750955629784,
      "loss": 0.8075,
      "num_input_tokens_seen": 15840720,
      "step": 27290
    },
    {
      "epoch": 4.065385761096217,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.029036063437363372,
      "loss": 0.8017,
      "num_input_tokens_seen": 15843568,
      "step": 27295
    },
    {
      "epoch": 4.066130473637176,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02903537568197193,
      "loss": 0.7769,
      "num_input_tokens_seen": 15846352,
      "step": 27300
    },
    {
      "epoch": 4.066875186178136,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.029034687689467084,
      "loss": 0.7623,
      "num_input_tokens_seen": 15849200,
      "step": 27305
    },
    {
      "epoch": 4.067619898719094,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029033999459860453,
      "loss": 0.8345,
      "num_input_tokens_seen": 15852304,
      "step": 27310
    },
    {
      "epoch": 4.068364611260054,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02903331099316366,
      "loss": 0.7902,
      "num_input_tokens_seen": 15855088,
      "step": 27315
    },
    {
      "epoch": 4.069109323801013,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029032622289388344,
      "loss": 0.8225,
      "num_input_tokens_seen": 15858288,
      "step": 27320
    },
    {
      "epoch": 4.069854036341972,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02903193334854614,
      "loss": 0.844,
      "num_input_tokens_seen": 15861264,
      "step": 27325
    },
    {
      "epoch": 4.070598748882931,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.02903124417064868,
      "loss": 0.7611,
      "num_input_tokens_seen": 15863888,
      "step": 27330
    },
    {
      "epoch": 4.071343461423891,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02903055475570761,
      "loss": 0.8353,
      "num_input_tokens_seen": 15866576,
      "step": 27335
    },
    {
      "epoch": 4.07208817396485,
      "grad_norm": 0.01025390625,
      "learning_rate": 0.02902986510373458,
      "loss": 0.7517,
      "num_input_tokens_seen": 15869680,
      "step": 27340
    },
    {
      "epoch": 4.072832886505808,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.029029175214741236,
      "loss": 0.8127,
      "num_input_tokens_seen": 15872656,
      "step": 27345
    },
    {
      "epoch": 4.073577599046768,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.029028485088739238,
      "loss": 0.8064,
      "num_input_tokens_seen": 15875600,
      "step": 27350
    },
    {
      "epoch": 4.074322311587727,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.029027794725740234,
      "loss": 0.7881,
      "num_input_tokens_seen": 15878416,
      "step": 27355
    },
    {
      "epoch": 4.075067024128686,
      "grad_norm": 0.017578125,
      "learning_rate": 0.029027104125755904,
      "loss": 0.8164,
      "num_input_tokens_seen": 15881328,
      "step": 27360
    },
    {
      "epoch": 4.075811736669645,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.029026413288797903,
      "loss": 0.8047,
      "num_input_tokens_seen": 15884368,
      "step": 27365
    },
    {
      "epoch": 4.076556449210605,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.029025722214877907,
      "loss": 0.7986,
      "num_input_tokens_seen": 15887344,
      "step": 27370
    },
    {
      "epoch": 4.0773011617515635,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029025030904007588,
      "loss": 0.8168,
      "num_input_tokens_seen": 15890096,
      "step": 27375
    },
    {
      "epoch": 4.078045874292523,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02902433935619863,
      "loss": 0.7772,
      "num_input_tokens_seen": 15892912,
      "step": 27380
    },
    {
      "epoch": 4.078790586833482,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.029023647571462707,
      "loss": 0.8023,
      "num_input_tokens_seen": 15895568,
      "step": 27385
    },
    {
      "epoch": 4.079535299374442,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.029022955549811512,
      "loss": 0.8043,
      "num_input_tokens_seen": 15898160,
      "step": 27390
    },
    {
      "epoch": 4.0802800119154,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.029022263291256737,
      "loss": 0.7824,
      "num_input_tokens_seen": 15901296,
      "step": 27395
    },
    {
      "epoch": 4.08102472445636,
      "grad_norm": 0.01171875,
      "learning_rate": 0.029021570795810078,
      "loss": 0.8036,
      "num_input_tokens_seen": 15903952,
      "step": 27400
    },
    {
      "epoch": 4.081769436997319,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.029020878063483226,
      "loss": 0.8138,
      "num_input_tokens_seen": 15907024,
      "step": 27405
    },
    {
      "epoch": 4.082514149538278,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02902018509428789,
      "loss": 0.8108,
      "num_input_tokens_seen": 15909744,
      "step": 27410
    },
    {
      "epoch": 4.083258862079237,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.02901949188823578,
      "loss": 0.8266,
      "num_input_tokens_seen": 15912688,
      "step": 27415
    },
    {
      "epoch": 4.084003574620197,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.029018798445338598,
      "loss": 0.7929,
      "num_input_tokens_seen": 15915856,
      "step": 27420
    },
    {
      "epoch": 4.084748287161156,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.02901810476560807,
      "loss": 0.7896,
      "num_input_tokens_seen": 15918864,
      "step": 27425
    },
    {
      "epoch": 4.085492999702115,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.029017410849055906,
      "loss": 0.8011,
      "num_input_tokens_seen": 15921552,
      "step": 27430
    },
    {
      "epoch": 4.086237712243074,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.029016716695693834,
      "loss": 0.8104,
      "num_input_tokens_seen": 15925008,
      "step": 27435
    },
    {
      "epoch": 4.086982424784034,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02901602230553358,
      "loss": 0.7872,
      "num_input_tokens_seen": 15927696,
      "step": 27440
    },
    {
      "epoch": 4.087727137324992,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02901532767858687,
      "loss": 0.7847,
      "num_input_tokens_seen": 15930576,
      "step": 27445
    },
    {
      "epoch": 4.088471849865952,
      "grad_norm": 0.011474609375,
      "learning_rate": 0.02901463281486545,
      "loss": 0.8046,
      "num_input_tokens_seen": 15933808,
      "step": 27450
    },
    {
      "epoch": 4.089216562406911,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029013937714381044,
      "loss": 0.8212,
      "num_input_tokens_seen": 15936624,
      "step": 27455
    },
    {
      "epoch": 4.08996127494787,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.02901324237714541,
      "loss": 0.8179,
      "num_input_tokens_seen": 15939344,
      "step": 27460
    },
    {
      "epoch": 4.090705987488829,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02901254680317029,
      "loss": 0.7939,
      "num_input_tokens_seen": 15941968,
      "step": 27465
    },
    {
      "epoch": 4.091450700029789,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02901185099246743,
      "loss": 0.8062,
      "num_input_tokens_seen": 15944720,
      "step": 27470
    },
    {
      "epoch": 4.092195412570748,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.029011154945048585,
      "loss": 0.8414,
      "num_input_tokens_seen": 15947600,
      "step": 27475
    },
    {
      "epoch": 4.092940125111707,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029010458660925522,
      "loss": 0.8055,
      "num_input_tokens_seen": 15950416,
      "step": 27480
    },
    {
      "epoch": 4.093684837652666,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02900976214011,
      "loss": 0.788,
      "num_input_tokens_seen": 15953168,
      "step": 27485
    },
    {
      "epoch": 4.094429550193626,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029009065382613785,
      "loss": 0.7981,
      "num_input_tokens_seen": 15956112,
      "step": 27490
    },
    {
      "epoch": 4.095174262734584,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.029008368388448648,
      "loss": 0.7881,
      "num_input_tokens_seen": 15958928,
      "step": 27495
    },
    {
      "epoch": 4.095918975275544,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.029007671157626362,
      "loss": 0.79,
      "num_input_tokens_seen": 15961808,
      "step": 27500
    },
    {
      "epoch": 4.096663687816503,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.029006973690158713,
      "loss": 0.8086,
      "num_input_tokens_seen": 15964624,
      "step": 27505
    },
    {
      "epoch": 4.0974084003574625,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.029006275986057477,
      "loss": 0.792,
      "num_input_tokens_seen": 15967632,
      "step": 27510
    },
    {
      "epoch": 4.098153112898421,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.029005578045334445,
      "loss": 0.7942,
      "num_input_tokens_seen": 15970352,
      "step": 27515
    },
    {
      "epoch": 4.09889782543938,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.029004879868001408,
      "loss": 0.7988,
      "num_input_tokens_seen": 15973552,
      "step": 27520
    },
    {
      "epoch": 4.09964253798034,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02900418145407015,
      "loss": 0.7938,
      "num_input_tokens_seen": 15976560,
      "step": 27525
    },
    {
      "epoch": 4.100387250521298,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.029003482803552492,
      "loss": 0.774,
      "num_input_tokens_seen": 15979568,
      "step": 27530
    },
    {
      "epoch": 4.101131963062258,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.029002783916460217,
      "loss": 0.7996,
      "num_input_tokens_seen": 15982512,
      "step": 27535
    },
    {
      "epoch": 4.101876675603217,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.029002084792805142,
      "loss": 0.8277,
      "num_input_tokens_seen": 15985520,
      "step": 27540
    },
    {
      "epoch": 4.102621388144176,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.029001385432599076,
      "loss": 0.7802,
      "num_input_tokens_seen": 15988112,
      "step": 27545
    },
    {
      "epoch": 4.103366100685135,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.029000685835853832,
      "loss": 0.809,
      "num_input_tokens_seen": 15990928,
      "step": 27550
    },
    {
      "epoch": 4.104110813226095,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028999986002581232,
      "loss": 0.7633,
      "num_input_tokens_seen": 15993520,
      "step": 27555
    },
    {
      "epoch": 4.104855525767054,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.028999285932793096,
      "loss": 0.8196,
      "num_input_tokens_seen": 15996336,
      "step": 27560
    },
    {
      "epoch": 4.105600238308013,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028998585626501254,
      "loss": 0.8022,
      "num_input_tokens_seen": 15999376,
      "step": 27565
    },
    {
      "epoch": 4.106344950848972,
      "grad_norm": 0.01953125,
      "learning_rate": 0.028997885083717534,
      "loss": 0.7872,
      "num_input_tokens_seen": 16002736,
      "step": 27570
    },
    {
      "epoch": 4.107089663389932,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.028997184304453773,
      "loss": 0.7992,
      "num_input_tokens_seen": 16005584,
      "step": 27575
    },
    {
      "epoch": 4.10783437593089,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.028996483288721807,
      "loss": 0.7841,
      "num_input_tokens_seen": 16008176,
      "step": 27580
    },
    {
      "epoch": 4.10857908847185,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.028995782036533482,
      "loss": 0.8331,
      "num_input_tokens_seen": 16011216,
      "step": 27585
    },
    {
      "epoch": 4.109323801012809,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.028995080547900644,
      "loss": 0.8043,
      "num_input_tokens_seen": 16014160,
      "step": 27590
    },
    {
      "epoch": 4.1100685135537685,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.028994378822835152,
      "loss": 0.79,
      "num_input_tokens_seen": 16016944,
      "step": 27595
    },
    {
      "epoch": 4.110813226094727,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.028993676861348846,
      "loss": 0.8203,
      "num_input_tokens_seen": 16019824,
      "step": 27600
    },
    {
      "epoch": 4.111557938635687,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.02899297466345359,
      "loss": 0.8038,
      "num_input_tokens_seen": 16022736,
      "step": 27605
    },
    {
      "epoch": 4.112302651176646,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.02899227222916125,
      "loss": 0.7987,
      "num_input_tokens_seen": 16025904,
      "step": 27610
    },
    {
      "epoch": 4.113047363717605,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02899156955848369,
      "loss": 0.7949,
      "num_input_tokens_seen": 16029168,
      "step": 27615
    },
    {
      "epoch": 4.113792076258564,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02899086665143279,
      "loss": 0.7905,
      "num_input_tokens_seen": 16031920,
      "step": 27620
    },
    {
      "epoch": 4.114536788799524,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.028990163508020413,
      "loss": 0.795,
      "num_input_tokens_seen": 16034864,
      "step": 27625
    },
    {
      "epoch": 4.115281501340482,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.028989460128258444,
      "loss": 0.8069,
      "num_input_tokens_seen": 16037584,
      "step": 27630
    },
    {
      "epoch": 4.116026213881442,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.028988756512158764,
      "loss": 0.8155,
      "num_input_tokens_seen": 16040496,
      "step": 27635
    },
    {
      "epoch": 4.116770926422401,
      "grad_norm": 0.01953125,
      "learning_rate": 0.028988052659733258,
      "loss": 0.7944,
      "num_input_tokens_seen": 16043280,
      "step": 27640
    },
    {
      "epoch": 4.1175156389633605,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02898734857099382,
      "loss": 0.7895,
      "num_input_tokens_seen": 16046160,
      "step": 27645
    },
    {
      "epoch": 4.118260351504319,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.028986644245952348,
      "loss": 0.8044,
      "num_input_tokens_seen": 16049040,
      "step": 27650
    },
    {
      "epoch": 4.119005064045279,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.028985939684620734,
      "loss": 0.7966,
      "num_input_tokens_seen": 16051696,
      "step": 27655
    },
    {
      "epoch": 4.119749776586238,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.028985234887010883,
      "loss": 0.8025,
      "num_input_tokens_seen": 16054832,
      "step": 27660
    },
    {
      "epoch": 4.120494489127197,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.0289845298531347,
      "loss": 0.8164,
      "num_input_tokens_seen": 16057808,
      "step": 27665
    },
    {
      "epoch": 4.121239201668156,
      "grad_norm": 0.029296875,
      "learning_rate": 0.0289838245830041,
      "loss": 0.8031,
      "num_input_tokens_seen": 16060656,
      "step": 27670
    },
    {
      "epoch": 4.121983914209116,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.028983119076631,
      "loss": 0.8091,
      "num_input_tokens_seen": 16063408,
      "step": 27675
    },
    {
      "epoch": 4.1227286267500745,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02898241333402731,
      "loss": 0.8175,
      "num_input_tokens_seen": 16066288,
      "step": 27680
    },
    {
      "epoch": 4.123473339291033,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.028981707355204965,
      "loss": 0.7983,
      "num_input_tokens_seen": 16068880,
      "step": 27685
    },
    {
      "epoch": 4.124218051831993,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.028981001140175877,
      "loss": 0.7736,
      "num_input_tokens_seen": 16071824,
      "step": 27690
    },
    {
      "epoch": 4.124962764372952,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02898029468895199,
      "loss": 0.8066,
      "num_input_tokens_seen": 16074576,
      "step": 27695
    },
    {
      "epoch": 4.125707476913911,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02897958800154523,
      "loss": 0.8051,
      "num_input_tokens_seen": 16077264,
      "step": 27700
    },
    {
      "epoch": 4.12645218945487,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.028978881077967537,
      "loss": 0.8161,
      "num_input_tokens_seen": 16079984,
      "step": 27705
    },
    {
      "epoch": 4.12719690199583,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.028978173918230862,
      "loss": 0.8046,
      "num_input_tokens_seen": 16082704,
      "step": 27710
    },
    {
      "epoch": 4.127941614536788,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.028977466522347138,
      "loss": 0.7886,
      "num_input_tokens_seen": 16085296,
      "step": 27715
    },
    {
      "epoch": 4.128686327077748,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02897675889032833,
      "loss": 0.8055,
      "num_input_tokens_seen": 16088080,
      "step": 27720
    },
    {
      "epoch": 4.129431039618707,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02897605102218638,
      "loss": 0.8259,
      "num_input_tokens_seen": 16090960,
      "step": 27725
    },
    {
      "epoch": 4.1301757521596665,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028975342917933255,
      "loss": 0.8062,
      "num_input_tokens_seen": 16093648,
      "step": 27730
    },
    {
      "epoch": 4.130920464700625,
      "grad_norm": 0.0126953125,
      "learning_rate": 0.028974634577580913,
      "loss": 0.7942,
      "num_input_tokens_seen": 16096528,
      "step": 27735
    },
    {
      "epoch": 4.131665177241585,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.028973926001141327,
      "loss": 0.8108,
      "num_input_tokens_seen": 16099536,
      "step": 27740
    },
    {
      "epoch": 4.132409889782544,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.02897321718862646,
      "loss": 0.8118,
      "num_input_tokens_seen": 16102480,
      "step": 27745
    },
    {
      "epoch": 4.133154602323503,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.028972508140048297,
      "loss": 0.8076,
      "num_input_tokens_seen": 16105360,
      "step": 27750
    },
    {
      "epoch": 4.133899314864462,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.028971798855418805,
      "loss": 0.781,
      "num_input_tokens_seen": 16108336,
      "step": 27755
    },
    {
      "epoch": 4.134644027405422,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02897108933474997,
      "loss": 0.792,
      "num_input_tokens_seen": 16111056,
      "step": 27760
    },
    {
      "epoch": 4.1353887399463805,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028970379578053783,
      "loss": 0.7967,
      "num_input_tokens_seen": 16113776,
      "step": 27765
    },
    {
      "epoch": 4.13613345248734,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.028969669585342233,
      "loss": 0.8311,
      "num_input_tokens_seen": 16116464,
      "step": 27770
    },
    {
      "epoch": 4.136878165028299,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.028968959356627313,
      "loss": 0.7756,
      "num_input_tokens_seen": 16119472,
      "step": 27775
    },
    {
      "epoch": 4.1376228775692585,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.028968248891921018,
      "loss": 0.7904,
      "num_input_tokens_seen": 16122480,
      "step": 27780
    },
    {
      "epoch": 4.138367590110217,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02896753819123536,
      "loss": 0.8016,
      "num_input_tokens_seen": 16125296,
      "step": 27785
    },
    {
      "epoch": 4.139112302651177,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.028966827254582335,
      "loss": 0.7947,
      "num_input_tokens_seen": 16128240,
      "step": 27790
    },
    {
      "epoch": 4.139857015192136,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.028966116081973962,
      "loss": 0.7964,
      "num_input_tokens_seen": 16131440,
      "step": 27795
    },
    {
      "epoch": 4.140601727733095,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.028965404673422252,
      "loss": 0.8283,
      "num_input_tokens_seen": 16134224,
      "step": 27800
    },
    {
      "epoch": 4.141346440274054,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.028964693028939222,
      "loss": 0.8215,
      "num_input_tokens_seen": 16136944,
      "step": 27805
    },
    {
      "epoch": 4.142091152815014,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0289639811485369,
      "loss": 0.797,
      "num_input_tokens_seen": 16139792,
      "step": 27810
    },
    {
      "epoch": 4.1428358653559725,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028963269032227307,
      "loss": 0.796,
      "num_input_tokens_seen": 16142608,
      "step": 27815
    },
    {
      "epoch": 4.143580577896932,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.028962556680022476,
      "loss": 0.783,
      "num_input_tokens_seen": 16145520,
      "step": 27820
    },
    {
      "epoch": 4.144325290437891,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02896184409193444,
      "loss": 0.8115,
      "num_input_tokens_seen": 16148304,
      "step": 27825
    },
    {
      "epoch": 4.1450700029788505,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.028961131267975237,
      "loss": 0.8352,
      "num_input_tokens_seen": 16151184,
      "step": 27830
    },
    {
      "epoch": 4.145814715519809,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.02896041820815691,
      "loss": 0.7896,
      "num_input_tokens_seen": 16154224,
      "step": 27835
    },
    {
      "epoch": 4.146559428060769,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.028959704912491512,
      "loss": 0.8158,
      "num_input_tokens_seen": 16156848,
      "step": 27840
    },
    {
      "epoch": 4.147304140601728,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02895899138099108,
      "loss": 0.7791,
      "num_input_tokens_seen": 16159760,
      "step": 27845
    },
    {
      "epoch": 4.148048853142687,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02895827761366768,
      "loss": 0.8055,
      "num_input_tokens_seen": 16162672,
      "step": 27850
    },
    {
      "epoch": 4.148793565683646,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02895756361053337,
      "loss": 0.786,
      "num_input_tokens_seen": 16165584,
      "step": 27855
    },
    {
      "epoch": 4.149538278224606,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.028956849371600202,
      "loss": 0.7933,
      "num_input_tokens_seen": 16168624,
      "step": 27860
    },
    {
      "epoch": 4.1502829907655645,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.028956134896880252,
      "loss": 0.8036,
      "num_input_tokens_seen": 16171440,
      "step": 27865
    },
    {
      "epoch": 4.151027703306523,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.028955420186385587,
      "loss": 0.8294,
      "num_input_tokens_seen": 16174416,
      "step": 27870
    },
    {
      "epoch": 4.151772415847483,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02895470524012828,
      "loss": 0.8041,
      "num_input_tokens_seen": 16177328,
      "step": 27875
    },
    {
      "epoch": 4.152517128388442,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.028953990058120414,
      "loss": 0.8196,
      "num_input_tokens_seen": 16180208,
      "step": 27880
    },
    {
      "epoch": 4.153261840929401,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.028953274640374064,
      "loss": 0.812,
      "num_input_tokens_seen": 16183440,
      "step": 27885
    },
    {
      "epoch": 4.15400655347036,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028952558986901324,
      "loss": 0.7887,
      "num_input_tokens_seen": 16186448,
      "step": 27890
    },
    {
      "epoch": 4.15475126601132,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.028951843097714276,
      "loss": 0.7977,
      "num_input_tokens_seen": 16189456,
      "step": 27895
    },
    {
      "epoch": 4.1554959785522785,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.028951126972825026,
      "loss": 0.8046,
      "num_input_tokens_seen": 16192496,
      "step": 27900
    },
    {
      "epoch": 4.156240691093238,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02895041061224566,
      "loss": 0.7851,
      "num_input_tokens_seen": 16195568,
      "step": 27905
    },
    {
      "epoch": 4.156985403634197,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.028949694015988285,
      "loss": 0.8022,
      "num_input_tokens_seen": 16198384,
      "step": 27910
    },
    {
      "epoch": 4.1577301161751565,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.028948977184065014,
      "loss": 0.7936,
      "num_input_tokens_seen": 16201072,
      "step": 27915
    },
    {
      "epoch": 4.158474828716115,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.028948260116487944,
      "loss": 0.7939,
      "num_input_tokens_seen": 16203920,
      "step": 27920
    },
    {
      "epoch": 4.159219541257075,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028947542813269195,
      "loss": 0.7952,
      "num_input_tokens_seen": 16206640,
      "step": 27925
    },
    {
      "epoch": 4.159964253798034,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02894682527442089,
      "loss": 0.7775,
      "num_input_tokens_seen": 16209328,
      "step": 27930
    },
    {
      "epoch": 4.160708966338993,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.028946107499955145,
      "loss": 0.8316,
      "num_input_tokens_seen": 16211984,
      "step": 27935
    },
    {
      "epoch": 4.161453678879952,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.028945389489884087,
      "loss": 0.7757,
      "num_input_tokens_seen": 16214960,
      "step": 27940
    },
    {
      "epoch": 4.162198391420912,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.02894467124421985,
      "loss": 0.8007,
      "num_input_tokens_seen": 16217904,
      "step": 27945
    },
    {
      "epoch": 4.1629431039618705,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.028943952762974564,
      "loss": 0.8169,
      "num_input_tokens_seen": 16220976,
      "step": 27950
    },
    {
      "epoch": 4.16368781650283,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02894323404616037,
      "loss": 0.7954,
      "num_input_tokens_seen": 16223760,
      "step": 27955
    },
    {
      "epoch": 4.164432529043789,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.028942515093789405,
      "loss": 0.7904,
      "num_input_tokens_seen": 16226672,
      "step": 27960
    },
    {
      "epoch": 4.165177241584749,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.028941795905873816,
      "loss": 0.8043,
      "num_input_tokens_seen": 16229776,
      "step": 27965
    },
    {
      "epoch": 4.165921954125707,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.028941076482425755,
      "loss": 0.7937,
      "num_input_tokens_seen": 16232336,
      "step": 27970
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.028940356823457376,
      "loss": 0.7972,
      "num_input_tokens_seen": 16235472,
      "step": 27975
    },
    {
      "epoch": 4.167411379207626,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02893963692898084,
      "loss": 0.7871,
      "num_input_tokens_seen": 16238448,
      "step": 27980
    },
    {
      "epoch": 4.168156091748585,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.028938916799008306,
      "loss": 0.8118,
      "num_input_tokens_seen": 16241296,
      "step": 27985
    },
    {
      "epoch": 4.168900804289544,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02893819643355194,
      "loss": 0.7932,
      "num_input_tokens_seen": 16244592,
      "step": 27990
    },
    {
      "epoch": 4.169645516830504,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.028937475832623908,
      "loss": 0.7819,
      "num_input_tokens_seen": 16247408,
      "step": 27995
    },
    {
      "epoch": 4.1703902293714625,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.028936754996236394,
      "loss": 0.8088,
      "num_input_tokens_seen": 16250544,
      "step": 28000
    },
    {
      "epoch": 4.171134941912422,
      "grad_norm": 0.02734375,
      "learning_rate": 0.028936033924401562,
      "loss": 0.7888,
      "num_input_tokens_seen": 16253232,
      "step": 28005
    },
    {
      "epoch": 4.171879654453381,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.0289353126171316,
      "loss": 0.7916,
      "num_input_tokens_seen": 16256368,
      "step": 28010
    },
    {
      "epoch": 4.172624366994341,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.028934591074438697,
      "loss": 0.8014,
      "num_input_tokens_seen": 16259312,
      "step": 28015
    },
    {
      "epoch": 4.173369079535299,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02893386929633504,
      "loss": 0.7754,
      "num_input_tokens_seen": 16262128,
      "step": 28020
    },
    {
      "epoch": 4.174113792076259,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.028933147282832826,
      "loss": 0.8043,
      "num_input_tokens_seen": 16265232,
      "step": 28025
    },
    {
      "epoch": 4.174858504617218,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.028932425033944248,
      "loss": 0.7839,
      "num_input_tokens_seen": 16268240,
      "step": 28030
    },
    {
      "epoch": 4.1756032171581765,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02893170254968151,
      "loss": 0.81,
      "num_input_tokens_seen": 16270896,
      "step": 28035
    },
    {
      "epoch": 4.176347929699136,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.028930979830056814,
      "loss": 0.8248,
      "num_input_tokens_seen": 16273488,
      "step": 28040
    },
    {
      "epoch": 4.177092642240095,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.028930256875082376,
      "loss": 0.8548,
      "num_input_tokens_seen": 16276048,
      "step": 28045
    },
    {
      "epoch": 4.177837354781055,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.0289295336847704,
      "loss": 0.8071,
      "num_input_tokens_seen": 16279120,
      "step": 28050
    },
    {
      "epoch": 4.178582067322013,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.028928810259133117,
      "loss": 0.7916,
      "num_input_tokens_seen": 16282352,
      "step": 28055
    },
    {
      "epoch": 4.179326779862973,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.028928086598182735,
      "loss": 0.8374,
      "num_input_tokens_seen": 16285008,
      "step": 28060
    },
    {
      "epoch": 4.180071492403932,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.028927362701931494,
      "loss": 0.8086,
      "num_input_tokens_seen": 16288048,
      "step": 28065
    },
    {
      "epoch": 4.180816204944891,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02892663857039161,
      "loss": 0.7769,
      "num_input_tokens_seen": 16291184,
      "step": 28070
    },
    {
      "epoch": 4.18156091748585,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02892591420357532,
      "loss": 0.835,
      "num_input_tokens_seen": 16293936,
      "step": 28075
    },
    {
      "epoch": 4.18230563002681,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.028925189601494865,
      "loss": 0.7863,
      "num_input_tokens_seen": 16297040,
      "step": 28080
    },
    {
      "epoch": 4.1830503425677685,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028924464764162487,
      "loss": 0.7946,
      "num_input_tokens_seen": 16300272,
      "step": 28085
    },
    {
      "epoch": 4.183795055108728,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028923739691590424,
      "loss": 0.8021,
      "num_input_tokens_seen": 16303312,
      "step": 28090
    },
    {
      "epoch": 4.184539767649687,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028923014383790937,
      "loss": 0.7985,
      "num_input_tokens_seen": 16306512,
      "step": 28095
    },
    {
      "epoch": 4.185284480190647,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02892228884077627,
      "loss": 0.8157,
      "num_input_tokens_seen": 16309104,
      "step": 28100
    },
    {
      "epoch": 4.186029192731605,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.028921563062558678,
      "loss": 0.7924,
      "num_input_tokens_seen": 16311920,
      "step": 28105
    },
    {
      "epoch": 4.186773905272565,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02892083704915043,
      "loss": 0.7712,
      "num_input_tokens_seen": 16314768,
      "step": 28110
    },
    {
      "epoch": 4.187518617813524,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02892011080056379,
      "loss": 0.8055,
      "num_input_tokens_seen": 16317392,
      "step": 28115
    },
    {
      "epoch": 4.188263330354483,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.028919384316811028,
      "loss": 0.8095,
      "num_input_tokens_seen": 16320432,
      "step": 28120
    },
    {
      "epoch": 4.189008042895442,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.028918657597904414,
      "loss": 0.8309,
      "num_input_tokens_seen": 16323120,
      "step": 28125
    },
    {
      "epoch": 4.189752755436402,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.028917930643856223,
      "loss": 0.8174,
      "num_input_tokens_seen": 16325904,
      "step": 28130
    },
    {
      "epoch": 4.190497467977361,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02891720345467874,
      "loss": 0.7904,
      "num_input_tokens_seen": 16328848,
      "step": 28135
    },
    {
      "epoch": 4.19124218051832,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028916476030384254,
      "loss": 0.801,
      "num_input_tokens_seen": 16331632,
      "step": 28140
    },
    {
      "epoch": 4.191986893059279,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.028915748370985046,
      "loss": 0.8033,
      "num_input_tokens_seen": 16334160,
      "step": 28145
    },
    {
      "epoch": 4.192731605600239,
      "grad_norm": 0.01953125,
      "learning_rate": 0.028915020476493414,
      "loss": 0.8148,
      "num_input_tokens_seen": 16336816,
      "step": 28150
    },
    {
      "epoch": 4.193476318141197,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02891429234692165,
      "loss": 0.8093,
      "num_input_tokens_seen": 16339600,
      "step": 28155
    },
    {
      "epoch": 4.194221030682157,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02891356398228206,
      "loss": 0.8073,
      "num_input_tokens_seen": 16342480,
      "step": 28160
    },
    {
      "epoch": 4.194965743223116,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02891283538258695,
      "loss": 0.8151,
      "num_input_tokens_seen": 16345616,
      "step": 28165
    },
    {
      "epoch": 4.195710455764075,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.028912106547848627,
      "loss": 0.8082,
      "num_input_tokens_seen": 16348272,
      "step": 28170
    },
    {
      "epoch": 4.196455168305034,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0289113774780794,
      "loss": 0.7951,
      "num_input_tokens_seen": 16351184,
      "step": 28175
    },
    {
      "epoch": 4.197199880845994,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.028910648173291593,
      "loss": 0.8067,
      "num_input_tokens_seen": 16354000,
      "step": 28180
    },
    {
      "epoch": 4.197944593386953,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.028909918633497525,
      "loss": 0.8029,
      "num_input_tokens_seen": 16356688,
      "step": 28185
    },
    {
      "epoch": 4.198689305927912,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02890918885870951,
      "loss": 0.7999,
      "num_input_tokens_seen": 16359728,
      "step": 28190
    },
    {
      "epoch": 4.199434018468871,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.028908458848939892,
      "loss": 0.8205,
      "num_input_tokens_seen": 16362736,
      "step": 28195
    },
    {
      "epoch": 4.200178731009831,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.028907728604201,
      "loss": 0.8082,
      "num_input_tokens_seen": 16365456,
      "step": 28200
    },
    {
      "epoch": 4.200923443550789,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.028906998124505166,
      "loss": 0.8069,
      "num_input_tokens_seen": 16368080,
      "step": 28205
    },
    {
      "epoch": 4.201668156091749,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.028906267409864732,
      "loss": 0.799,
      "num_input_tokens_seen": 16371088,
      "step": 28210
    },
    {
      "epoch": 4.202412868632708,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.028905536460292043,
      "loss": 0.7914,
      "num_input_tokens_seen": 16374224,
      "step": 28215
    },
    {
      "epoch": 4.203157581173667,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.028904805275799448,
      "loss": 0.827,
      "num_input_tokens_seen": 16376848,
      "step": 28220
    },
    {
      "epoch": 4.203902293714626,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.0289040738563993,
      "loss": 0.8039,
      "num_input_tokens_seen": 16380112,
      "step": 28225
    },
    {
      "epoch": 4.204647006255585,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.028903342202103964,
      "loss": 0.7926,
      "num_input_tokens_seen": 16382960,
      "step": 28230
    },
    {
      "epoch": 4.205391718796545,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.028902610312925785,
      "loss": 0.796,
      "num_input_tokens_seen": 16385776,
      "step": 28235
    },
    {
      "epoch": 4.206136431337503,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02890187818887714,
      "loss": 0.7928,
      "num_input_tokens_seen": 16388592,
      "step": 28240
    },
    {
      "epoch": 4.206881143878463,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.028901145829970387,
      "loss": 0.8158,
      "num_input_tokens_seen": 16391504,
      "step": 28245
    },
    {
      "epoch": 4.207625856419422,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.028900413236217903,
      "loss": 0.8243,
      "num_input_tokens_seen": 16394672,
      "step": 28250
    },
    {
      "epoch": 4.208370568960381,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.028899680407632064,
      "loss": 0.7836,
      "num_input_tokens_seen": 16397616,
      "step": 28255
    },
    {
      "epoch": 4.20911528150134,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028898947344225257,
      "loss": 0.8005,
      "num_input_tokens_seen": 16400400,
      "step": 28260
    },
    {
      "epoch": 4.2098599940423,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.028898214046009858,
      "loss": 0.7995,
      "num_input_tokens_seen": 16403312,
      "step": 28265
    },
    {
      "epoch": 4.210604706583259,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.028897480512998258,
      "loss": 0.7981,
      "num_input_tokens_seen": 16406160,
      "step": 28270
    },
    {
      "epoch": 4.211349419124218,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.02889674674520285,
      "loss": 0.8116,
      "num_input_tokens_seen": 16409296,
      "step": 28275
    },
    {
      "epoch": 4.212094131665177,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.028896012742636028,
      "loss": 0.8168,
      "num_input_tokens_seen": 16412144,
      "step": 28280
    },
    {
      "epoch": 4.212838844206137,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.028895278505310193,
      "loss": 0.8094,
      "num_input_tokens_seen": 16415056,
      "step": 28285
    },
    {
      "epoch": 4.213583556747095,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028894544033237757,
      "loss": 0.7958,
      "num_input_tokens_seen": 16417904,
      "step": 28290
    },
    {
      "epoch": 4.214328269288055,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.028893809326431118,
      "loss": 0.8025,
      "num_input_tokens_seen": 16420432,
      "step": 28295
    },
    {
      "epoch": 4.215072981829014,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02889307438490269,
      "loss": 0.8036,
      "num_input_tokens_seen": 16423408,
      "step": 28300
    },
    {
      "epoch": 4.2158176943699734,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02889233920866489,
      "loss": 0.8098,
      "num_input_tokens_seen": 16426576,
      "step": 28305
    },
    {
      "epoch": 4.216562406910932,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02889160379773014,
      "loss": 0.8034,
      "num_input_tokens_seen": 16429424,
      "step": 28310
    },
    {
      "epoch": 4.217307119451892,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.028890868152110862,
      "loss": 0.7883,
      "num_input_tokens_seen": 16432368,
      "step": 28315
    },
    {
      "epoch": 4.218051831992851,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028890132271819486,
      "loss": 0.7928,
      "num_input_tokens_seen": 16435216,
      "step": 28320
    },
    {
      "epoch": 4.21879654453381,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02888939615686844,
      "loss": 0.8031,
      "num_input_tokens_seen": 16438576,
      "step": 28325
    },
    {
      "epoch": 4.219541257074769,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.028888659807270164,
      "loss": 0.7936,
      "num_input_tokens_seen": 16441392,
      "step": 28330
    },
    {
      "epoch": 4.220285969615729,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.0288879232230371,
      "loss": 0.8086,
      "num_input_tokens_seen": 16444208,
      "step": 28335
    },
    {
      "epoch": 4.221030682156687,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.028887186404181683,
      "loss": 0.8167,
      "num_input_tokens_seen": 16446896,
      "step": 28340
    },
    {
      "epoch": 4.221775394697647,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.028886449350716367,
      "loss": 0.7839,
      "num_input_tokens_seen": 16449680,
      "step": 28345
    },
    {
      "epoch": 4.222520107238606,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028885712062653602,
      "loss": 0.7788,
      "num_input_tokens_seen": 16452528,
      "step": 28350
    },
    {
      "epoch": 4.2232648197795655,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.02888497454000585,
      "loss": 0.7974,
      "num_input_tokens_seen": 16455824,
      "step": 28355
    },
    {
      "epoch": 4.224009532320524,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02888423678278556,
      "loss": 0.7978,
      "num_input_tokens_seen": 16458672,
      "step": 28360
    },
    {
      "epoch": 4.224754244861484,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.0288834987910052,
      "loss": 0.7922,
      "num_input_tokens_seen": 16461584,
      "step": 28365
    },
    {
      "epoch": 4.225498957402443,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02888276056467724,
      "loss": 0.8065,
      "num_input_tokens_seen": 16464368,
      "step": 28370
    },
    {
      "epoch": 4.226243669943402,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.028882022103814155,
      "loss": 0.8167,
      "num_input_tokens_seen": 16467440,
      "step": 28375
    },
    {
      "epoch": 4.226988382484361,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.028881283408428406,
      "loss": 0.8222,
      "num_input_tokens_seen": 16470192,
      "step": 28380
    },
    {
      "epoch": 4.22773309502532,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02888054447853249,
      "loss": 0.7872,
      "num_input_tokens_seen": 16473552,
      "step": 28385
    },
    {
      "epoch": 4.2284778075662794,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.02887980531413888,
      "loss": 0.8307,
      "num_input_tokens_seen": 16476368,
      "step": 28390
    },
    {
      "epoch": 4.229222520107238,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.028879065915260065,
      "loss": 0.8063,
      "num_input_tokens_seen": 16479120,
      "step": 28395
    },
    {
      "epoch": 4.229967232648198,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02887832628190854,
      "loss": 0.7901,
      "num_input_tokens_seen": 16482064,
      "step": 28400
    },
    {
      "epoch": 4.230711945189157,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0288775864140968,
      "loss": 0.807,
      "num_input_tokens_seen": 16484688,
      "step": 28405
    },
    {
      "epoch": 4.231456657730116,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02887684631183733,
      "loss": 0.7908,
      "num_input_tokens_seen": 16487536,
      "step": 28410
    },
    {
      "epoch": 4.232201370271075,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028876105975142657,
      "loss": 0.8341,
      "num_input_tokens_seen": 16490000,
      "step": 28415
    },
    {
      "epoch": 4.232946082812035,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.028875365404025275,
      "loss": 0.8385,
      "num_input_tokens_seen": 16492880,
      "step": 28420
    },
    {
      "epoch": 4.233690795352993,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028874624598497694,
      "loss": 0.7762,
      "num_input_tokens_seen": 16495600,
      "step": 28425
    },
    {
      "epoch": 4.234435507893953,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.028873883558572434,
      "loss": 0.7979,
      "num_input_tokens_seen": 16498480,
      "step": 28430
    },
    {
      "epoch": 4.235180220434912,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.028873142284262007,
      "loss": 0.8216,
      "num_input_tokens_seen": 16501424,
      "step": 28435
    },
    {
      "epoch": 4.2359249329758715,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.028872400775578948,
      "loss": 0.8027,
      "num_input_tokens_seen": 16504400,
      "step": 28440
    },
    {
      "epoch": 4.23666964551683,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.028871659032535774,
      "loss": 0.7931,
      "num_input_tokens_seen": 16507248,
      "step": 28445
    },
    {
      "epoch": 4.23741435805779,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02887091705514502,
      "loss": 0.804,
      "num_input_tokens_seen": 16510032,
      "step": 28450
    },
    {
      "epoch": 4.238159070598749,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02887017484341922,
      "loss": 0.7948,
      "num_input_tokens_seen": 16512816,
      "step": 28455
    },
    {
      "epoch": 4.238903783139708,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02886943239737091,
      "loss": 0.8183,
      "num_input_tokens_seen": 16515504,
      "step": 28460
    },
    {
      "epoch": 4.239648495680667,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02886868971701264,
      "loss": 0.8096,
      "num_input_tokens_seen": 16518128,
      "step": 28465
    },
    {
      "epoch": 4.240393208221627,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.028867946802356954,
      "loss": 0.7951,
      "num_input_tokens_seen": 16521040,
      "step": 28470
    },
    {
      "epoch": 4.2411379207625854,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.028867203653416395,
      "loss": 0.8216,
      "num_input_tokens_seen": 16524112,
      "step": 28475
    },
    {
      "epoch": 4.241882633303545,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.028866460270203526,
      "loss": 0.8047,
      "num_input_tokens_seen": 16526800,
      "step": 28480
    },
    {
      "epoch": 4.242627345844504,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02886571665273091,
      "loss": 0.799,
      "num_input_tokens_seen": 16529488,
      "step": 28485
    },
    {
      "epoch": 4.2433720583854635,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.0288649728010111,
      "loss": 0.8025,
      "num_input_tokens_seen": 16532336,
      "step": 28490
    },
    {
      "epoch": 4.244116770926422,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.028864228715056665,
      "loss": 0.8093,
      "num_input_tokens_seen": 16535024,
      "step": 28495
    },
    {
      "epoch": 4.244861483467382,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02886348439488018,
      "loss": 0.814,
      "num_input_tokens_seen": 16537872,
      "step": 28500
    },
    {
      "epoch": 4.245606196008341,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.028862739840494214,
      "loss": 0.7904,
      "num_input_tokens_seen": 16540624,
      "step": 28505
    },
    {
      "epoch": 4.2463509085493,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.028861995051911345,
      "loss": 0.8155,
      "num_input_tokens_seen": 16543632,
      "step": 28510
    },
    {
      "epoch": 4.247095621090259,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02886125002914416,
      "loss": 0.8086,
      "num_input_tokens_seen": 16546480,
      "step": 28515
    },
    {
      "epoch": 4.247840333631219,
      "grad_norm": 0.025390625,
      "learning_rate": 0.028860504772205246,
      "loss": 0.8123,
      "num_input_tokens_seen": 16549328,
      "step": 28520
    },
    {
      "epoch": 4.2485850461721775,
      "grad_norm": 0.0126953125,
      "learning_rate": 0.028859759281107192,
      "loss": 0.7889,
      "num_input_tokens_seen": 16552144,
      "step": 28525
    },
    {
      "epoch": 4.249329758713137,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.02885901355586259,
      "loss": 0.7953,
      "num_input_tokens_seen": 16555120,
      "step": 28530
    },
    {
      "epoch": 4.250074471254096,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.028858267596484036,
      "loss": 0.8041,
      "num_input_tokens_seen": 16558032,
      "step": 28535
    },
    {
      "epoch": 4.2508191837950555,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02885752140298414,
      "loss": 0.7973,
      "num_input_tokens_seen": 16561040,
      "step": 28540
    },
    {
      "epoch": 4.251563896336014,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.028856774975375504,
      "loss": 0.7986,
      "num_input_tokens_seen": 16563792,
      "step": 28545
    },
    {
      "epoch": 4.252308608876973,
      "grad_norm": 0.021484375,
      "learning_rate": 0.028856028313670738,
      "loss": 0.7934,
      "num_input_tokens_seen": 16566672,
      "step": 28550
    },
    {
      "epoch": 4.253053321417933,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.028855281417882454,
      "loss": 0.7989,
      "num_input_tokens_seen": 16569648,
      "step": 28555
    },
    {
      "epoch": 4.253798033958892,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.028854534288023276,
      "loss": 0.8214,
      "num_input_tokens_seen": 16572432,
      "step": 28560
    },
    {
      "epoch": 4.254542746499851,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.028853786924105815,
      "loss": 0.8088,
      "num_input_tokens_seen": 16575088,
      "step": 28565
    },
    {
      "epoch": 4.25528745904081,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.028853039326142714,
      "loss": 0.8104,
      "num_input_tokens_seen": 16578096,
      "step": 28570
    },
    {
      "epoch": 4.2560321715817695,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.028852291494146587,
      "loss": 0.8121,
      "num_input_tokens_seen": 16580944,
      "step": 28575
    },
    {
      "epoch": 4.256776884122728,
      "grad_norm": 0.033203125,
      "learning_rate": 0.028851543428130073,
      "loss": 0.8067,
      "num_input_tokens_seen": 16583664,
      "step": 28580
    },
    {
      "epoch": 4.257521596663688,
      "grad_norm": 0.021484375,
      "learning_rate": 0.028850795128105816,
      "loss": 0.8101,
      "num_input_tokens_seen": 16586640,
      "step": 28585
    },
    {
      "epoch": 4.258266309204647,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.028850046594086447,
      "loss": 0.7872,
      "num_input_tokens_seen": 16589424,
      "step": 28590
    },
    {
      "epoch": 4.259011021745606,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028849297826084624,
      "loss": 0.7935,
      "num_input_tokens_seen": 16592112,
      "step": 28595
    },
    {
      "epoch": 4.259755734286565,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.028848548824112988,
      "loss": 0.8116,
      "num_input_tokens_seen": 16595120,
      "step": 28600
    },
    {
      "epoch": 4.260500446827525,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02884779958818419,
      "loss": 0.8004,
      "num_input_tokens_seen": 16597808,
      "step": 28605
    },
    {
      "epoch": 4.2612451593684835,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.0288470501183109,
      "loss": 0.8087,
      "num_input_tokens_seen": 16600656,
      "step": 28610
    },
    {
      "epoch": 4.261989871909443,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.028846300414505765,
      "loss": 0.8122,
      "num_input_tokens_seen": 16603408,
      "step": 28615
    },
    {
      "epoch": 4.262734584450402,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02884555047678146,
      "loss": 0.7927,
      "num_input_tokens_seen": 16606640,
      "step": 28620
    },
    {
      "epoch": 4.2634792969913615,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028844800305150648,
      "loss": 0.7889,
      "num_input_tokens_seen": 16609360,
      "step": 28625
    },
    {
      "epoch": 4.26422400953232,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.02884404989962601,
      "loss": 0.7953,
      "num_input_tokens_seen": 16612112,
      "step": 28630
    },
    {
      "epoch": 4.26496872207328,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.028843299260220216,
      "loss": 0.7995,
      "num_input_tokens_seen": 16615248,
      "step": 28635
    },
    {
      "epoch": 4.265713434614239,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.028842548386945957,
      "loss": 0.8114,
      "num_input_tokens_seen": 16618256,
      "step": 28640
    },
    {
      "epoch": 4.266458147155198,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02884179727981591,
      "loss": 0.8001,
      "num_input_tokens_seen": 16621616,
      "step": 28645
    },
    {
      "epoch": 4.267202859696157,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028841045938842763,
      "loss": 0.798,
      "num_input_tokens_seen": 16624336,
      "step": 28650
    },
    {
      "epoch": 4.267947572237117,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.02884029436403921,
      "loss": 0.8111,
      "num_input_tokens_seen": 16627376,
      "step": 28655
    },
    {
      "epoch": 4.2686922847780755,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.028839542555417957,
      "loss": 0.7942,
      "num_input_tokens_seen": 16630512,
      "step": 28660
    },
    {
      "epoch": 4.269436997319035,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.028838790512991694,
      "loss": 0.7896,
      "num_input_tokens_seen": 16633424,
      "step": 28665
    },
    {
      "epoch": 4.270181709859994,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.028838038236773136,
      "loss": 0.7937,
      "num_input_tokens_seen": 16636336,
      "step": 28670
    },
    {
      "epoch": 4.2709264224009535,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02883728572677498,
      "loss": 0.8169,
      "num_input_tokens_seen": 16638928,
      "step": 28675
    },
    {
      "epoch": 4.271671134941912,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.028836532983009946,
      "loss": 0.8178,
      "num_input_tokens_seen": 16641872,
      "step": 28680
    },
    {
      "epoch": 4.272415847482872,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.028835780005490752,
      "loss": 0.7904,
      "num_input_tokens_seen": 16644560,
      "step": 28685
    },
    {
      "epoch": 4.273160560023831,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.028835026794230115,
      "loss": 0.8185,
      "num_input_tokens_seen": 16647344,
      "step": 28690
    },
    {
      "epoch": 4.27390527256479,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.028834273349240762,
      "loss": 0.8119,
      "num_input_tokens_seen": 16650224,
      "step": 28695
    },
    {
      "epoch": 4.274649985105749,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028833519670535424,
      "loss": 0.7908,
      "num_input_tokens_seen": 16653424,
      "step": 28700
    },
    {
      "epoch": 4.275394697646709,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02883276575812683,
      "loss": 0.7785,
      "num_input_tokens_seen": 16656112,
      "step": 28705
    },
    {
      "epoch": 4.2761394101876675,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028832011612027714,
      "loss": 0.7966,
      "num_input_tokens_seen": 16658832,
      "step": 28710
    },
    {
      "epoch": 4.276884122728626,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028831257232250825,
      "loss": 0.8094,
      "num_input_tokens_seen": 16661840,
      "step": 28715
    },
    {
      "epoch": 4.277628835269586,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.0288305026188089,
      "loss": 0.796,
      "num_input_tokens_seen": 16665008,
      "step": 28720
    },
    {
      "epoch": 4.278373547810546,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02882974777171469,
      "loss": 0.8114,
      "num_input_tokens_seen": 16668080,
      "step": 28725
    },
    {
      "epoch": 4.279118260351504,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028828992690980944,
      "loss": 0.8229,
      "num_input_tokens_seen": 16670928,
      "step": 28730
    },
    {
      "epoch": 4.279862972892463,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02882823737662043,
      "loss": 0.7977,
      "num_input_tokens_seen": 16673296,
      "step": 28735
    },
    {
      "epoch": 4.280607685433423,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.028827481828645894,
      "loss": 0.7949,
      "num_input_tokens_seen": 16676208,
      "step": 28740
    },
    {
      "epoch": 4.2813523979743815,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02882672604707011,
      "loss": 0.7697,
      "num_input_tokens_seen": 16679248,
      "step": 28745
    },
    {
      "epoch": 4.282097110515341,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.028825970031905835,
      "loss": 0.8053,
      "num_input_tokens_seen": 16682128,
      "step": 28750
    },
    {
      "epoch": 4.2828418230563,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02882521378316585,
      "loss": 0.828,
      "num_input_tokens_seen": 16685040,
      "step": 28755
    },
    {
      "epoch": 4.2835865355972595,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028824457300862937,
      "loss": 0.7926,
      "num_input_tokens_seen": 16687696,
      "step": 28760
    },
    {
      "epoch": 4.284331248138218,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02882370058500986,
      "loss": 0.813,
      "num_input_tokens_seen": 16690448,
      "step": 28765
    },
    {
      "epoch": 4.285075960679178,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02882294363561942,
      "loss": 0.7882,
      "num_input_tokens_seen": 16693264,
      "step": 28770
    },
    {
      "epoch": 4.285820673220137,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02882218645270439,
      "loss": 0.7977,
      "num_input_tokens_seen": 16695984,
      "step": 28775
    },
    {
      "epoch": 4.286565385761096,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.028821429036277567,
      "loss": 0.809,
      "num_input_tokens_seen": 16698864,
      "step": 28780
    },
    {
      "epoch": 4.287310098302055,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02882067138635175,
      "loss": 0.7987,
      "num_input_tokens_seen": 16701648,
      "step": 28785
    },
    {
      "epoch": 4.288054810843015,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02881991350293974,
      "loss": 0.8067,
      "num_input_tokens_seen": 16704528,
      "step": 28790
    },
    {
      "epoch": 4.2887995233839735,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.028819155386054333,
      "loss": 0.8158,
      "num_input_tokens_seen": 16707344,
      "step": 28795
    },
    {
      "epoch": 4.289544235924933,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02881839703570834,
      "loss": 0.7983,
      "num_input_tokens_seen": 16710320,
      "step": 28800
    },
    {
      "epoch": 4.290288948465892,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.028817638451914578,
      "loss": 0.8046,
      "num_input_tokens_seen": 16713520,
      "step": 28805
    },
    {
      "epoch": 4.291033661006852,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.028816879634685856,
      "loss": 0.8165,
      "num_input_tokens_seen": 16716304,
      "step": 28810
    },
    {
      "epoch": 4.29177837354781,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.028816120584034998,
      "loss": 0.793,
      "num_input_tokens_seen": 16719152,
      "step": 28815
    },
    {
      "epoch": 4.29252308608877,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02881536129997482,
      "loss": 0.8,
      "num_input_tokens_seen": 16722096,
      "step": 28820
    },
    {
      "epoch": 4.293267798629729,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.028814601782518153,
      "loss": 0.8151,
      "num_input_tokens_seen": 16724880,
      "step": 28825
    },
    {
      "epoch": 4.294012511170688,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.028813842031677833,
      "loss": 0.8017,
      "num_input_tokens_seen": 16727792,
      "step": 28830
    },
    {
      "epoch": 4.294757223711647,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.028813082047466694,
      "loss": 0.8132,
      "num_input_tokens_seen": 16730896,
      "step": 28835
    },
    {
      "epoch": 4.295501936252607,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02881232182989757,
      "loss": 0.7828,
      "num_input_tokens_seen": 16733872,
      "step": 28840
    },
    {
      "epoch": 4.2962466487935655,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.028811561378983306,
      "loss": 0.7956,
      "num_input_tokens_seen": 16736528,
      "step": 28845
    },
    {
      "epoch": 4.296991361334525,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.028810800694736752,
      "loss": 0.8124,
      "num_input_tokens_seen": 16739344,
      "step": 28850
    },
    {
      "epoch": 4.297736073875484,
      "grad_norm": 0.01953125,
      "learning_rate": 0.028810039777170754,
      "loss": 0.7824,
      "num_input_tokens_seen": 16742416,
      "step": 28855
    },
    {
      "epoch": 4.298480786416444,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02880927862629817,
      "loss": 0.7988,
      "num_input_tokens_seen": 16745456,
      "step": 28860
    },
    {
      "epoch": 4.299225498957402,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.028808517242131862,
      "loss": 0.7958,
      "num_input_tokens_seen": 16748432,
      "step": 28865
    },
    {
      "epoch": 4.299970211498362,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.028807755624684687,
      "loss": 0.8158,
      "num_input_tokens_seen": 16751408,
      "step": 28870
    },
    {
      "epoch": 4.300714924039321,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02880699377396952,
      "loss": 0.8046,
      "num_input_tokens_seen": 16754384,
      "step": 28875
    },
    {
      "epoch": 4.30145963658028,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028806231689999216,
      "loss": 0.7934,
      "num_input_tokens_seen": 16757104,
      "step": 28880
    },
    {
      "epoch": 4.302204349121239,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02880546937278667,
      "loss": 0.7944,
      "num_input_tokens_seen": 16759824,
      "step": 28885
    },
    {
      "epoch": 4.302949061662199,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.028804706822344744,
      "loss": 0.8003,
      "num_input_tokens_seen": 16762800,
      "step": 28890
    },
    {
      "epoch": 4.303693774203158,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.028803944038686326,
      "loss": 0.7937,
      "num_input_tokens_seen": 16765616,
      "step": 28895
    },
    {
      "epoch": 4.304438486744116,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.028803181021824303,
      "loss": 0.8254,
      "num_input_tokens_seen": 16768272,
      "step": 28900
    },
    {
      "epoch": 4.305183199285076,
      "grad_norm": 0.01953125,
      "learning_rate": 0.028802417771771567,
      "loss": 0.8038,
      "num_input_tokens_seen": 16771088,
      "step": 28905
    },
    {
      "epoch": 4.305927911826035,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.028801654288541013,
      "loss": 0.8041,
      "num_input_tokens_seen": 16774352,
      "step": 28910
    },
    {
      "epoch": 4.306672624366994,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.028800890572145535,
      "loss": 0.8051,
      "num_input_tokens_seen": 16777296,
      "step": 28915
    },
    {
      "epoch": 4.307417336907953,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.028800126622598038,
      "loss": 0.8048,
      "num_input_tokens_seen": 16780144,
      "step": 28920
    },
    {
      "epoch": 4.308162049448913,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02879936243991143,
      "loss": 0.8152,
      "num_input_tokens_seen": 16782992,
      "step": 28925
    },
    {
      "epoch": 4.3089067619898715,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.028798598024098614,
      "loss": 0.7976,
      "num_input_tokens_seen": 16786192,
      "step": 28930
    },
    {
      "epoch": 4.309651474530831,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02879783337517251,
      "loss": 0.7927,
      "num_input_tokens_seen": 16789200,
      "step": 28935
    },
    {
      "epoch": 4.31039618707179,
      "grad_norm": 0.01171875,
      "learning_rate": 0.028797068493146036,
      "loss": 0.8133,
      "num_input_tokens_seen": 16792080,
      "step": 28940
    },
    {
      "epoch": 4.31114089961275,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.028796303378032115,
      "loss": 0.8173,
      "num_input_tokens_seen": 16794928,
      "step": 28945
    },
    {
      "epoch": 4.311885612153708,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02879553802984366,
      "loss": 0.808,
      "num_input_tokens_seen": 16797776,
      "step": 28950
    },
    {
      "epoch": 4.312630324694668,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.028794772448593624,
      "loss": 0.7968,
      "num_input_tokens_seen": 16800848,
      "step": 28955
    },
    {
      "epoch": 4.313375037235627,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.028794006634294923,
      "loss": 0.8113,
      "num_input_tokens_seen": 16803632,
      "step": 28960
    },
    {
      "epoch": 4.314119749776586,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.0287932405869605,
      "loss": 0.7849,
      "num_input_tokens_seen": 16806224,
      "step": 28965
    },
    {
      "epoch": 4.314864462317545,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.028792474306603293,
      "loss": 0.8253,
      "num_input_tokens_seen": 16808912,
      "step": 28970
    },
    {
      "epoch": 4.315609174858505,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028791707793236254,
      "loss": 0.8158,
      "num_input_tokens_seen": 16811728,
      "step": 28975
    },
    {
      "epoch": 4.316353887399464,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.028790941046872326,
      "loss": 0.8283,
      "num_input_tokens_seen": 16815024,
      "step": 28980
    },
    {
      "epoch": 4.317098599940423,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.028790174067524467,
      "loss": 0.7938,
      "num_input_tokens_seen": 16817712,
      "step": 28985
    },
    {
      "epoch": 4.317843312481382,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.028789406855205632,
      "loss": 0.7963,
      "num_input_tokens_seen": 16820240,
      "step": 28990
    },
    {
      "epoch": 4.318588025022342,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.028788639409928783,
      "loss": 0.7992,
      "num_input_tokens_seen": 16822992,
      "step": 28995
    },
    {
      "epoch": 4.3193327375633,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028787871731706886,
      "loss": 0.804,
      "num_input_tokens_seen": 16825872,
      "step": 29000
    },
    {
      "epoch": 4.32007745010426,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.02878710382055291,
      "loss": 0.8061,
      "num_input_tokens_seen": 16828688,
      "step": 29005
    },
    {
      "epoch": 4.320822162645219,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.028786335676479827,
      "loss": 0.805,
      "num_input_tokens_seen": 16831632,
      "step": 29010
    },
    {
      "epoch": 4.321566875186178,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.028785567299500617,
      "loss": 0.7983,
      "num_input_tokens_seen": 16834384,
      "step": 29015
    },
    {
      "epoch": 4.322311587727137,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.028784798689628254,
      "loss": 0.8197,
      "num_input_tokens_seen": 16837168,
      "step": 29020
    },
    {
      "epoch": 4.323056300268097,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.028784029846875732,
      "loss": 0.8095,
      "num_input_tokens_seen": 16840432,
      "step": 29025
    },
    {
      "epoch": 4.323801012809056,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.028783260771256036,
      "loss": 0.8073,
      "num_input_tokens_seen": 16843088,
      "step": 29030
    },
    {
      "epoch": 4.324545725350015,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.02878249146278215,
      "loss": 0.7959,
      "num_input_tokens_seen": 16846000,
      "step": 29035
    },
    {
      "epoch": 4.325290437890974,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.028781721921467086,
      "loss": 0.7855,
      "num_input_tokens_seen": 16848752,
      "step": 29040
    },
    {
      "epoch": 4.326035150431934,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.028780952147323832,
      "loss": 0.8135,
      "num_input_tokens_seen": 16851728,
      "step": 29045
    },
    {
      "epoch": 4.326779862972892,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.028780182140365405,
      "loss": 0.8156,
      "num_input_tokens_seen": 16854832,
      "step": 29050
    },
    {
      "epoch": 4.327524575513852,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.028779411900604797,
      "loss": 0.7999,
      "num_input_tokens_seen": 16857680,
      "step": 29055
    },
    {
      "epoch": 4.328269288054811,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.028778641428055036,
      "loss": 0.8047,
      "num_input_tokens_seen": 16860784,
      "step": 29060
    },
    {
      "epoch": 4.32901400059577,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.02877787072272913,
      "loss": 0.803,
      "num_input_tokens_seen": 16864080,
      "step": 29065
    },
    {
      "epoch": 4.329758713136729,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.028777099784640103,
      "loss": 0.7933,
      "num_input_tokens_seen": 16866960,
      "step": 29070
    },
    {
      "epoch": 4.330503425677689,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.02877632861380098,
      "loss": 0.8079,
      "num_input_tokens_seen": 16870128,
      "step": 29075
    },
    {
      "epoch": 4.331248138218648,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.028775557210224782,
      "loss": 0.8021,
      "num_input_tokens_seen": 16873072,
      "step": 29080
    },
    {
      "epoch": 4.331992850759606,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.028774785573924544,
      "loss": 0.8014,
      "num_input_tokens_seen": 16875984,
      "step": 29085
    },
    {
      "epoch": 4.332737563300566,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.02877401370491331,
      "loss": 0.7911,
      "num_input_tokens_seen": 16879056,
      "step": 29090
    },
    {
      "epoch": 4.333482275841525,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02877324160320411,
      "loss": 0.8162,
      "num_input_tokens_seen": 16881776,
      "step": 29095
    },
    {
      "epoch": 4.334226988382484,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02877246926880999,
      "loss": 0.7956,
      "num_input_tokens_seen": 16884592,
      "step": 29100
    },
    {
      "epoch": 4.334971700923443,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.028771696701744003,
      "loss": 0.7967,
      "num_input_tokens_seen": 16887344,
      "step": 29105
    },
    {
      "epoch": 4.335716413464403,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.0287709239020192,
      "loss": 0.7903,
      "num_input_tokens_seen": 16890576,
      "step": 29110
    },
    {
      "epoch": 4.336461126005362,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.028770150869648625,
      "loss": 0.7816,
      "num_input_tokens_seen": 16893488,
      "step": 29115
    },
    {
      "epoch": 4.337205838546321,
      "grad_norm": 0.01953125,
      "learning_rate": 0.028769377604645353,
      "loss": 0.8034,
      "num_input_tokens_seen": 16896528,
      "step": 29120
    },
    {
      "epoch": 4.33795055108728,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.028768604107022436,
      "loss": 0.8105,
      "num_input_tokens_seen": 16899568,
      "step": 29125
    },
    {
      "epoch": 4.33869526362824,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.028767830376792952,
      "loss": 0.8048,
      "num_input_tokens_seen": 16902256,
      "step": 29130
    },
    {
      "epoch": 4.339439976169198,
      "grad_norm": 0.01171875,
      "learning_rate": 0.028767056413969965,
      "loss": 0.8034,
      "num_input_tokens_seen": 16904976,
      "step": 29135
    },
    {
      "epoch": 4.340184688710158,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02876628221856655,
      "loss": 0.8054,
      "num_input_tokens_seen": 16907760,
      "step": 29140
    },
    {
      "epoch": 4.340929401251117,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.028765507790595786,
      "loss": 0.812,
      "num_input_tokens_seen": 16910640,
      "step": 29145
    },
    {
      "epoch": 4.3416741137920765,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.028764733130070767,
      "loss": 0.8362,
      "num_input_tokens_seen": 16913296,
      "step": 29150
    },
    {
      "epoch": 4.342418826333035,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.028763958237004568,
      "loss": 0.7985,
      "num_input_tokens_seen": 16916112,
      "step": 29155
    },
    {
      "epoch": 4.343163538873995,
      "grad_norm": 0.021484375,
      "learning_rate": 0.028763183111410278,
      "loss": 0.8081,
      "num_input_tokens_seen": 16918960,
      "step": 29160
    },
    {
      "epoch": 4.343908251414954,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028762407753301,
      "loss": 0.8219,
      "num_input_tokens_seen": 16922064,
      "step": 29165
    },
    {
      "epoch": 4.344652963955913,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.028761632162689836,
      "loss": 0.8112,
      "num_input_tokens_seen": 16924912,
      "step": 29170
    },
    {
      "epoch": 4.345397676496872,
      "grad_norm": 0.029296875,
      "learning_rate": 0.028760856339589882,
      "loss": 0.8045,
      "num_input_tokens_seen": 16927760,
      "step": 29175
    },
    {
      "epoch": 4.346142389037832,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.02876008028401424,
      "loss": 0.7986,
      "num_input_tokens_seen": 16930352,
      "step": 29180
    },
    {
      "epoch": 4.34688710157879,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02875930399597603,
      "loss": 0.7969,
      "num_input_tokens_seen": 16933040,
      "step": 29185
    },
    {
      "epoch": 4.34763181411975,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.028758527475488363,
      "loss": 0.8027,
      "num_input_tokens_seen": 16935632,
      "step": 29190
    },
    {
      "epoch": 4.348376526660709,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02875775072256436,
      "loss": 0.8265,
      "num_input_tokens_seen": 16938768,
      "step": 29195
    },
    {
      "epoch": 4.3491212392016685,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02875697373721714,
      "loss": 0.7933,
      "num_input_tokens_seen": 16941936,
      "step": 29200
    },
    {
      "epoch": 4.349865951742627,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02875619651945983,
      "loss": 0.8157,
      "num_input_tokens_seen": 16944816,
      "step": 29205
    },
    {
      "epoch": 4.350610664283587,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.028755419069305562,
      "loss": 0.806,
      "num_input_tokens_seen": 16947760,
      "step": 29210
    },
    {
      "epoch": 4.351355376824546,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02875464138676747,
      "loss": 0.8233,
      "num_input_tokens_seen": 16950480,
      "step": 29215
    },
    {
      "epoch": 4.352100089365505,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.028753863471858692,
      "loss": 0.8094,
      "num_input_tokens_seen": 16953296,
      "step": 29220
    },
    {
      "epoch": 4.352844801906464,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.028753085324592368,
      "loss": 0.7987,
      "num_input_tokens_seen": 16956112,
      "step": 29225
    },
    {
      "epoch": 4.353589514447424,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02875230694498164,
      "loss": 0.8042,
      "num_input_tokens_seen": 16958992,
      "step": 29230
    },
    {
      "epoch": 4.3543342269883825,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02875152833303967,
      "loss": 0.8111,
      "num_input_tokens_seen": 16961904,
      "step": 29235
    },
    {
      "epoch": 4.355078939529342,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.028750749488779606,
      "loss": 0.8109,
      "num_input_tokens_seen": 16964784,
      "step": 29240
    },
    {
      "epoch": 4.355823652070301,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.028749970412214602,
      "loss": 0.8132,
      "num_input_tokens_seen": 16967664,
      "step": 29245
    },
    {
      "epoch": 4.35656836461126,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.028749191103357823,
      "loss": 0.7905,
      "num_input_tokens_seen": 16970384,
      "step": 29250
    },
    {
      "epoch": 4.357313077152219,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.028748411562222437,
      "loss": 0.8069,
      "num_input_tokens_seen": 16973328,
      "step": 29255
    },
    {
      "epoch": 4.358057789693178,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028747631788821608,
      "loss": 0.8083,
      "num_input_tokens_seen": 16976176,
      "step": 29260
    },
    {
      "epoch": 4.358802502234138,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.028746851783168513,
      "loss": 0.7915,
      "num_input_tokens_seen": 16979024,
      "step": 29265
    },
    {
      "epoch": 4.359547214775096,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.028746071545276326,
      "loss": 0.8021,
      "num_input_tokens_seen": 16982192,
      "step": 29270
    },
    {
      "epoch": 4.360291927316056,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.028745291075158236,
      "loss": 0.8125,
      "num_input_tokens_seen": 16985104,
      "step": 29275
    },
    {
      "epoch": 4.361036639857015,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028744510372827423,
      "loss": 0.8092,
      "num_input_tokens_seen": 16988112,
      "step": 29280
    },
    {
      "epoch": 4.3617813523979745,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.028743729438297075,
      "loss": 0.7869,
      "num_input_tokens_seen": 16990960,
      "step": 29285
    },
    {
      "epoch": 4.362526064938933,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.028742948271580384,
      "loss": 0.7845,
      "num_input_tokens_seen": 16993840,
      "step": 29290
    },
    {
      "epoch": 4.363270777479893,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.028742166872690558,
      "loss": 0.7816,
      "num_input_tokens_seen": 16996816,
      "step": 29295
    },
    {
      "epoch": 4.364015490020852,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.028741385241640778,
      "loss": 0.7909,
      "num_input_tokens_seen": 16999920,
      "step": 29300
    },
    {
      "epoch": 4.364760202561811,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.028740603378444265,
      "loss": 0.7952,
      "num_input_tokens_seen": 17002704,
      "step": 29305
    },
    {
      "epoch": 4.36550491510277,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02873982128311423,
      "loss": 0.7948,
      "num_input_tokens_seen": 17005552,
      "step": 29310
    },
    {
      "epoch": 4.36624962764373,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.028739038955663872,
      "loss": 0.8091,
      "num_input_tokens_seen": 17008528,
      "step": 29315
    },
    {
      "epoch": 4.3669943401846885,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02873825639610642,
      "loss": 0.7983,
      "num_input_tokens_seen": 17011216,
      "step": 29320
    },
    {
      "epoch": 4.367739052725648,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02873747360445508,
      "loss": 0.794,
      "num_input_tokens_seen": 17014192,
      "step": 29325
    },
    {
      "epoch": 4.368483765266607,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02873669058072309,
      "loss": 0.8125,
      "num_input_tokens_seen": 17017072,
      "step": 29330
    },
    {
      "epoch": 4.3692284778075665,
      "grad_norm": 0.017578125,
      "learning_rate": 0.02873590732492368,
      "loss": 0.7691,
      "num_input_tokens_seen": 17019952,
      "step": 29335
    },
    {
      "epoch": 4.369973190348525,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02873512383707007,
      "loss": 0.7743,
      "num_input_tokens_seen": 17022864,
      "step": 29340
    },
    {
      "epoch": 4.370717902889485,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.028734340117175503,
      "loss": 0.8227,
      "num_input_tokens_seen": 17025616,
      "step": 29345
    },
    {
      "epoch": 4.371462615430444,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.028733556165253218,
      "loss": 0.7928,
      "num_input_tokens_seen": 17028464,
      "step": 29350
    },
    {
      "epoch": 4.372207327971403,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02873277198131646,
      "loss": 0.7917,
      "num_input_tokens_seen": 17031504,
      "step": 29355
    },
    {
      "epoch": 4.372952040512362,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02873198756537848,
      "loss": 0.7817,
      "num_input_tokens_seen": 17034384,
      "step": 29360
    },
    {
      "epoch": 4.373696753053322,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.028731202917452524,
      "loss": 0.7882,
      "num_input_tokens_seen": 17037296,
      "step": 29365
    },
    {
      "epoch": 4.3744414655942805,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02873041803755185,
      "loss": 0.7914,
      "num_input_tokens_seen": 17040144,
      "step": 29370
    },
    {
      "epoch": 4.37518617813524,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.028729632925689717,
      "loss": 0.8113,
      "num_input_tokens_seen": 17043120,
      "step": 29375
    },
    {
      "epoch": 4.375930890676199,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.028728847581879394,
      "loss": 0.7939,
      "num_input_tokens_seen": 17045904,
      "step": 29380
    },
    {
      "epoch": 4.3766756032171585,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.028728062006134144,
      "loss": 0.7757,
      "num_input_tokens_seen": 17048496,
      "step": 29385
    },
    {
      "epoch": 4.377420315758117,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.028727276198467235,
      "loss": 0.7878,
      "num_input_tokens_seen": 17051600,
      "step": 29390
    },
    {
      "epoch": 4.378165028299077,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.028726490158891952,
      "loss": 0.8034,
      "num_input_tokens_seen": 17054448,
      "step": 29395
    },
    {
      "epoch": 4.378909740840036,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.028725703887421565,
      "loss": 0.8164,
      "num_input_tokens_seen": 17057424,
      "step": 29400
    },
    {
      "epoch": 4.379654453380995,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.028724917384069362,
      "loss": 0.7801,
      "num_input_tokens_seen": 17060336,
      "step": 29405
    },
    {
      "epoch": 4.380399165921954,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.028724130648848628,
      "loss": 0.8119,
      "num_input_tokens_seen": 17063216,
      "step": 29410
    },
    {
      "epoch": 4.381143878462913,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.028723343681772658,
      "loss": 0.7975,
      "num_input_tokens_seen": 17066320,
      "step": 29415
    },
    {
      "epoch": 4.3818885910038725,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.028722556482854742,
      "loss": 0.8062,
      "num_input_tokens_seen": 17069456,
      "step": 29420
    },
    {
      "epoch": 4.382633303544832,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.028721769052108184,
      "loss": 0.8106,
      "num_input_tokens_seen": 17072176,
      "step": 29425
    },
    {
      "epoch": 4.383378016085791,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.028720981389546277,
      "loss": 0.8139,
      "num_input_tokens_seen": 17075056,
      "step": 29430
    },
    {
      "epoch": 4.38412272862675,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.028720193495182345,
      "loss": 0.8292,
      "num_input_tokens_seen": 17077712,
      "step": 29435
    },
    {
      "epoch": 4.384867441167709,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02871940536902968,
      "loss": 0.8245,
      "num_input_tokens_seen": 17080688,
      "step": 29440
    },
    {
      "epoch": 4.385612153708668,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.028718617011101608,
      "loss": 0.8232,
      "num_input_tokens_seen": 17083792,
      "step": 29445
    },
    {
      "epoch": 4.386356866249628,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.028717828421411445,
      "loss": 0.8151,
      "num_input_tokens_seen": 17086800,
      "step": 29450
    },
    {
      "epoch": 4.3871015787905865,
      "grad_norm": 0.01055908203125,
      "learning_rate": 0.028717039599972512,
      "loss": 0.7856,
      "num_input_tokens_seen": 17089616,
      "step": 29455
    },
    {
      "epoch": 4.387846291331546,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.028716250546798137,
      "loss": 0.8003,
      "num_input_tokens_seen": 17092592,
      "step": 29460
    },
    {
      "epoch": 4.388591003872505,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.028715461261901647,
      "loss": 0.8048,
      "num_input_tokens_seen": 17095440,
      "step": 29465
    },
    {
      "epoch": 4.3893357164134645,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02871467174529638,
      "loss": 0.7855,
      "num_input_tokens_seen": 17098000,
      "step": 29470
    },
    {
      "epoch": 4.390080428954423,
      "grad_norm": 0.0123291015625,
      "learning_rate": 0.02871388199699568,
      "loss": 0.7998,
      "num_input_tokens_seen": 17100816,
      "step": 29475
    },
    {
      "epoch": 4.390825141495383,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.028713092017012873,
      "loss": 0.7971,
      "num_input_tokens_seen": 17103600,
      "step": 29480
    },
    {
      "epoch": 4.391569854036342,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.028712301805361315,
      "loss": 0.7998,
      "num_input_tokens_seen": 17106416,
      "step": 29485
    },
    {
      "epoch": 4.392314566577301,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.028711511362054357,
      "loss": 0.8124,
      "num_input_tokens_seen": 17109648,
      "step": 29490
    },
    {
      "epoch": 4.39305927911826,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02871072068710535,
      "loss": 0.8286,
      "num_input_tokens_seen": 17112464,
      "step": 29495
    },
    {
      "epoch": 4.39380399165922,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.02870992978052765,
      "loss": 0.8333,
      "num_input_tokens_seen": 17115568,
      "step": 29500
    },
    {
      "epoch": 4.3945487042001785,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02870913864233463,
      "loss": 0.8117,
      "num_input_tokens_seen": 17118736,
      "step": 29505
    },
    {
      "epoch": 4.395293416741138,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.028708347272539632,
      "loss": 0.8091,
      "num_input_tokens_seen": 17121296,
      "step": 29510
    },
    {
      "epoch": 4.396038129282097,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.028707555671156047,
      "loss": 0.8005,
      "num_input_tokens_seen": 17123952,
      "step": 29515
    },
    {
      "epoch": 4.396782841823057,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02870676383819724,
      "loss": 0.7965,
      "num_input_tokens_seen": 17126896,
      "step": 29520
    },
    {
      "epoch": 4.397527554364015,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.028705971773676585,
      "loss": 0.814,
      "num_input_tokens_seen": 17129744,
      "step": 29525
    },
    {
      "epoch": 4.398272266904975,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.028705179477607472,
      "loss": 0.7872,
      "num_input_tokens_seen": 17132656,
      "step": 29530
    },
    {
      "epoch": 4.399016979445934,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.028704386950003283,
      "loss": 0.8159,
      "num_input_tokens_seen": 17135408,
      "step": 29535
    },
    {
      "epoch": 4.399761691986893,
      "grad_norm": 0.010986328125,
      "learning_rate": 0.028703594190877396,
      "loss": 0.8018,
      "num_input_tokens_seen": 17138352,
      "step": 29540
    },
    {
      "epoch": 4.400506404527852,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.028702801200243218,
      "loss": 0.8103,
      "num_input_tokens_seen": 17141168,
      "step": 29545
    },
    {
      "epoch": 4.401251117068812,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02870200797811414,
      "loss": 0.8181,
      "num_input_tokens_seen": 17144144,
      "step": 29550
    },
    {
      "epoch": 4.4019958296097705,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.028701214524503565,
      "loss": 0.7945,
      "num_input_tokens_seen": 17147024,
      "step": 29555
    },
    {
      "epoch": 4.40274054215073,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.0287004208394249,
      "loss": 0.7948,
      "num_input_tokens_seen": 17149744,
      "step": 29560
    },
    {
      "epoch": 4.403485254691689,
      "grad_norm": 0.01092529296875,
      "learning_rate": 0.02869962692289154,
      "loss": 0.8005,
      "num_input_tokens_seen": 17152784,
      "step": 29565
    },
    {
      "epoch": 4.404229967232649,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02869883277491691,
      "loss": 0.802,
      "num_input_tokens_seen": 17155760,
      "step": 29570
    },
    {
      "epoch": 4.404974679773607,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028698038395514422,
      "loss": 0.7959,
      "num_input_tokens_seen": 17158384,
      "step": 29575
    },
    {
      "epoch": 4.405719392314566,
      "grad_norm": 0.01116943359375,
      "learning_rate": 0.028697243784697502,
      "loss": 0.8069,
      "num_input_tokens_seen": 17161328,
      "step": 29580
    },
    {
      "epoch": 4.406464104855526,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.028696448942479566,
      "loss": 0.8041,
      "num_input_tokens_seen": 17163984,
      "step": 29585
    },
    {
      "epoch": 4.407208817396485,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.028695653868874047,
      "loss": 0.8186,
      "num_input_tokens_seen": 17166864,
      "step": 29590
    },
    {
      "epoch": 4.407953529937444,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.028694858563894376,
      "loss": 0.8093,
      "num_input_tokens_seen": 17169584,
      "step": 29595
    },
    {
      "epoch": 4.408698242478403,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.028694063027553986,
      "loss": 0.8034,
      "num_input_tokens_seen": 17172304,
      "step": 29600
    },
    {
      "epoch": 4.409442955019363,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02869326725986632,
      "loss": 0.7937,
      "num_input_tokens_seen": 17175408,
      "step": 29605
    },
    {
      "epoch": 4.410187667560321,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.028692471260844826,
      "loss": 0.8139,
      "num_input_tokens_seen": 17178352,
      "step": 29610
    },
    {
      "epoch": 4.410932380101281,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.028691675030502937,
      "loss": 0.816,
      "num_input_tokens_seen": 17181392,
      "step": 29615
    },
    {
      "epoch": 4.41167709264224,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02869087856885412,
      "loss": 0.8085,
      "num_input_tokens_seen": 17184240,
      "step": 29620
    },
    {
      "epoch": 4.412421805183199,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.028690081875911825,
      "loss": 0.803,
      "num_input_tokens_seen": 17187344,
      "step": 29625
    },
    {
      "epoch": 4.413166517724158,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.028689284951689507,
      "loss": 0.807,
      "num_input_tokens_seen": 17190064,
      "step": 29630
    },
    {
      "epoch": 4.413911230265118,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.028688487796200635,
      "loss": 0.8036,
      "num_input_tokens_seen": 17193136,
      "step": 29635
    },
    {
      "epoch": 4.4146559428060765,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.028687690409458674,
      "loss": 0.796,
      "num_input_tokens_seen": 17196048,
      "step": 29640
    },
    {
      "epoch": 4.415400655347036,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.028686892791477094,
      "loss": 0.7981,
      "num_input_tokens_seen": 17198992,
      "step": 29645
    },
    {
      "epoch": 4.416145367887995,
      "grad_norm": 0.021484375,
      "learning_rate": 0.028686094942269372,
      "loss": 0.8038,
      "num_input_tokens_seen": 17201872,
      "step": 29650
    },
    {
      "epoch": 4.416890080428955,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.028685296861848987,
      "loss": 0.7947,
      "num_input_tokens_seen": 17204720,
      "step": 29655
    },
    {
      "epoch": 4.417634792969913,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02868449855022942,
      "loss": 0.8242,
      "num_input_tokens_seen": 17207568,
      "step": 29660
    },
    {
      "epoch": 4.418379505510873,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.028683700007424158,
      "loss": 0.7825,
      "num_input_tokens_seen": 17210384,
      "step": 29665
    },
    {
      "epoch": 4.419124218051832,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.028682901233446693,
      "loss": 0.8128,
      "num_input_tokens_seen": 17213392,
      "step": 29670
    },
    {
      "epoch": 4.419868930592791,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.028682102228310514,
      "loss": 0.8033,
      "num_input_tokens_seen": 17216368,
      "step": 29675
    },
    {
      "epoch": 4.42061364313375,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028681302992029128,
      "loss": 0.7882,
      "num_input_tokens_seen": 17219280,
      "step": 29680
    },
    {
      "epoch": 4.42135835567471,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.028680503524616034,
      "loss": 0.7868,
      "num_input_tokens_seen": 17222128,
      "step": 29685
    },
    {
      "epoch": 4.422103068215669,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.028679703826084733,
      "loss": 0.7985,
      "num_input_tokens_seen": 17224688,
      "step": 29690
    },
    {
      "epoch": 4.422847780756628,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02867890389644874,
      "loss": 0.8133,
      "num_input_tokens_seen": 17227792,
      "step": 29695
    },
    {
      "epoch": 4.423592493297587,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.02867810373572157,
      "loss": 0.7984,
      "num_input_tokens_seen": 17230672,
      "step": 29700
    },
    {
      "epoch": 4.424337205838547,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.028677303343916737,
      "loss": 0.8323,
      "num_input_tokens_seen": 17233584,
      "step": 29705
    },
    {
      "epoch": 4.425081918379505,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02867650272104777,
      "loss": 0.8331,
      "num_input_tokens_seen": 17236240,
      "step": 29710
    },
    {
      "epoch": 4.425826630920465,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.028675701867128187,
      "loss": 0.827,
      "num_input_tokens_seen": 17239216,
      "step": 29715
    },
    {
      "epoch": 4.426571343461424,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.028674900782171518,
      "loss": 0.8114,
      "num_input_tokens_seen": 17241840,
      "step": 29720
    },
    {
      "epoch": 4.427316056002383,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.028674099466191302,
      "loss": 0.8122,
      "num_input_tokens_seen": 17245136,
      "step": 29725
    },
    {
      "epoch": 4.428060768543342,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02867329791920107,
      "loss": 0.7976,
      "num_input_tokens_seen": 17247984,
      "step": 29730
    },
    {
      "epoch": 4.428805481084302,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02867249614121437,
      "loss": 0.8037,
      "num_input_tokens_seen": 17251056,
      "step": 29735
    },
    {
      "epoch": 4.429550193625261,
      "grad_norm": 0.017578125,
      "learning_rate": 0.02867169413224474,
      "loss": 0.7934,
      "num_input_tokens_seen": 17254000,
      "step": 29740
    },
    {
      "epoch": 4.43029490616622,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.028670891892305737,
      "loss": 0.8015,
      "num_input_tokens_seen": 17256944,
      "step": 29745
    },
    {
      "epoch": 4.431039618707179,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02867008942141091,
      "loss": 0.8073,
      "num_input_tokens_seen": 17259888,
      "step": 29750
    },
    {
      "epoch": 4.431784331248139,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02866928671957381,
      "loss": 0.7995,
      "num_input_tokens_seen": 17262704,
      "step": 29755
    },
    {
      "epoch": 4.432529043789097,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.028668483786808006,
      "loss": 0.7938,
      "num_input_tokens_seen": 17265520,
      "step": 29760
    },
    {
      "epoch": 4.433273756330056,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.028667680623127065,
      "loss": 0.8085,
      "num_input_tokens_seen": 17268496,
      "step": 29765
    },
    {
      "epoch": 4.434018468871016,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.028666877228544548,
      "loss": 0.8093,
      "num_input_tokens_seen": 17271280,
      "step": 29770
    },
    {
      "epoch": 4.434763181411975,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.02866607360307403,
      "loss": 0.8154,
      "num_input_tokens_seen": 17274256,
      "step": 29775
    },
    {
      "epoch": 4.435507893952934,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.02866526974672909,
      "loss": 0.8069,
      "num_input_tokens_seen": 17277360,
      "step": 29780
    },
    {
      "epoch": 4.436252606493893,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.028664465659523304,
      "loss": 0.7846,
      "num_input_tokens_seen": 17280464,
      "step": 29785
    },
    {
      "epoch": 4.436997319034853,
      "grad_norm": 0.0108642578125,
      "learning_rate": 0.028663661341470262,
      "loss": 0.792,
      "num_input_tokens_seen": 17282928,
      "step": 29790
    },
    {
      "epoch": 4.437742031575811,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.028662856792583544,
      "loss": 0.814,
      "num_input_tokens_seen": 17285712,
      "step": 29795
    },
    {
      "epoch": 4.438486744116771,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.028662052012876752,
      "loss": 0.7939,
      "num_input_tokens_seen": 17288560,
      "step": 29800
    },
    {
      "epoch": 4.43923145665773,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.02866124700236347,
      "loss": 0.7977,
      "num_input_tokens_seen": 17291824,
      "step": 29805
    },
    {
      "epoch": 4.439976169198689,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.028660441761057313,
      "loss": 0.8224,
      "num_input_tokens_seen": 17294832,
      "step": 29810
    },
    {
      "epoch": 4.440720881739648,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.028659636288971873,
      "loss": 0.7906,
      "num_input_tokens_seen": 17297968,
      "step": 29815
    },
    {
      "epoch": 4.441465594280608,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.02865883058612076,
      "loss": 0.8062,
      "num_input_tokens_seen": 17300912,
      "step": 29820
    },
    {
      "epoch": 4.442210306821567,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.028658024652517586,
      "loss": 0.8051,
      "num_input_tokens_seen": 17303760,
      "step": 29825
    },
    {
      "epoch": 4.442955019362526,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.028657218488175967,
      "loss": 0.8172,
      "num_input_tokens_seen": 17306704,
      "step": 29830
    },
    {
      "epoch": 4.443699731903485,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.02865641209310952,
      "loss": 0.8009,
      "num_input_tokens_seen": 17309680,
      "step": 29835
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.028655605467331872,
      "loss": 0.7941,
      "num_input_tokens_seen": 17312432,
      "step": 29840
    },
    {
      "epoch": 4.445189156985403,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.028654798610856652,
      "loss": 0.8076,
      "num_input_tokens_seen": 17315248,
      "step": 29845
    },
    {
      "epoch": 4.445933869526363,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.028653991523697485,
      "loss": 0.7852,
      "num_input_tokens_seen": 17318000,
      "step": 29850
    },
    {
      "epoch": 4.446678582067322,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.028653184205868005,
      "loss": 0.8191,
      "num_input_tokens_seen": 17320688,
      "step": 29855
    },
    {
      "epoch": 4.4474232946082815,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.02865237665738186,
      "loss": 0.8143,
      "num_input_tokens_seen": 17323312,
      "step": 29860
    },
    {
      "epoch": 4.44816800714924,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.028651568878252685,
      "loss": 0.8014,
      "num_input_tokens_seen": 17326320,
      "step": 29865
    },
    {
      "epoch": 4.4489127196902,
      "grad_norm": 0.03125,
      "learning_rate": 0.02865076086849413,
      "loss": 0.7912,
      "num_input_tokens_seen": 17329328,
      "step": 29870
    },
    {
      "epoch": 4.449657432231159,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.028649952628119844,
      "loss": 0.7981,
      "num_input_tokens_seen": 17332368,
      "step": 29875
    },
    {
      "epoch": 4.450402144772118,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.028649144157143477,
      "loss": 0.8073,
      "num_input_tokens_seen": 17335184,
      "step": 29880
    },
    {
      "epoch": 4.451146857313077,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.028648335455578697,
      "loss": 0.7834,
      "num_input_tokens_seen": 17337872,
      "step": 29885
    },
    {
      "epoch": 4.451891569854037,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.028647526523439156,
      "loss": 0.7742,
      "num_input_tokens_seen": 17340848,
      "step": 29890
    },
    {
      "epoch": 4.452636282394995,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02864671736073853,
      "loss": 0.8094,
      "num_input_tokens_seen": 17343984,
      "step": 29895
    },
    {
      "epoch": 4.453380994935955,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.028645907967490482,
      "loss": 0.793,
      "num_input_tokens_seen": 17347152,
      "step": 29900
    },
    {
      "epoch": 4.454125707476914,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02864509834370869,
      "loss": 0.8088,
      "num_input_tokens_seen": 17350224,
      "step": 29905
    },
    {
      "epoch": 4.4548704200178735,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.028644288489406828,
      "loss": 0.8129,
      "num_input_tokens_seen": 17353008,
      "step": 29910
    },
    {
      "epoch": 4.455615132558832,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.028643478404598577,
      "loss": 0.7928,
      "num_input_tokens_seen": 17356240,
      "step": 29915
    },
    {
      "epoch": 4.456359845099792,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.028642668089297628,
      "loss": 0.8264,
      "num_input_tokens_seen": 17359248,
      "step": 29920
    },
    {
      "epoch": 4.457104557640751,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.028641857543517668,
      "loss": 0.8164,
      "num_input_tokens_seen": 17362384,
      "step": 29925
    },
    {
      "epoch": 4.457849270181709,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.02864104676727239,
      "loss": 0.7989,
      "num_input_tokens_seen": 17365456,
      "step": 29930
    },
    {
      "epoch": 4.458593982722669,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02864023576057549,
      "loss": 0.7892,
      "num_input_tokens_seen": 17368368,
      "step": 29935
    },
    {
      "epoch": 4.459338695263629,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02863942452344067,
      "loss": 0.8131,
      "num_input_tokens_seen": 17371312,
      "step": 29940
    },
    {
      "epoch": 4.4600834078045875,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.028638613055881636,
      "loss": 0.7873,
      "num_input_tokens_seen": 17374000,
      "step": 29945
    },
    {
      "epoch": 4.460828120345546,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02863780135791209,
      "loss": 0.8117,
      "num_input_tokens_seen": 17376976,
      "step": 29950
    },
    {
      "epoch": 4.461572832886506,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.028636989429545758,
      "loss": 0.791,
      "num_input_tokens_seen": 17380112,
      "step": 29955
    },
    {
      "epoch": 4.462317545427465,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02863617727079635,
      "loss": 0.7993,
      "num_input_tokens_seen": 17383120,
      "step": 29960
    },
    {
      "epoch": 4.463062257968424,
      "grad_norm": 0.0126953125,
      "learning_rate": 0.02863536488167758,
      "loss": 0.8171,
      "num_input_tokens_seen": 17386064,
      "step": 29965
    },
    {
      "epoch": 4.463806970509383,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.028634552262203183,
      "loss": 0.819,
      "num_input_tokens_seen": 17388560,
      "step": 29970
    },
    {
      "epoch": 4.464551683050343,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02863373941238688,
      "loss": 0.7904,
      "num_input_tokens_seen": 17391504,
      "step": 29975
    },
    {
      "epoch": 4.465296395591301,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.028632926332242414,
      "loss": 0.8185,
      "num_input_tokens_seen": 17394288,
      "step": 29980
    },
    {
      "epoch": 4.466041108132261,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02863211302178351,
      "loss": 0.7705,
      "num_input_tokens_seen": 17397360,
      "step": 29985
    },
    {
      "epoch": 4.46678582067322,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028631299481023907,
      "loss": 0.7935,
      "num_input_tokens_seen": 17400432,
      "step": 29990
    },
    {
      "epoch": 4.4675305332141795,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.028630485709977355,
      "loss": 0.8127,
      "num_input_tokens_seen": 17403408,
      "step": 29995
    },
    {
      "epoch": 4.468275245755138,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028629671708657607,
      "loss": 0.7707,
      "num_input_tokens_seen": 17406608,
      "step": 30000
    },
    {
      "epoch": 4.469019958296098,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.028628857477078398,
      "loss": 0.7969,
      "num_input_tokens_seen": 17409360,
      "step": 30005
    },
    {
      "epoch": 4.469764670837057,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.0286280430152535,
      "loss": 0.8239,
      "num_input_tokens_seen": 17412464,
      "step": 30010
    },
    {
      "epoch": 4.470509383378016,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.028627228323196665,
      "loss": 0.8087,
      "num_input_tokens_seen": 17415344,
      "step": 30015
    },
    {
      "epoch": 4.471254095918975,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.02862641340092166,
      "loss": 0.7876,
      "num_input_tokens_seen": 17418224,
      "step": 30020
    },
    {
      "epoch": 4.471998808459935,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02862559824844224,
      "loss": 0.8228,
      "num_input_tokens_seen": 17421232,
      "step": 30025
    },
    {
      "epoch": 4.4727435210008935,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.028624782865772195,
      "loss": 0.8014,
      "num_input_tokens_seen": 17423920,
      "step": 30030
    },
    {
      "epoch": 4.473488233541853,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.028623967252925292,
      "loss": 0.8014,
      "num_input_tokens_seen": 17426736,
      "step": 30035
    },
    {
      "epoch": 4.474232946082812,
      "grad_norm": 0.01953125,
      "learning_rate": 0.028623151409915303,
      "loss": 0.8074,
      "num_input_tokens_seen": 17429712,
      "step": 30040
    },
    {
      "epoch": 4.4749776586237715,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.02862233533675602,
      "loss": 0.8086,
      "num_input_tokens_seen": 17432656,
      "step": 30045
    },
    {
      "epoch": 4.47572237116473,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02862151903346123,
      "loss": 0.829,
      "num_input_tokens_seen": 17435696,
      "step": 30050
    },
    {
      "epoch": 4.47646708370569,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.028620702500044712,
      "loss": 0.7985,
      "num_input_tokens_seen": 17438768,
      "step": 30055
    },
    {
      "epoch": 4.477211796246649,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02861988573652028,
      "loss": 0.7832,
      "num_input_tokens_seen": 17441648,
      "step": 30060
    },
    {
      "epoch": 4.477956508787608,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.028619068742901713,
      "loss": 0.8283,
      "num_input_tokens_seen": 17444464,
      "step": 30065
    },
    {
      "epoch": 4.478701221328567,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.028618251519202824,
      "loss": 0.7999,
      "num_input_tokens_seen": 17447184,
      "step": 30070
    },
    {
      "epoch": 4.479445933869527,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.028617434065437417,
      "loss": 0.8181,
      "num_input_tokens_seen": 17450160,
      "step": 30075
    },
    {
      "epoch": 4.4801906464104855,
      "grad_norm": 0.029296875,
      "learning_rate": 0.028616616381619296,
      "loss": 0.7969,
      "num_input_tokens_seen": 17453360,
      "step": 30080
    },
    {
      "epoch": 4.480935358951445,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.028615798467762285,
      "loss": 0.8062,
      "num_input_tokens_seen": 17456144,
      "step": 30085
    },
    {
      "epoch": 4.481680071492404,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.0286149803238802,
      "loss": 0.8042,
      "num_input_tokens_seen": 17459152,
      "step": 30090
    },
    {
      "epoch": 4.4824247840333635,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.028614161949986856,
      "loss": 0.7962,
      "num_input_tokens_seen": 17462032,
      "step": 30095
    },
    {
      "epoch": 4.483169496574322,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.028613343346096085,
      "loss": 0.8153,
      "num_input_tokens_seen": 17464848,
      "step": 30100
    },
    {
      "epoch": 4.483914209115282,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.028612524512221713,
      "loss": 0.7971,
      "num_input_tokens_seen": 17467696,
      "step": 30105
    },
    {
      "epoch": 4.484658921656241,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.028611705448377572,
      "loss": 0.8003,
      "num_input_tokens_seen": 17470480,
      "step": 30110
    },
    {
      "epoch": 4.4854036341971995,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.028610886154577506,
      "loss": 0.7916,
      "num_input_tokens_seen": 17473232,
      "step": 30115
    },
    {
      "epoch": 4.486148346738159,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.028610066630835353,
      "loss": 0.791,
      "num_input_tokens_seen": 17475984,
      "step": 30120
    },
    {
      "epoch": 4.486893059279118,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028609246877164953,
      "loss": 0.7847,
      "num_input_tokens_seen": 17479056,
      "step": 30125
    },
    {
      "epoch": 4.4876377718200775,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.028608426893580157,
      "loss": 0.7859,
      "num_input_tokens_seen": 17482032,
      "step": 30130
    },
    {
      "epoch": 4.488382484361036,
      "grad_norm": 0.01068115234375,
      "learning_rate": 0.02860760668009483,
      "loss": 0.8001,
      "num_input_tokens_seen": 17484880,
      "step": 30135
    },
    {
      "epoch": 4.489127196901996,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.028606786236722807,
      "loss": 0.803,
      "num_input_tokens_seen": 17487664,
      "step": 30140
    },
    {
      "epoch": 4.489871909442955,
      "grad_norm": 0.017578125,
      "learning_rate": 0.028605965563477966,
      "loss": 0.7945,
      "num_input_tokens_seen": 17490576,
      "step": 30145
    },
    {
      "epoch": 4.490616621983914,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.028605144660374157,
      "loss": 0.7855,
      "num_input_tokens_seen": 17493488,
      "step": 30150
    },
    {
      "epoch": 4.491361334524873,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.028604323527425268,
      "loss": 0.7929,
      "num_input_tokens_seen": 17496240,
      "step": 30155
    },
    {
      "epoch": 4.492106047065833,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.02860350216464515,
      "loss": 0.782,
      "num_input_tokens_seen": 17498800,
      "step": 30160
    },
    {
      "epoch": 4.4928507596067915,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.028602680572047696,
      "loss": 0.7869,
      "num_input_tokens_seen": 17501456,
      "step": 30165
    },
    {
      "epoch": 4.493595472147751,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.02860185874964678,
      "loss": 0.8151,
      "num_input_tokens_seen": 17504368,
      "step": 30170
    },
    {
      "epoch": 4.49434018468871,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02860103669745628,
      "loss": 0.7877,
      "num_input_tokens_seen": 17507536,
      "step": 30175
    },
    {
      "epoch": 4.4950848972296695,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02860021441549009,
      "loss": 0.8154,
      "num_input_tokens_seen": 17510768,
      "step": 30180
    },
    {
      "epoch": 4.495829609770628,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028599391903762095,
      "loss": 0.8078,
      "num_input_tokens_seen": 17513744,
      "step": 30185
    },
    {
      "epoch": 4.496574322311588,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.028598569162286203,
      "loss": 0.7986,
      "num_input_tokens_seen": 17516976,
      "step": 30190
    },
    {
      "epoch": 4.497319034852547,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.0285977461910763,
      "loss": 0.8207,
      "num_input_tokens_seen": 17520112,
      "step": 30195
    },
    {
      "epoch": 4.498063747393506,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.0285969229901463,
      "loss": 0.802,
      "num_input_tokens_seen": 17522832,
      "step": 30200
    },
    {
      "epoch": 4.498808459934465,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.028596099559510105,
      "loss": 0.7754,
      "num_input_tokens_seen": 17525744,
      "step": 30205
    },
    {
      "epoch": 4.499553172475425,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02859527589918163,
      "loss": 0.7869,
      "num_input_tokens_seen": 17528464,
      "step": 30210
    },
    {
      "epoch": 4.5002978850163835,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02859445200917478,
      "loss": 0.7848,
      "num_input_tokens_seen": 17531408,
      "step": 30215
    },
    {
      "epoch": 4.501042597557343,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.028593627889503478,
      "loss": 0.7982,
      "num_input_tokens_seen": 17534224,
      "step": 30220
    },
    {
      "epoch": 4.501787310098302,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.028592803540181653,
      "loss": 0.8071,
      "num_input_tokens_seen": 17537200,
      "step": 30225
    },
    {
      "epoch": 4.5025320226392616,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.028591978961223227,
      "loss": 0.8049,
      "num_input_tokens_seen": 17540272,
      "step": 30230
    },
    {
      "epoch": 4.50327673518022,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028591154152642132,
      "loss": 0.8054,
      "num_input_tokens_seen": 17543280,
      "step": 30235
    },
    {
      "epoch": 4.50402144772118,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.0285903291144523,
      "loss": 0.8058,
      "num_input_tokens_seen": 17546288,
      "step": 30240
    },
    {
      "epoch": 4.504766160262139,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02858950384666767,
      "loss": 0.7946,
      "num_input_tokens_seen": 17549520,
      "step": 30245
    },
    {
      "epoch": 4.505510872803098,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02858867834930218,
      "loss": 0.7838,
      "num_input_tokens_seen": 17552368,
      "step": 30250
    },
    {
      "epoch": 4.506255585344057,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.028587852622369788,
      "loss": 0.8291,
      "num_input_tokens_seen": 17555312,
      "step": 30255
    },
    {
      "epoch": 4.507000297885017,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.02858702666588443,
      "loss": 0.8086,
      "num_input_tokens_seen": 17558416,
      "step": 30260
    },
    {
      "epoch": 4.5077450104259755,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.028586200479860067,
      "loss": 0.827,
      "num_input_tokens_seen": 17561328,
      "step": 30265
    },
    {
      "epoch": 4.508489722966935,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.028585374064310654,
      "loss": 0.7812,
      "num_input_tokens_seen": 17564432,
      "step": 30270
    },
    {
      "epoch": 4.509234435507894,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.028584547419250157,
      "loss": 0.796,
      "num_input_tokens_seen": 17567376,
      "step": 30275
    },
    {
      "epoch": 4.509979148048853,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02858372054469254,
      "loss": 0.8057,
      "num_input_tokens_seen": 17570288,
      "step": 30280
    },
    {
      "epoch": 4.510723860589812,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.028582893440651767,
      "loss": 0.8202,
      "num_input_tokens_seen": 17573456,
      "step": 30285
    },
    {
      "epoch": 4.511468573130772,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.028582066107141816,
      "loss": 0.7792,
      "num_input_tokens_seen": 17576176,
      "step": 30290
    },
    {
      "epoch": 4.512213285671731,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02858123854417666,
      "loss": 0.777,
      "num_input_tokens_seen": 17578800,
      "step": 30295
    },
    {
      "epoch": 4.5129579982126895,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.028580410751770282,
      "loss": 0.7737,
      "num_input_tokens_seen": 17581712,
      "step": 30300
    },
    {
      "epoch": 4.513702710753649,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.028579582729936666,
      "loss": 0.8357,
      "num_input_tokens_seen": 17584688,
      "step": 30305
    },
    {
      "epoch": 4.514447423294608,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.028578754478689808,
      "loss": 0.7916,
      "num_input_tokens_seen": 17587504,
      "step": 30310
    },
    {
      "epoch": 4.5151921358355676,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028577925998043686,
      "loss": 0.7901,
      "num_input_tokens_seen": 17590128,
      "step": 30315
    },
    {
      "epoch": 4.515936848376526,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.028577097288012308,
      "loss": 0.8093,
      "num_input_tokens_seen": 17592976,
      "step": 30320
    },
    {
      "epoch": 4.516681560917486,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02857626834860967,
      "loss": 0.802,
      "num_input_tokens_seen": 17595984,
      "step": 30325
    },
    {
      "epoch": 4.517426273458445,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.028575439179849775,
      "loss": 0.8222,
      "num_input_tokens_seen": 17598832,
      "step": 30330
    },
    {
      "epoch": 4.518170985999404,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.028574609781746637,
      "loss": 0.7898,
      "num_input_tokens_seen": 17601552,
      "step": 30335
    },
    {
      "epoch": 4.518915698540363,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02857378015431426,
      "loss": 0.8201,
      "num_input_tokens_seen": 17604592,
      "step": 30340
    },
    {
      "epoch": 4.519660411081323,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.028572950297566662,
      "loss": 0.8155,
      "num_input_tokens_seen": 17607536,
      "step": 30345
    },
    {
      "epoch": 4.5204051236222815,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.028572120211517865,
      "loss": 0.7808,
      "num_input_tokens_seen": 17610672,
      "step": 30350
    },
    {
      "epoch": 4.521149836163241,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.028571289896181892,
      "loss": 0.7976,
      "num_input_tokens_seen": 17613520,
      "step": 30355
    },
    {
      "epoch": 4.5218945487042,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.02857045935157277,
      "loss": 0.8259,
      "num_input_tokens_seen": 17616624,
      "step": 30360
    },
    {
      "epoch": 4.52263926124516,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.02856962857770453,
      "loss": 0.8207,
      "num_input_tokens_seen": 17619792,
      "step": 30365
    },
    {
      "epoch": 4.523383973786118,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0285687975745912,
      "loss": 0.8173,
      "num_input_tokens_seen": 17622992,
      "step": 30370
    },
    {
      "epoch": 4.524128686327078,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.028567966342246832,
      "loss": 0.7878,
      "num_input_tokens_seen": 17625840,
      "step": 30375
    },
    {
      "epoch": 4.524873398868037,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.028567134880685463,
      "loss": 0.8064,
      "num_input_tokens_seen": 17628496,
      "step": 30380
    },
    {
      "epoch": 4.525618111408996,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02856630318992114,
      "loss": 0.8006,
      "num_input_tokens_seen": 17631248,
      "step": 30385
    },
    {
      "epoch": 4.526362823949955,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.028565471269967908,
      "loss": 0.8231,
      "num_input_tokens_seen": 17634032,
      "step": 30390
    },
    {
      "epoch": 4.527107536490915,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.028564639120839827,
      "loss": 0.7963,
      "num_input_tokens_seen": 17637328,
      "step": 30395
    },
    {
      "epoch": 4.5278522490318736,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.028563806742550955,
      "loss": 0.8014,
      "num_input_tokens_seen": 17640432,
      "step": 30400
    },
    {
      "epoch": 4.528596961572833,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.028562974135115358,
      "loss": 0.8045,
      "num_input_tokens_seen": 17643152,
      "step": 30405
    },
    {
      "epoch": 4.529341674113792,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.028562141298547093,
      "loss": 0.8,
      "num_input_tokens_seen": 17646000,
      "step": 30410
    },
    {
      "epoch": 4.530086386654752,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028561308232860235,
      "loss": 0.8149,
      "num_input_tokens_seen": 17648880,
      "step": 30415
    },
    {
      "epoch": 4.53083109919571,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02856047493806886,
      "loss": 0.7717,
      "num_input_tokens_seen": 17651632,
      "step": 30420
    },
    {
      "epoch": 4.53157581173667,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.028559641414187042,
      "loss": 0.7844,
      "num_input_tokens_seen": 17654448,
      "step": 30425
    },
    {
      "epoch": 4.532320524277629,
      "grad_norm": 0.017578125,
      "learning_rate": 0.028558807661228865,
      "loss": 0.8059,
      "num_input_tokens_seen": 17657680,
      "step": 30430
    },
    {
      "epoch": 4.533065236818588,
      "grad_norm": 0.021484375,
      "learning_rate": 0.028557973679208412,
      "loss": 0.815,
      "num_input_tokens_seen": 17660528,
      "step": 30435
    },
    {
      "epoch": 4.533809949359547,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.028557139468139772,
      "loss": 0.8027,
      "num_input_tokens_seen": 17663600,
      "step": 30440
    },
    {
      "epoch": 4.534554661900506,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02855630502803704,
      "loss": 0.7953,
      "num_input_tokens_seen": 17666480,
      "step": 30445
    },
    {
      "epoch": 4.535299374441466,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.028555470358914314,
      "loss": 0.817,
      "num_input_tokens_seen": 17669520,
      "step": 30450
    },
    {
      "epoch": 4.536044086982425,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.028554635460785693,
      "loss": 0.7987,
      "num_input_tokens_seen": 17672208,
      "step": 30455
    },
    {
      "epoch": 4.536788799523384,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.028553800333665284,
      "loss": 0.8005,
      "num_input_tokens_seen": 17675152,
      "step": 30460
    },
    {
      "epoch": 4.537533512064343,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.02855296497756719,
      "loss": 0.8007,
      "num_input_tokens_seen": 17678448,
      "step": 30465
    },
    {
      "epoch": 4.538278224605302,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.028552129392505527,
      "loss": 0.7935,
      "num_input_tokens_seen": 17681488,
      "step": 30470
    },
    {
      "epoch": 4.539022937146262,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.028551293578494414,
      "loss": 0.8112,
      "num_input_tokens_seen": 17684432,
      "step": 30475
    },
    {
      "epoch": 4.539767649687221,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028550457535547972,
      "loss": 0.805,
      "num_input_tokens_seen": 17687280,
      "step": 30480
    },
    {
      "epoch": 4.5405123622281796,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.028549621263680318,
      "loss": 0.8167,
      "num_input_tokens_seen": 17690160,
      "step": 30485
    },
    {
      "epoch": 4.541257074769139,
      "grad_norm": 0.01080322265625,
      "learning_rate": 0.028548784762905583,
      "loss": 0.7753,
      "num_input_tokens_seen": 17693008,
      "step": 30490
    },
    {
      "epoch": 4.542001787310098,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.028547948033237902,
      "loss": 0.7862,
      "num_input_tokens_seen": 17695952,
      "step": 30495
    },
    {
      "epoch": 4.542746499851058,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02854711107469141,
      "loss": 0.809,
      "num_input_tokens_seen": 17698736,
      "step": 30500
    },
    {
      "epoch": 4.543491212392016,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.028546273887280243,
      "loss": 0.7808,
      "num_input_tokens_seen": 17701552,
      "step": 30505
    },
    {
      "epoch": 4.544235924932976,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.028545436471018548,
      "loss": 0.8111,
      "num_input_tokens_seen": 17704528,
      "step": 30510
    },
    {
      "epoch": 4.544980637473935,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.028544598825920465,
      "loss": 0.7816,
      "num_input_tokens_seen": 17707408,
      "step": 30515
    },
    {
      "epoch": 4.545725350014894,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.028543760952000157,
      "loss": 0.8389,
      "num_input_tokens_seen": 17710128,
      "step": 30520
    },
    {
      "epoch": 4.546470062555853,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02854292284927177,
      "loss": 0.7794,
      "num_input_tokens_seen": 17712976,
      "step": 30525
    },
    {
      "epoch": 4.547214775096813,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028542084517749467,
      "loss": 0.811,
      "num_input_tokens_seen": 17715984,
      "step": 30530
    },
    {
      "epoch": 4.547959487637772,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.02854124595744741,
      "loss": 0.8053,
      "num_input_tokens_seen": 17718928,
      "step": 30535
    },
    {
      "epoch": 4.548704200178731,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.028540407168379765,
      "loss": 0.7817,
      "num_input_tokens_seen": 17721680,
      "step": 30540
    },
    {
      "epoch": 4.54944891271969,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028539568150560704,
      "loss": 0.8037,
      "num_input_tokens_seen": 17724400,
      "step": 30545
    },
    {
      "epoch": 4.55019362526065,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.0285387289040044,
      "loss": 0.7961,
      "num_input_tokens_seen": 17727280,
      "step": 30550
    },
    {
      "epoch": 4.550938337801608,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02853788942872503,
      "loss": 0.8266,
      "num_input_tokens_seen": 17730448,
      "step": 30555
    },
    {
      "epoch": 4.551683050342568,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.028537049724736776,
      "loss": 0.7899,
      "num_input_tokens_seen": 17733328,
      "step": 30560
    },
    {
      "epoch": 4.552427762883527,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02853620979205383,
      "loss": 0.815,
      "num_input_tokens_seen": 17736144,
      "step": 30565
    },
    {
      "epoch": 4.553172475424486,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02853536963069037,
      "loss": 0.8089,
      "num_input_tokens_seen": 17739152,
      "step": 30570
    },
    {
      "epoch": 4.553917187965445,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.028534529240660602,
      "loss": 0.7972,
      "num_input_tokens_seen": 17742128,
      "step": 30575
    },
    {
      "epoch": 4.554661900506405,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.028533688621978717,
      "loss": 0.797,
      "num_input_tokens_seen": 17744976,
      "step": 30580
    },
    {
      "epoch": 4.555406613047364,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.028532847774658917,
      "loss": 0.8091,
      "num_input_tokens_seen": 17747984,
      "step": 30585
    },
    {
      "epoch": 4.556151325588323,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.02853200669871541,
      "loss": 0.801,
      "num_input_tokens_seen": 17750992,
      "step": 30590
    },
    {
      "epoch": 4.556896038129282,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0285311653941624,
      "loss": 0.8026,
      "num_input_tokens_seen": 17753968,
      "step": 30595
    },
    {
      "epoch": 4.557640750670242,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.028530323861014107,
      "loss": 0.7823,
      "num_input_tokens_seen": 17756784,
      "step": 30600
    },
    {
      "epoch": 4.5583854632112,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.028529482099284742,
      "loss": 0.8356,
      "num_input_tokens_seen": 17759984,
      "step": 30605
    },
    {
      "epoch": 4.559130175752159,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.028528640108988527,
      "loss": 0.7953,
      "num_input_tokens_seen": 17763280,
      "step": 30610
    },
    {
      "epoch": 4.559874888293119,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02852779789013969,
      "loss": 0.7899,
      "num_input_tokens_seen": 17765872,
      "step": 30615
    },
    {
      "epoch": 4.5606196008340785,
      "grad_norm": 0.021484375,
      "learning_rate": 0.028526955442752456,
      "loss": 0.7978,
      "num_input_tokens_seen": 17768816,
      "step": 30620
    },
    {
      "epoch": 4.561364313375037,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.028526112766841055,
      "loss": 0.7827,
      "num_input_tokens_seen": 17771920,
      "step": 30625
    },
    {
      "epoch": 4.562109025915996,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02852526986241973,
      "loss": 0.79,
      "num_input_tokens_seen": 17774640,
      "step": 30630
    },
    {
      "epoch": 4.562853738456956,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.028524426729502714,
      "loss": 0.8062,
      "num_input_tokens_seen": 17777264,
      "step": 30635
    },
    {
      "epoch": 4.563598450997915,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.028523583368104254,
      "loss": 0.8222,
      "num_input_tokens_seen": 17780432,
      "step": 30640
    },
    {
      "epoch": 4.564343163538874,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.028522739778238604,
      "loss": 0.8046,
      "num_input_tokens_seen": 17783312,
      "step": 30645
    },
    {
      "epoch": 4.565087876079833,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.02852189595992,
      "loss": 0.8094,
      "num_input_tokens_seen": 17786064,
      "step": 30650
    },
    {
      "epoch": 4.565832588620792,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.02852105191316271,
      "loss": 0.807,
      "num_input_tokens_seen": 17788912,
      "step": 30655
    },
    {
      "epoch": 4.566577301161751,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02852020763798099,
      "loss": 0.7752,
      "num_input_tokens_seen": 17791376,
      "step": 30660
    },
    {
      "epoch": 4.567322013702711,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.028519363134389106,
      "loss": 0.8466,
      "num_input_tokens_seen": 17794320,
      "step": 30665
    },
    {
      "epoch": 4.56806672624367,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02851851840240132,
      "loss": 0.8426,
      "num_input_tokens_seen": 17797488,
      "step": 30670
    },
    {
      "epoch": 4.568811438784629,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.028517673442031908,
      "loss": 0.7943,
      "num_input_tokens_seen": 17800208,
      "step": 30675
    },
    {
      "epoch": 4.569556151325588,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02851682825329514,
      "loss": 0.7839,
      "num_input_tokens_seen": 17802960,
      "step": 30680
    },
    {
      "epoch": 4.570300863866548,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.028515982836205298,
      "loss": 0.7995,
      "num_input_tokens_seen": 17805488,
      "step": 30685
    },
    {
      "epoch": 4.571045576407506,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.02851513719077666,
      "loss": 0.823,
      "num_input_tokens_seen": 17808208,
      "step": 30690
    },
    {
      "epoch": 4.571790288948466,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.028514291317023516,
      "loss": 0.8132,
      "num_input_tokens_seen": 17810896,
      "step": 30695
    },
    {
      "epoch": 4.572535001489425,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.028513445214960156,
      "loss": 0.7962,
      "num_input_tokens_seen": 17813648,
      "step": 30700
    },
    {
      "epoch": 4.5732797140303845,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.028512598884600875,
      "loss": 0.8187,
      "num_input_tokens_seen": 17816272,
      "step": 30705
    },
    {
      "epoch": 4.574024426571343,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.028511752325959965,
      "loss": 0.8145,
      "num_input_tokens_seen": 17819248,
      "step": 30710
    },
    {
      "epoch": 4.574769139112303,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02851090553905174,
      "loss": 0.7957,
      "num_input_tokens_seen": 17822064,
      "step": 30715
    },
    {
      "epoch": 4.575513851653262,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02851005852389049,
      "loss": 0.8134,
      "num_input_tokens_seen": 17824816,
      "step": 30720
    },
    {
      "epoch": 4.576258564194221,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02850921128049054,
      "loss": 0.813,
      "num_input_tokens_seen": 17827600,
      "step": 30725
    },
    {
      "epoch": 4.57700327673518,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02850836380886619,
      "loss": 0.8114,
      "num_input_tokens_seen": 17830768,
      "step": 30730
    },
    {
      "epoch": 4.57774798927614,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.028507516109031763,
      "loss": 0.8131,
      "num_input_tokens_seen": 17833616,
      "step": 30735
    },
    {
      "epoch": 4.578492701817098,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.028506668181001577,
      "loss": 0.7904,
      "num_input_tokens_seen": 17836464,
      "step": 30740
    },
    {
      "epoch": 4.579237414358058,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.02850582002478997,
      "loss": 0.8038,
      "num_input_tokens_seen": 17839280,
      "step": 30745
    },
    {
      "epoch": 4.579982126899017,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02850497164041125,
      "loss": 0.8007,
      "num_input_tokens_seen": 17842768,
      "step": 30750
    },
    {
      "epoch": 4.5807268394399765,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02850412302787976,
      "loss": 0.8075,
      "num_input_tokens_seen": 17845488,
      "step": 30755
    },
    {
      "epoch": 4.581471551980935,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02850327418720984,
      "loss": 0.7873,
      "num_input_tokens_seen": 17848304,
      "step": 30760
    },
    {
      "epoch": 4.582216264521895,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02850242511841583,
      "loss": 0.82,
      "num_input_tokens_seen": 17851312,
      "step": 30765
    },
    {
      "epoch": 4.582960977062854,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.028501575821512068,
      "loss": 0.8098,
      "num_input_tokens_seen": 17854000,
      "step": 30770
    },
    {
      "epoch": 4.583705689603813,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.0285007262965129,
      "loss": 0.8209,
      "num_input_tokens_seen": 17856848,
      "step": 30775
    },
    {
      "epoch": 4.584450402144772,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.02849987654343269,
      "loss": 0.7888,
      "num_input_tokens_seen": 17859760,
      "step": 30780
    },
    {
      "epoch": 4.585195114685732,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.028499026562285784,
      "loss": 0.7996,
      "num_input_tokens_seen": 17862448,
      "step": 30785
    },
    {
      "epoch": 4.5859398272266905,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.028498176353086546,
      "loss": 0.8158,
      "num_input_tokens_seen": 17865392,
      "step": 30790
    },
    {
      "epoch": 4.586684539767649,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.028497325915849334,
      "loss": 0.8002,
      "num_input_tokens_seen": 17868048,
      "step": 30795
    },
    {
      "epoch": 4.587429252308609,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028496475250588522,
      "loss": 0.7907,
      "num_input_tokens_seen": 17870768,
      "step": 30800
    },
    {
      "epoch": 4.5881739648495685,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.028495624357318474,
      "loss": 0.8201,
      "num_input_tokens_seen": 17873520,
      "step": 30805
    },
    {
      "epoch": 4.588918677390527,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.02849477323605357,
      "loss": 0.7987,
      "num_input_tokens_seen": 17876464,
      "step": 30810
    },
    {
      "epoch": 4.589663389931486,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02849392188680819,
      "loss": 0.7898,
      "num_input_tokens_seen": 17879152,
      "step": 30815
    },
    {
      "epoch": 4.590408102472446,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.028493070309596708,
      "loss": 0.8048,
      "num_input_tokens_seen": 17882000,
      "step": 30820
    },
    {
      "epoch": 4.591152815013404,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.028492218504433526,
      "loss": 0.8125,
      "num_input_tokens_seen": 17884912,
      "step": 30825
    },
    {
      "epoch": 4.591897527554364,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02849136647133302,
      "loss": 0.7947,
      "num_input_tokens_seen": 17887632,
      "step": 30830
    },
    {
      "epoch": 4.592642240095323,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02849051421030959,
      "loss": 0.7989,
      "num_input_tokens_seen": 17890480,
      "step": 30835
    },
    {
      "epoch": 4.5933869526362825,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02848966172137763,
      "loss": 0.821,
      "num_input_tokens_seen": 17893392,
      "step": 30840
    },
    {
      "epoch": 4.594131665177241,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02848880900455155,
      "loss": 0.8013,
      "num_input_tokens_seen": 17896144,
      "step": 30845
    },
    {
      "epoch": 4.594876377718201,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02848795605984575,
      "loss": 0.8022,
      "num_input_tokens_seen": 17899248,
      "step": 30850
    },
    {
      "epoch": 4.59562109025916,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.028487102887274646,
      "loss": 0.7941,
      "num_input_tokens_seen": 17902000,
      "step": 30855
    },
    {
      "epoch": 4.596365802800119,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02848624948685264,
      "loss": 0.7888,
      "num_input_tokens_seen": 17904720,
      "step": 30860
    },
    {
      "epoch": 4.597110515341078,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.028485395858594158,
      "loss": 0.8009,
      "num_input_tokens_seen": 17907472,
      "step": 30865
    },
    {
      "epoch": 4.597855227882038,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.028484542002513616,
      "loss": 0.8023,
      "num_input_tokens_seen": 17910032,
      "step": 30870
    },
    {
      "epoch": 4.5985999404229965,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.028483687918625447,
      "loss": 0.7837,
      "num_input_tokens_seen": 17913040,
      "step": 30875
    },
    {
      "epoch": 4.599344652963956,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.028482833606944074,
      "loss": 0.8024,
      "num_input_tokens_seen": 17915984,
      "step": 30880
    },
    {
      "epoch": 4.600089365504915,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028481979067483926,
      "loss": 0.7989,
      "num_input_tokens_seen": 17918640,
      "step": 30885
    },
    {
      "epoch": 4.6008340780458745,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.028481124300259446,
      "loss": 0.8036,
      "num_input_tokens_seen": 17921808,
      "step": 30890
    },
    {
      "epoch": 4.601578790586833,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02848026930528507,
      "loss": 0.7996,
      "num_input_tokens_seen": 17924816,
      "step": 30895
    },
    {
      "epoch": 4.602323503127793,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028479414082575254,
      "loss": 0.7745,
      "num_input_tokens_seen": 17927664,
      "step": 30900
    },
    {
      "epoch": 4.603068215668752,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028478558632144428,
      "loss": 0.7827,
      "num_input_tokens_seen": 17930288,
      "step": 30905
    },
    {
      "epoch": 4.603812928209711,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.028477702954007058,
      "loss": 0.776,
      "num_input_tokens_seen": 17933072,
      "step": 30910
    },
    {
      "epoch": 4.60455764075067,
      "grad_norm": 0.01953125,
      "learning_rate": 0.028476847048177593,
      "loss": 0.8337,
      "num_input_tokens_seen": 17935856,
      "step": 30915
    },
    {
      "epoch": 4.60530235329163,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.028475990914670492,
      "loss": 0.7831,
      "num_input_tokens_seen": 17938800,
      "step": 30920
    },
    {
      "epoch": 4.6060470658325885,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.028475134553500224,
      "loss": 0.824,
      "num_input_tokens_seen": 17942032,
      "step": 30925
    },
    {
      "epoch": 4.606791778373548,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028474277964681254,
      "loss": 0.8151,
      "num_input_tokens_seen": 17945072,
      "step": 30930
    },
    {
      "epoch": 4.607536490914507,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.028473421148228054,
      "loss": 0.8206,
      "num_input_tokens_seen": 17948048,
      "step": 30935
    },
    {
      "epoch": 4.6082812034554665,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.02847256410415509,
      "loss": 0.8152,
      "num_input_tokens_seen": 17951216,
      "step": 30940
    },
    {
      "epoch": 4.609025915996425,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.028471706832476854,
      "loss": 0.8097,
      "num_input_tokens_seen": 17954160,
      "step": 30945
    },
    {
      "epoch": 4.609770628537385,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02847084933320782,
      "loss": 0.7989,
      "num_input_tokens_seen": 17957104,
      "step": 30950
    },
    {
      "epoch": 4.610515341078344,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.028469991606362478,
      "loss": 0.7996,
      "num_input_tokens_seen": 17960080,
      "step": 30955
    },
    {
      "epoch": 4.6112600536193025,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02846913365195532,
      "loss": 0.7943,
      "num_input_tokens_seen": 17963152,
      "step": 30960
    },
    {
      "epoch": 4.612004766160262,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.028468275470000838,
      "loss": 0.8041,
      "num_input_tokens_seen": 17966064,
      "step": 30965
    },
    {
      "epoch": 4.612749478701222,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.028467417060513533,
      "loss": 0.8024,
      "num_input_tokens_seen": 17969040,
      "step": 30970
    },
    {
      "epoch": 4.6134941912421805,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.028466558423507902,
      "loss": 0.789,
      "num_input_tokens_seen": 17972208,
      "step": 30975
    },
    {
      "epoch": 4.614238903783139,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028465699558998452,
      "loss": 0.8107,
      "num_input_tokens_seen": 17974896,
      "step": 30980
    },
    {
      "epoch": 4.614983616324099,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028464840466999694,
      "loss": 0.8001,
      "num_input_tokens_seen": 17977776,
      "step": 30985
    },
    {
      "epoch": 4.615728328865059,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.028463981147526143,
      "loss": 0.8067,
      "num_input_tokens_seen": 17980368,
      "step": 30990
    },
    {
      "epoch": 4.616473041406017,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.028463121600592313,
      "loss": 0.823,
      "num_input_tokens_seen": 17983280,
      "step": 30995
    },
    {
      "epoch": 4.617217753946976,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.028462261826212726,
      "loss": 0.8116,
      "num_input_tokens_seen": 17986384,
      "step": 31000
    },
    {
      "epoch": 4.617962466487936,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.028461401824401912,
      "loss": 0.7903,
      "num_input_tokens_seen": 17989328,
      "step": 31005
    },
    {
      "epoch": 4.6187071790288945,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.028460541595174395,
      "loss": 0.7838,
      "num_input_tokens_seen": 17992112,
      "step": 31010
    },
    {
      "epoch": 4.619451891569854,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028459681138544704,
      "loss": 0.7948,
      "num_input_tokens_seen": 17995024,
      "step": 31015
    },
    {
      "epoch": 4.620196604110813,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.028458820454527384,
      "loss": 0.7971,
      "num_input_tokens_seen": 17998160,
      "step": 31020
    },
    {
      "epoch": 4.6209413166517725,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02845795954313697,
      "loss": 0.7919,
      "num_input_tokens_seen": 18000496,
      "step": 31025
    },
    {
      "epoch": 4.621686029192731,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.028457098404388006,
      "loss": 0.8128,
      "num_input_tokens_seen": 18003312,
      "step": 31030
    },
    {
      "epoch": 4.622430741733691,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.028456237038295043,
      "loss": 0.8273,
      "num_input_tokens_seen": 18006160,
      "step": 31035
    },
    {
      "epoch": 4.62317545427465,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.028455375444872633,
      "loss": 0.8074,
      "num_input_tokens_seen": 18008976,
      "step": 31040
    },
    {
      "epoch": 4.623920166815609,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.028454513624135327,
      "loss": 0.7967,
      "num_input_tokens_seen": 18011792,
      "step": 31045
    },
    {
      "epoch": 4.624664879356568,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.028453651576097694,
      "loss": 0.7993,
      "num_input_tokens_seen": 18014992,
      "step": 31050
    },
    {
      "epoch": 4.625409591897528,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.028452789300774287,
      "loss": 0.8082,
      "num_input_tokens_seen": 18017904,
      "step": 31055
    },
    {
      "epoch": 4.6261543044384865,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.028451926798179678,
      "loss": 0.8092,
      "num_input_tokens_seen": 18020848,
      "step": 31060
    },
    {
      "epoch": 4.626899016979446,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.028451064068328438,
      "loss": 0.7959,
      "num_input_tokens_seen": 18023728,
      "step": 31065
    },
    {
      "epoch": 4.627643729520405,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02845020111123514,
      "loss": 0.8086,
      "num_input_tokens_seen": 18026480,
      "step": 31070
    },
    {
      "epoch": 4.628388442061365,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.028449337926914366,
      "loss": 0.8071,
      "num_input_tokens_seen": 18029104,
      "step": 31075
    },
    {
      "epoch": 4.629133154602323,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.028448474515380698,
      "loss": 0.8273,
      "num_input_tokens_seen": 18032016,
      "step": 31080
    },
    {
      "epoch": 4.629877867143283,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.028447610876648717,
      "loss": 0.8037,
      "num_input_tokens_seen": 18034832,
      "step": 31085
    },
    {
      "epoch": 4.630622579684242,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02844674701073302,
      "loss": 0.8016,
      "num_input_tokens_seen": 18037616,
      "step": 31090
    },
    {
      "epoch": 4.631367292225201,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.028445882917648203,
      "loss": 0.8052,
      "num_input_tokens_seen": 18040976,
      "step": 31095
    },
    {
      "epoch": 4.63211200476616,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.028445018597408857,
      "loss": 0.7901,
      "num_input_tokens_seen": 18044592,
      "step": 31100
    },
    {
      "epoch": 4.63285671730712,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.028444154050029584,
      "loss": 0.7956,
      "num_input_tokens_seen": 18047504,
      "step": 31105
    },
    {
      "epoch": 4.6336014298480785,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.028443289275525,
      "loss": 0.8323,
      "num_input_tokens_seen": 18050352,
      "step": 31110
    },
    {
      "epoch": 4.634346142389038,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.028442424273909694,
      "loss": 0.8215,
      "num_input_tokens_seen": 18053136,
      "step": 31115
    },
    {
      "epoch": 4.635090854929997,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.028441559045198305,
      "loss": 0.8121,
      "num_input_tokens_seen": 18056208,
      "step": 31120
    },
    {
      "epoch": 4.635835567470957,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02844069358940543,
      "loss": 0.8086,
      "num_input_tokens_seen": 18059344,
      "step": 31125
    },
    {
      "epoch": 4.636580280011915,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.0284398279065457,
      "loss": 0.8138,
      "num_input_tokens_seen": 18062224,
      "step": 31130
    },
    {
      "epoch": 4.637324992552875,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.028438961996633736,
      "loss": 0.7965,
      "num_input_tokens_seen": 18065008,
      "step": 31135
    },
    {
      "epoch": 4.638069705093834,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.028438095859684168,
      "loss": 0.8149,
      "num_input_tokens_seen": 18067536,
      "step": 31140
    },
    {
      "epoch": 4.6388144176347925,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02843722949571163,
      "loss": 0.8109,
      "num_input_tokens_seen": 18070608,
      "step": 31145
    },
    {
      "epoch": 4.639559130175752,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02843636290473076,
      "loss": 0.7888,
      "num_input_tokens_seen": 18073488,
      "step": 31150
    },
    {
      "epoch": 4.640303842716712,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.028435496086756187,
      "loss": 0.8113,
      "num_input_tokens_seen": 18076336,
      "step": 31155
    },
    {
      "epoch": 4.641048555257671,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.028434629041802564,
      "loss": 0.8079,
      "num_input_tokens_seen": 18079088,
      "step": 31160
    },
    {
      "epoch": 4.641793267798629,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.02843376176988454,
      "loss": 0.8183,
      "num_input_tokens_seen": 18082224,
      "step": 31165
    },
    {
      "epoch": 4.642537980339589,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.028432894271016765,
      "loss": 0.7915,
      "num_input_tokens_seen": 18085200,
      "step": 31170
    },
    {
      "epoch": 4.643282692880548,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.02843202654521389,
      "loss": 0.791,
      "num_input_tokens_seen": 18087952,
      "step": 31175
    },
    {
      "epoch": 4.644027405421507,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.028431158592490586,
      "loss": 0.8124,
      "num_input_tokens_seen": 18090992,
      "step": 31180
    },
    {
      "epoch": 4.644772117962466,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.028430290412861502,
      "loss": 0.7997,
      "num_input_tokens_seen": 18093872,
      "step": 31185
    },
    {
      "epoch": 4.645516830503426,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.028429422006341312,
      "loss": 0.7933,
      "num_input_tokens_seen": 18096624,
      "step": 31190
    },
    {
      "epoch": 4.6462615430443845,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.028428553372944684,
      "loss": 0.7904,
      "num_input_tokens_seen": 18099504,
      "step": 31195
    },
    {
      "epoch": 4.647006255585344,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.0284276845126863,
      "loss": 0.8002,
      "num_input_tokens_seen": 18102448,
      "step": 31200
    },
    {
      "epoch": 4.647750968126303,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02842681542558083,
      "loss": 0.7927,
      "num_input_tokens_seen": 18105488,
      "step": 31205
    },
    {
      "epoch": 4.648495680667263,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.028425946111642963,
      "loss": 0.7972,
      "num_input_tokens_seen": 18108336,
      "step": 31210
    },
    {
      "epoch": 4.649240393208221,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.028425076570887383,
      "loss": 0.7844,
      "num_input_tokens_seen": 18111248,
      "step": 31215
    },
    {
      "epoch": 4.649985105749181,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.028424206803328776,
      "loss": 0.7889,
      "num_input_tokens_seen": 18114416,
      "step": 31220
    },
    {
      "epoch": 4.65072981829014,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.028423336808981837,
      "loss": 0.783,
      "num_input_tokens_seen": 18117360,
      "step": 31225
    },
    {
      "epoch": 4.651474530831099,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.028422466587861267,
      "loss": 0.7989,
      "num_input_tokens_seen": 18120208,
      "step": 31230
    },
    {
      "epoch": 4.652219243372058,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02842159613998177,
      "loss": 0.7973,
      "num_input_tokens_seen": 18122896,
      "step": 31235
    },
    {
      "epoch": 4.652963955913018,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.028420725465358045,
      "loss": 0.8019,
      "num_input_tokens_seen": 18125680,
      "step": 31240
    },
    {
      "epoch": 4.653708668453977,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.0284198545640048,
      "loss": 0.815,
      "num_input_tokens_seen": 18128688,
      "step": 31245
    },
    {
      "epoch": 4.654453380994936,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.028418983435936754,
      "loss": 0.7937,
      "num_input_tokens_seen": 18131536,
      "step": 31250
    },
    {
      "epoch": 4.655198093535895,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02841811208116862,
      "loss": 0.8055,
      "num_input_tokens_seen": 18134256,
      "step": 31255
    },
    {
      "epoch": 4.655942806076855,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.028417240499715123,
      "loss": 0.8015,
      "num_input_tokens_seen": 18137040,
      "step": 31260
    },
    {
      "epoch": 4.656687518617813,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.028416368691590983,
      "loss": 0.7842,
      "num_input_tokens_seen": 18139920,
      "step": 31265
    },
    {
      "epoch": 4.657432231158773,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.028415496656810927,
      "loss": 0.8282,
      "num_input_tokens_seen": 18142800,
      "step": 31270
    },
    {
      "epoch": 4.658176943699732,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028414624395389693,
      "loss": 0.7818,
      "num_input_tokens_seen": 18145456,
      "step": 31275
    },
    {
      "epoch": 4.658921656240691,
      "grad_norm": 0.015625,
      "learning_rate": 0.02841375190734201,
      "loss": 0.8235,
      "num_input_tokens_seen": 18148528,
      "step": 31280
    },
    {
      "epoch": 4.65966636878165,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02841287919268263,
      "loss": 0.7916,
      "num_input_tokens_seen": 18151280,
      "step": 31285
    },
    {
      "epoch": 4.66041108132261,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02841200625142628,
      "loss": 0.8464,
      "num_input_tokens_seen": 18154096,
      "step": 31290
    },
    {
      "epoch": 4.661155793863569,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.02841113308358772,
      "loss": 0.8117,
      "num_input_tokens_seen": 18157040,
      "step": 31295
    },
    {
      "epoch": 4.661900506404528,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028410259689181695,
      "loss": 0.8213,
      "num_input_tokens_seen": 18159984,
      "step": 31300
    },
    {
      "epoch": 4.662645218945487,
      "grad_norm": 0.01953125,
      "learning_rate": 0.028409386068222964,
      "loss": 0.7912,
      "num_input_tokens_seen": 18162832,
      "step": 31305
    },
    {
      "epoch": 4.663389931486446,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02840851222072628,
      "loss": 0.7933,
      "num_input_tokens_seen": 18165872,
      "step": 31310
    },
    {
      "epoch": 4.664134644027405,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028407638146706417,
      "loss": 0.808,
      "num_input_tokens_seen": 18168784,
      "step": 31315
    },
    {
      "epoch": 4.664879356568365,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02840676384617813,
      "loss": 0.8088,
      "num_input_tokens_seen": 18171920,
      "step": 31320
    },
    {
      "epoch": 4.665624069109324,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02840588931915619,
      "loss": 0.8085,
      "num_input_tokens_seen": 18174576,
      "step": 31325
    },
    {
      "epoch": 4.666368781650283,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.028405014565655383,
      "loss": 0.7949,
      "num_input_tokens_seen": 18177648,
      "step": 31330
    },
    {
      "epoch": 4.667113494191242,
      "grad_norm": 0.02734375,
      "learning_rate": 0.028404139585690474,
      "loss": 0.7953,
      "num_input_tokens_seen": 18180688,
      "step": 31335
    },
    {
      "epoch": 4.667858206732202,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.028403264379276254,
      "loss": 0.8228,
      "num_input_tokens_seen": 18183440,
      "step": 31340
    },
    {
      "epoch": 4.668602919273161,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.0284023889464275,
      "loss": 0.7888,
      "num_input_tokens_seen": 18186384,
      "step": 31345
    },
    {
      "epoch": 4.669347631814119,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.028401513287159008,
      "loss": 0.8037,
      "num_input_tokens_seen": 18189168,
      "step": 31350
    },
    {
      "epoch": 4.670092344355079,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.028400637401485566,
      "loss": 0.8082,
      "num_input_tokens_seen": 18192400,
      "step": 31355
    },
    {
      "epoch": 4.670837056896038,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028399761289421985,
      "loss": 0.7956,
      "num_input_tokens_seen": 18195408,
      "step": 31360
    },
    {
      "epoch": 4.671581769436997,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02839888495098305,
      "loss": 0.7958,
      "num_input_tokens_seen": 18198480,
      "step": 31365
    },
    {
      "epoch": 4.672326481977956,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02839800838618357,
      "loss": 0.7996,
      "num_input_tokens_seen": 18201296,
      "step": 31370
    },
    {
      "epoch": 4.673071194518916,
      "grad_norm": 0.01422119140625,
      "learning_rate": 0.028397131595038354,
      "loss": 0.8124,
      "num_input_tokens_seen": 18204464,
      "step": 31375
    },
    {
      "epoch": 4.673815907059875,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02839625457756222,
      "loss": 0.7968,
      "num_input_tokens_seen": 18207312,
      "step": 31380
    },
    {
      "epoch": 4.674560619600834,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.028395377333769983,
      "loss": 0.7964,
      "num_input_tokens_seen": 18210160,
      "step": 31385
    },
    {
      "epoch": 4.675305332141793,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028394499863676453,
      "loss": 0.8232,
      "num_input_tokens_seen": 18213040,
      "step": 31390
    },
    {
      "epoch": 4.676050044682753,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02839362216729647,
      "loss": 0.8126,
      "num_input_tokens_seen": 18216208,
      "step": 31395
    },
    {
      "epoch": 4.676794757223711,
      "grad_norm": 0.0107421875,
      "learning_rate": 0.028392744244644847,
      "loss": 0.8005,
      "num_input_tokens_seen": 18218864,
      "step": 31400
    },
    {
      "epoch": 4.677539469764671,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02839186609573642,
      "loss": 0.7942,
      "num_input_tokens_seen": 18221680,
      "step": 31405
    },
    {
      "epoch": 4.67828418230563,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02839098772058603,
      "loss": 0.7931,
      "num_input_tokens_seen": 18224496,
      "step": 31410
    },
    {
      "epoch": 4.6790288948465895,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02839010911920851,
      "loss": 0.7776,
      "num_input_tokens_seen": 18227280,
      "step": 31415
    },
    {
      "epoch": 4.679773607387548,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02838923029161871,
      "loss": 0.7789,
      "num_input_tokens_seen": 18230064,
      "step": 31420
    },
    {
      "epoch": 4.680518319928508,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.02838835123783147,
      "loss": 0.7918,
      "num_input_tokens_seen": 18232880,
      "step": 31425
    },
    {
      "epoch": 4.681263032469467,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02838747195786164,
      "loss": 0.8441,
      "num_input_tokens_seen": 18235536,
      "step": 31430
    },
    {
      "epoch": 4.682007745010426,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.028386592451724084,
      "loss": 0.7848,
      "num_input_tokens_seen": 18238096,
      "step": 31435
    },
    {
      "epoch": 4.682752457551385,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.028385712719433652,
      "loss": 0.7864,
      "num_input_tokens_seen": 18240912,
      "step": 31440
    },
    {
      "epoch": 4.683497170092345,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.028384832761005205,
      "loss": 0.814,
      "num_input_tokens_seen": 18243984,
      "step": 31445
    },
    {
      "epoch": 4.684241882633303,
      "grad_norm": 0.02734375,
      "learning_rate": 0.028383952576453614,
      "loss": 0.8481,
      "num_input_tokens_seen": 18246992,
      "step": 31450
    },
    {
      "epoch": 4.684986595174263,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.02838307216579375,
      "loss": 0.798,
      "num_input_tokens_seen": 18249776,
      "step": 31455
    },
    {
      "epoch": 4.685731307715222,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.02838219152904048,
      "loss": 0.8152,
      "num_input_tokens_seen": 18252688,
      "step": 31460
    },
    {
      "epoch": 4.6864760202561815,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.028381310666208682,
      "loss": 0.8005,
      "num_input_tokens_seen": 18255696,
      "step": 31465
    },
    {
      "epoch": 4.68722073279714,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.028380429577313245,
      "loss": 0.8186,
      "num_input_tokens_seen": 18258672,
      "step": 31470
    },
    {
      "epoch": 4.687965445338099,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.028379548262369048,
      "loss": 0.7668,
      "num_input_tokens_seen": 18261424,
      "step": 31475
    },
    {
      "epoch": 4.688710157879059,
      "grad_norm": 0.01129150390625,
      "learning_rate": 0.028378666721390985,
      "loss": 0.7887,
      "num_input_tokens_seen": 18264336,
      "step": 31480
    },
    {
      "epoch": 4.689454870420018,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.028377784954393942,
      "loss": 0.7948,
      "num_input_tokens_seen": 18267504,
      "step": 31485
    },
    {
      "epoch": 4.690199582960977,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.028376902961392822,
      "loss": 0.8087,
      "num_input_tokens_seen": 18270384,
      "step": 31490
    },
    {
      "epoch": 4.690944295501936,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.02837602074240252,
      "loss": 0.8056,
      "num_input_tokens_seen": 18273424,
      "step": 31495
    },
    {
      "epoch": 4.6916890080428955,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02837513829743794,
      "loss": 0.7842,
      "num_input_tokens_seen": 18276400,
      "step": 31500
    },
    {
      "epoch": 4.692433720583855,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.028374255626513994,
      "loss": 0.8139,
      "num_input_tokens_seen": 18279216,
      "step": 31505
    },
    {
      "epoch": 4.693178433124814,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.02837337272964559,
      "loss": 0.8341,
      "num_input_tokens_seen": 18282256,
      "step": 31510
    },
    {
      "epoch": 4.693923145665773,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02837248960684765,
      "loss": 0.7846,
      "num_input_tokens_seen": 18285232,
      "step": 31515
    },
    {
      "epoch": 4.694667858206732,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02837160625813509,
      "loss": 0.7914,
      "num_input_tokens_seen": 18288496,
      "step": 31520
    },
    {
      "epoch": 4.695412570747691,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02837072268352283,
      "loss": 0.8165,
      "num_input_tokens_seen": 18291504,
      "step": 31525
    },
    {
      "epoch": 4.696157283288651,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.028369838883025796,
      "loss": 0.799,
      "num_input_tokens_seen": 18294128,
      "step": 31530
    },
    {
      "epoch": 4.696901995829609,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02836895485665893,
      "loss": 0.8018,
      "num_input_tokens_seen": 18297232,
      "step": 31535
    },
    {
      "epoch": 4.697646708370569,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.028368070604437155,
      "loss": 0.8123,
      "num_input_tokens_seen": 18300240,
      "step": 31540
    },
    {
      "epoch": 4.698391420911528,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.028367186126375413,
      "loss": 0.7912,
      "num_input_tokens_seen": 18302864,
      "step": 31545
    },
    {
      "epoch": 4.6991361334524875,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.028366301422488648,
      "loss": 0.8112,
      "num_input_tokens_seen": 18306032,
      "step": 31550
    },
    {
      "epoch": 4.699880845993446,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.02836541649279181,
      "loss": 0.8017,
      "num_input_tokens_seen": 18308880,
      "step": 31555
    },
    {
      "epoch": 4.700625558534406,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02836453133729984,
      "loss": 0.8079,
      "num_input_tokens_seen": 18311952,
      "step": 31560
    },
    {
      "epoch": 4.701370271075365,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.028363645956027694,
      "loss": 0.8098,
      "num_input_tokens_seen": 18314960,
      "step": 31565
    },
    {
      "epoch": 4.702114983616324,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.028362760348990337,
      "loss": 0.7928,
      "num_input_tokens_seen": 18317648,
      "step": 31570
    },
    {
      "epoch": 4.702859696157283,
      "grad_norm": 0.01953125,
      "learning_rate": 0.028361874516202718,
      "loss": 0.7992,
      "num_input_tokens_seen": 18320656,
      "step": 31575
    },
    {
      "epoch": 4.703604408698243,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.028360988457679818,
      "loss": 0.812,
      "num_input_tokens_seen": 18323504,
      "step": 31580
    },
    {
      "epoch": 4.7043491212392015,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02836010217343659,
      "loss": 0.8227,
      "num_input_tokens_seen": 18326320,
      "step": 31585
    },
    {
      "epoch": 4.705093833780161,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02835921566348802,
      "loss": 0.8267,
      "num_input_tokens_seen": 18329424,
      "step": 31590
    },
    {
      "epoch": 4.70583854632112,
      "grad_norm": 0.0234375,
      "learning_rate": 0.028358328927849077,
      "loss": 0.8062,
      "num_input_tokens_seen": 18332208,
      "step": 31595
    },
    {
      "epoch": 4.7065832588620795,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02835744196653474,
      "loss": 0.7957,
      "num_input_tokens_seen": 18334928,
      "step": 31600
    },
    {
      "epoch": 4.707327971403038,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.028356554779560006,
      "loss": 0.7982,
      "num_input_tokens_seen": 18337872,
      "step": 31605
    },
    {
      "epoch": 4.708072683943998,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.028355667366939848,
      "loss": 0.805,
      "num_input_tokens_seen": 18340912,
      "step": 31610
    },
    {
      "epoch": 4.708817396484957,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.028354779728689262,
      "loss": 0.7982,
      "num_input_tokens_seen": 18343600,
      "step": 31615
    },
    {
      "epoch": 4.709562109025916,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.028353891864823247,
      "loss": 0.8168,
      "num_input_tokens_seen": 18346352,
      "step": 31620
    },
    {
      "epoch": 4.710306821566875,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.028353003775356805,
      "loss": 0.8105,
      "num_input_tokens_seen": 18349104,
      "step": 31625
    },
    {
      "epoch": 4.711051534107835,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02835211546030493,
      "loss": 0.7936,
      "num_input_tokens_seen": 18351920,
      "step": 31630
    },
    {
      "epoch": 4.7117962466487935,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02835122691968264,
      "loss": 0.7936,
      "num_input_tokens_seen": 18354736,
      "step": 31635
    },
    {
      "epoch": 4.712540959189753,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.028350338153504937,
      "loss": 0.7983,
      "num_input_tokens_seen": 18357936,
      "step": 31640
    },
    {
      "epoch": 4.713285671730712,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.028349449161786847,
      "loss": 0.8164,
      "num_input_tokens_seen": 18360880,
      "step": 31645
    },
    {
      "epoch": 4.7140303842716715,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02834855994454337,
      "loss": 0.8101,
      "num_input_tokens_seen": 18363952,
      "step": 31650
    },
    {
      "epoch": 4.71477509681263,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.028347670501789546,
      "loss": 0.8084,
      "num_input_tokens_seen": 18366928,
      "step": 31655
    },
    {
      "epoch": 4.715519809353589,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.028346780833540395,
      "loss": 0.7793,
      "num_input_tokens_seen": 18369776,
      "step": 31660
    },
    {
      "epoch": 4.716264521894549,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.028345890939810943,
      "loss": 0.8208,
      "num_input_tokens_seen": 18372752,
      "step": 31665
    },
    {
      "epoch": 4.717009234435508,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02834500082061623,
      "loss": 0.792,
      "num_input_tokens_seen": 18375376,
      "step": 31670
    },
    {
      "epoch": 4.717753946976467,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.028344110475971292,
      "loss": 0.7879,
      "num_input_tokens_seen": 18378096,
      "step": 31675
    },
    {
      "epoch": 4.718498659517426,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02834321990589117,
      "loss": 0.8127,
      "num_input_tokens_seen": 18380880,
      "step": 31680
    },
    {
      "epoch": 4.7192433720583855,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.028342329110390908,
      "loss": 0.8024,
      "num_input_tokens_seen": 18384080,
      "step": 31685
    },
    {
      "epoch": 4.719988084599344,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.028341438089485558,
      "loss": 0.802,
      "num_input_tokens_seen": 18386992,
      "step": 31690
    },
    {
      "epoch": 4.720732797140304,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.028340546843190168,
      "loss": 0.8094,
      "num_input_tokens_seen": 18390128,
      "step": 31695
    },
    {
      "epoch": 4.721477509681263,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.028339655371519803,
      "loss": 0.8039,
      "num_input_tokens_seen": 18393200,
      "step": 31700
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.028338763674489516,
      "loss": 0.7831,
      "num_input_tokens_seen": 18395888,
      "step": 31705
    },
    {
      "epoch": 4.722966934763181,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.02833787175211437,
      "loss": 0.8149,
      "num_input_tokens_seen": 18398832,
      "step": 31710
    },
    {
      "epoch": 4.723711647304141,
      "grad_norm": 0.01190185546875,
      "learning_rate": 0.028336979604409437,
      "loss": 0.7966,
      "num_input_tokens_seen": 18401744,
      "step": 31715
    },
    {
      "epoch": 4.7244563598450995,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.028336087231389792,
      "loss": 0.7991,
      "num_input_tokens_seen": 18404720,
      "step": 31720
    },
    {
      "epoch": 4.725201072386059,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02833519463307051,
      "loss": 0.8064,
      "num_input_tokens_seen": 18407536,
      "step": 31725
    },
    {
      "epoch": 4.725945784927018,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02833430180946666,
      "loss": 0.8087,
      "num_input_tokens_seen": 18410544,
      "step": 31730
    },
    {
      "epoch": 4.7266904974679775,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.028333408760593343,
      "loss": 0.7898,
      "num_input_tokens_seen": 18413424,
      "step": 31735
    },
    {
      "epoch": 4.727435210008936,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02833251548646563,
      "loss": 0.8036,
      "num_input_tokens_seen": 18416208,
      "step": 31740
    },
    {
      "epoch": 4.728179922549896,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02833162198709862,
      "loss": 0.8147,
      "num_input_tokens_seen": 18419728,
      "step": 31745
    },
    {
      "epoch": 4.728924635090855,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.0283307282625074,
      "loss": 0.799,
      "num_input_tokens_seen": 18422512,
      "step": 31750
    },
    {
      "epoch": 4.729669347631814,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.028329834312707084,
      "loss": 0.8187,
      "num_input_tokens_seen": 18425232,
      "step": 31755
    },
    {
      "epoch": 4.730414060172773,
      "grad_norm": 0.011474609375,
      "learning_rate": 0.02832894013771276,
      "loss": 0.8067,
      "num_input_tokens_seen": 18428016,
      "step": 31760
    },
    {
      "epoch": 4.731158772713733,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.028328045737539546,
      "loss": 0.8044,
      "num_input_tokens_seen": 18431408,
      "step": 31765
    },
    {
      "epoch": 4.7319034852546915,
      "grad_norm": 0.01116943359375,
      "learning_rate": 0.028327151112202537,
      "loss": 0.8039,
      "num_input_tokens_seen": 18434192,
      "step": 31770
    },
    {
      "epoch": 4.732648197795651,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.02832625626171686,
      "loss": 0.7922,
      "num_input_tokens_seen": 18437008,
      "step": 31775
    },
    {
      "epoch": 4.73339291033661,
      "grad_norm": 0.0107421875,
      "learning_rate": 0.028325361186097627,
      "loss": 0.8024,
      "num_input_tokens_seen": 18439888,
      "step": 31780
    },
    {
      "epoch": 4.73413762287757,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02832446588535996,
      "loss": 0.8181,
      "num_input_tokens_seen": 18442704,
      "step": 31785
    },
    {
      "epoch": 4.734882335418528,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.028323570359518976,
      "loss": 0.7976,
      "num_input_tokens_seen": 18445776,
      "step": 31790
    },
    {
      "epoch": 4.735627047959488,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.028322674608589823,
      "loss": 0.8064,
      "num_input_tokens_seen": 18448720,
      "step": 31795
    },
    {
      "epoch": 4.736371760500447,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.028321778632587623,
      "loss": 0.8089,
      "num_input_tokens_seen": 18451376,
      "step": 31800
    },
    {
      "epoch": 4.737116473041406,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028320882431527507,
      "loss": 0.8123,
      "num_input_tokens_seen": 18454448,
      "step": 31805
    },
    {
      "epoch": 4.737861185582365,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.028319986005424624,
      "loss": 0.8018,
      "num_input_tokens_seen": 18457456,
      "step": 31810
    },
    {
      "epoch": 4.738605898123325,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.028319089354294114,
      "loss": 0.8079,
      "num_input_tokens_seen": 18460144,
      "step": 31815
    },
    {
      "epoch": 4.7393506106642835,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02831819247815113,
      "loss": 0.8006,
      "num_input_tokens_seen": 18463280,
      "step": 31820
    },
    {
      "epoch": 4.740095323205242,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.028317295377010814,
      "loss": 0.81,
      "num_input_tokens_seen": 18466000,
      "step": 31825
    },
    {
      "epoch": 4.740840035746202,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02831639805088833,
      "loss": 0.7933,
      "num_input_tokens_seen": 18468816,
      "step": 31830
    },
    {
      "epoch": 4.741584748287162,
      "grad_norm": 0.0234375,
      "learning_rate": 0.028315500499798838,
      "loss": 0.7921,
      "num_input_tokens_seen": 18472048,
      "step": 31835
    },
    {
      "epoch": 4.74232946082812,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.028314602723757496,
      "loss": 0.8061,
      "num_input_tokens_seen": 18474992,
      "step": 31840
    },
    {
      "epoch": 4.743074173369079,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02831370472277947,
      "loss": 0.7914,
      "num_input_tokens_seen": 18477616,
      "step": 31845
    },
    {
      "epoch": 4.743818885910039,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.028312806496879936,
      "loss": 0.79,
      "num_input_tokens_seen": 18480560,
      "step": 31850
    },
    {
      "epoch": 4.744563598450998,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.028311908046074072,
      "loss": 0.7992,
      "num_input_tokens_seen": 18483152,
      "step": 31855
    },
    {
      "epoch": 4.745308310991957,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028311009370377047,
      "loss": 0.7912,
      "num_input_tokens_seen": 18485968,
      "step": 31860
    },
    {
      "epoch": 4.746053023532916,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02831011046980405,
      "loss": 0.8019,
      "num_input_tokens_seen": 18488944,
      "step": 31865
    },
    {
      "epoch": 4.746797736073876,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.028309211344370262,
      "loss": 0.7888,
      "num_input_tokens_seen": 18492048,
      "step": 31870
    },
    {
      "epoch": 4.747542448614834,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.028308311994090875,
      "loss": 0.8185,
      "num_input_tokens_seen": 18494768,
      "step": 31875
    },
    {
      "epoch": 4.748287161155794,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.028307412418981086,
      "loss": 0.8139,
      "num_input_tokens_seen": 18497776,
      "step": 31880
    },
    {
      "epoch": 4.749031873696753,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.028306512619056083,
      "loss": 0.7898,
      "num_input_tokens_seen": 18500496,
      "step": 31885
    },
    {
      "epoch": 4.749776586237712,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.028305612594331078,
      "loss": 0.8191,
      "num_input_tokens_seen": 18503568,
      "step": 31890
    },
    {
      "epoch": 4.750521298778671,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.028304712344821268,
      "loss": 0.811,
      "num_input_tokens_seen": 18506736,
      "step": 31895
    },
    {
      "epoch": 4.751266011319631,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02830381187054187,
      "loss": 0.7888,
      "num_input_tokens_seen": 18509872,
      "step": 31900
    },
    {
      "epoch": 4.7520107238605895,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02830291117150809,
      "loss": 0.8093,
      "num_input_tokens_seen": 18513200,
      "step": 31905
    },
    {
      "epoch": 4.752755436401549,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.028302010247735147,
      "loss": 0.7874,
      "num_input_tokens_seen": 18515856,
      "step": 31910
    },
    {
      "epoch": 4.753500148942508,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02830110909923826,
      "loss": 0.804,
      "num_input_tokens_seen": 18518768,
      "step": 31915
    },
    {
      "epoch": 4.754244861483468,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02830020772603265,
      "loss": 0.8191,
      "num_input_tokens_seen": 18521776,
      "step": 31920
    },
    {
      "epoch": 4.754989574024426,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.028299306128133552,
      "loss": 0.8044,
      "num_input_tokens_seen": 18524624,
      "step": 31925
    },
    {
      "epoch": 4.755734286565386,
      "grad_norm": 0.011474609375,
      "learning_rate": 0.028298404305556194,
      "loss": 0.8086,
      "num_input_tokens_seen": 18527536,
      "step": 31930
    },
    {
      "epoch": 4.756478999106345,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.028297502258315812,
      "loss": 0.8095,
      "num_input_tokens_seen": 18530288,
      "step": 31935
    },
    {
      "epoch": 4.757223711647304,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02829659998642764,
      "loss": 0.8011,
      "num_input_tokens_seen": 18533264,
      "step": 31940
    },
    {
      "epoch": 4.757968424188263,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.028295697489906926,
      "loss": 0.7979,
      "num_input_tokens_seen": 18536144,
      "step": 31945
    },
    {
      "epoch": 4.758713136729223,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.028294794768768918,
      "loss": 0.8001,
      "num_input_tokens_seen": 18538800,
      "step": 31950
    },
    {
      "epoch": 4.759457849270182,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.028293891823028865,
      "loss": 0.8055,
      "num_input_tokens_seen": 18541456,
      "step": 31955
    },
    {
      "epoch": 4.760202561811141,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.02829298865270202,
      "loss": 0.7872,
      "num_input_tokens_seen": 18544176,
      "step": 31960
    },
    {
      "epoch": 4.7609472743521,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.02829208525780364,
      "loss": 0.8387,
      "num_input_tokens_seen": 18548368,
      "step": 31965
    },
    {
      "epoch": 4.76169198689306,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.028291181638348994,
      "loss": 0.8076,
      "num_input_tokens_seen": 18551216,
      "step": 31970
    },
    {
      "epoch": 4.762436699434018,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02829027779435334,
      "loss": 0.8233,
      "num_input_tokens_seen": 18553968,
      "step": 31975
    },
    {
      "epoch": 4.763181411974978,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.02828937372583195,
      "loss": 0.8026,
      "num_input_tokens_seen": 18556816,
      "step": 31980
    },
    {
      "epoch": 4.763926124515937,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.028288469432800098,
      "loss": 0.7825,
      "num_input_tokens_seen": 18559760,
      "step": 31985
    },
    {
      "epoch": 4.764670837056896,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.028287564915273056,
      "loss": 0.7882,
      "num_input_tokens_seen": 18562544,
      "step": 31990
    },
    {
      "epoch": 4.765415549597855,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02828666017326612,
      "loss": 0.8033,
      "num_input_tokens_seen": 18565488,
      "step": 31995
    },
    {
      "epoch": 4.766160262138815,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.028285755206794556,
      "loss": 0.81,
      "num_input_tokens_seen": 18568624,
      "step": 32000
    },
    {
      "epoch": 4.766904974679774,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.028284850015873664,
      "loss": 0.8,
      "num_input_tokens_seen": 18571984,
      "step": 32005
    },
    {
      "epoch": 4.767649687220732,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028283944600518734,
      "loss": 0.7962,
      "num_input_tokens_seen": 18574800,
      "step": 32010
    },
    {
      "epoch": 4.768394399761692,
      "grad_norm": 0.017578125,
      "learning_rate": 0.02828303896074506,
      "loss": 0.7807,
      "num_input_tokens_seen": 18577552,
      "step": 32015
    },
    {
      "epoch": 4.769139112302652,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028282133096567942,
      "loss": 0.8294,
      "num_input_tokens_seen": 18580688,
      "step": 32020
    },
    {
      "epoch": 4.76988382484361,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028281227008002684,
      "loss": 0.7996,
      "num_input_tokens_seen": 18583600,
      "step": 32025
    },
    {
      "epoch": 4.770628537384569,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.028280320695064594,
      "loss": 0.7954,
      "num_input_tokens_seen": 18587024,
      "step": 32030
    },
    {
      "epoch": 4.771373249925529,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.028279414157768988,
      "loss": 0.7787,
      "num_input_tokens_seen": 18589552,
      "step": 32035
    },
    {
      "epoch": 4.772117962466488,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.028278507396131174,
      "loss": 0.8012,
      "num_input_tokens_seen": 18592496,
      "step": 32040
    },
    {
      "epoch": 4.772862675007447,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02827760041016647,
      "loss": 0.8029,
      "num_input_tokens_seen": 18595248,
      "step": 32045
    },
    {
      "epoch": 4.773607387548406,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028276693199890204,
      "loss": 0.8042,
      "num_input_tokens_seen": 18598032,
      "step": 32050
    },
    {
      "epoch": 4.774352100089366,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.028275785765317703,
      "loss": 0.7937,
      "num_input_tokens_seen": 18601200,
      "step": 32055
    },
    {
      "epoch": 4.775096812630324,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.028274878106464292,
      "loss": 0.8054,
      "num_input_tokens_seen": 18604176,
      "step": 32060
    },
    {
      "epoch": 4.775841525171284,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.028273970223345306,
      "loss": 0.8216,
      "num_input_tokens_seen": 18607024,
      "step": 32065
    },
    {
      "epoch": 4.776586237712243,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.028273062115976084,
      "loss": 0.8277,
      "num_input_tokens_seen": 18609648,
      "step": 32070
    },
    {
      "epoch": 4.777330950253202,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.028272153784371968,
      "loss": 0.8159,
      "num_input_tokens_seen": 18612720,
      "step": 32075
    },
    {
      "epoch": 4.778075662794161,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.028271245228548306,
      "loss": 0.8006,
      "num_input_tokens_seen": 18615408,
      "step": 32080
    },
    {
      "epoch": 4.778820375335121,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02827033644852044,
      "loss": 0.7966,
      "num_input_tokens_seen": 18618192,
      "step": 32085
    },
    {
      "epoch": 4.77956508787608,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.028269427444303732,
      "loss": 0.8077,
      "num_input_tokens_seen": 18621104,
      "step": 32090
    },
    {
      "epoch": 4.780309800417039,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02826851821591353,
      "loss": 0.778,
      "num_input_tokens_seen": 18624112,
      "step": 32095
    },
    {
      "epoch": 4.781054512957998,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.0282676087633652,
      "loss": 0.7933,
      "num_input_tokens_seen": 18626864,
      "step": 32100
    },
    {
      "epoch": 4.781799225498958,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.028266699086674105,
      "loss": 0.8197,
      "num_input_tokens_seen": 18629936,
      "step": 32105
    },
    {
      "epoch": 4.782543938039916,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.028265789185855608,
      "loss": 0.8199,
      "num_input_tokens_seen": 18632784,
      "step": 32110
    },
    {
      "epoch": 4.783288650580876,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028264879060925088,
      "loss": 0.8123,
      "num_input_tokens_seen": 18635728,
      "step": 32115
    },
    {
      "epoch": 4.784033363121835,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02826396871189792,
      "loss": 0.8023,
      "num_input_tokens_seen": 18638448,
      "step": 32120
    },
    {
      "epoch": 4.7847780756627944,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.028263058138789484,
      "loss": 0.8022,
      "num_input_tokens_seen": 18641296,
      "step": 32125
    },
    {
      "epoch": 4.785522788203753,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.028262147341615154,
      "loss": 0.7918,
      "num_input_tokens_seen": 18644528,
      "step": 32130
    },
    {
      "epoch": 4.786267500744713,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.028261236320390327,
      "loss": 0.7685,
      "num_input_tokens_seen": 18647760,
      "step": 32135
    },
    {
      "epoch": 4.787012213285672,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.02826032507513039,
      "loss": 0.8114,
      "num_input_tokens_seen": 18650736,
      "step": 32140
    },
    {
      "epoch": 4.787756925826631,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.028259413605850738,
      "loss": 0.7854,
      "num_input_tokens_seen": 18653712,
      "step": 32145
    },
    {
      "epoch": 4.78850163836759,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.028258501912566774,
      "loss": 0.8245,
      "num_input_tokens_seen": 18656752,
      "step": 32150
    },
    {
      "epoch": 4.78924635090855,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.028257589995293892,
      "loss": 0.7764,
      "num_input_tokens_seen": 18659632,
      "step": 32155
    },
    {
      "epoch": 4.789991063449508,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.0282566778540475,
      "loss": 0.8273,
      "num_input_tokens_seen": 18662480,
      "step": 32160
    },
    {
      "epoch": 4.790735775990468,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02825576548884301,
      "loss": 0.801,
      "num_input_tokens_seen": 18665232,
      "step": 32165
    },
    {
      "epoch": 4.791480488531427,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.028254852899695832,
      "loss": 0.8065,
      "num_input_tokens_seen": 18668208,
      "step": 32170
    },
    {
      "epoch": 4.792225201072386,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.028253940086621392,
      "loss": 0.7951,
      "num_input_tokens_seen": 18671280,
      "step": 32175
    },
    {
      "epoch": 4.792969913613345,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.0282530270496351,
      "loss": 0.8066,
      "num_input_tokens_seen": 18674352,
      "step": 32180
    },
    {
      "epoch": 4.793714626154305,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02825211378875239,
      "loss": 0.7964,
      "num_input_tokens_seen": 18677296,
      "step": 32185
    },
    {
      "epoch": 4.794459338695264,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.028251200303988682,
      "loss": 0.7923,
      "num_input_tokens_seen": 18681008,
      "step": 32190
    },
    {
      "epoch": 4.795204051236222,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.028250286595359417,
      "loss": 0.8401,
      "num_input_tokens_seen": 18684112,
      "step": 32195
    },
    {
      "epoch": 4.795948763777182,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02824937266288002,
      "loss": 0.7895,
      "num_input_tokens_seen": 18686864,
      "step": 32200
    },
    {
      "epoch": 4.796693476318142,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.028248458506565946,
      "loss": 0.8312,
      "num_input_tokens_seen": 18689840,
      "step": 32205
    },
    {
      "epoch": 4.7974381888591004,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028247544126432624,
      "loss": 0.7992,
      "num_input_tokens_seen": 18692752,
      "step": 32210
    },
    {
      "epoch": 4.798182901400059,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02824662952249551,
      "loss": 0.7955,
      "num_input_tokens_seen": 18695408,
      "step": 32215
    },
    {
      "epoch": 4.798927613941019,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02824571469477006,
      "loss": 0.7889,
      "num_input_tokens_seen": 18698224,
      "step": 32220
    },
    {
      "epoch": 4.799672326481978,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.028244799643271715,
      "loss": 0.8008,
      "num_input_tokens_seen": 18701072,
      "step": 32225
    },
    {
      "epoch": 4.800417039022937,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028243884368015944,
      "loss": 0.8039,
      "num_input_tokens_seen": 18703824,
      "step": 32230
    },
    {
      "epoch": 4.801161751563896,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02824296886901821,
      "loss": 0.8013,
      "num_input_tokens_seen": 18706992,
      "step": 32235
    },
    {
      "epoch": 4.801906464104856,
      "grad_norm": 0.025390625,
      "learning_rate": 0.028242053146293973,
      "loss": 0.8058,
      "num_input_tokens_seen": 18709968,
      "step": 32240
    },
    {
      "epoch": 4.802651176645814,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.028241137199858702,
      "loss": 0.8135,
      "num_input_tokens_seen": 18712816,
      "step": 32245
    },
    {
      "epoch": 4.803395889186774,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.02824022102972788,
      "loss": 0.801,
      "num_input_tokens_seen": 18716944,
      "step": 32250
    },
    {
      "epoch": 4.804140601727733,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.028239304635916986,
      "loss": 0.798,
      "num_input_tokens_seen": 18719792,
      "step": 32255
    },
    {
      "epoch": 4.8048853142686925,
      "grad_norm": 0.0126953125,
      "learning_rate": 0.02823838801844149,
      "loss": 0.797,
      "num_input_tokens_seen": 18722512,
      "step": 32260
    },
    {
      "epoch": 4.805630026809651,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.028237471177316883,
      "loss": 0.8022,
      "num_input_tokens_seen": 18725424,
      "step": 32265
    },
    {
      "epoch": 4.806374739350611,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028236554112558657,
      "loss": 0.8091,
      "num_input_tokens_seen": 18728304,
      "step": 32270
    },
    {
      "epoch": 4.80711945189157,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.0282356368241823,
      "loss": 0.8153,
      "num_input_tokens_seen": 18731152,
      "step": 32275
    },
    {
      "epoch": 4.807864164432529,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.028234719312203314,
      "loss": 0.8129,
      "num_input_tokens_seen": 18733968,
      "step": 32280
    },
    {
      "epoch": 4.808608876973488,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.028233801576637192,
      "loss": 0.7958,
      "num_input_tokens_seen": 18736816,
      "step": 32285
    },
    {
      "epoch": 4.809353589514448,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.028232883617499448,
      "loss": 0.7909,
      "num_input_tokens_seen": 18739696,
      "step": 32290
    },
    {
      "epoch": 4.8100983020554064,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.02823196543480558,
      "loss": 0.8209,
      "num_input_tokens_seen": 18742672,
      "step": 32295
    },
    {
      "epoch": 4.810843014596366,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028231047028571106,
      "loss": 0.7939,
      "num_input_tokens_seen": 18745808,
      "step": 32300
    },
    {
      "epoch": 4.811587727137325,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.02823012839881154,
      "loss": 0.7962,
      "num_input_tokens_seen": 18748688,
      "step": 32305
    },
    {
      "epoch": 4.8123324396782845,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.0282292095455424,
      "loss": 0.793,
      "num_input_tokens_seen": 18751664,
      "step": 32310
    },
    {
      "epoch": 4.813077152219243,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.028228290468779213,
      "loss": 0.7891,
      "num_input_tokens_seen": 18754384,
      "step": 32315
    },
    {
      "epoch": 4.813821864760203,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.0282273711685375,
      "loss": 0.7974,
      "num_input_tokens_seen": 18757168,
      "step": 32320
    },
    {
      "epoch": 4.814566577301162,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.0282264516448328,
      "loss": 0.7856,
      "num_input_tokens_seen": 18760368,
      "step": 32325
    },
    {
      "epoch": 4.815311289842121,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.028225531897680636,
      "loss": 0.8069,
      "num_input_tokens_seen": 18763472,
      "step": 32330
    },
    {
      "epoch": 4.81605600238308,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.028224611927096556,
      "loss": 0.7978,
      "num_input_tokens_seen": 18766352,
      "step": 32335
    },
    {
      "epoch": 4.816800714924039,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.028223691733096096,
      "loss": 0.7804,
      "num_input_tokens_seen": 18769264,
      "step": 32340
    },
    {
      "epoch": 4.8175454274649985,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.028222771315694808,
      "loss": 0.8194,
      "num_input_tokens_seen": 18772272,
      "step": 32345
    },
    {
      "epoch": 4.818290140005958,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.028221850674908237,
      "loss": 0.805,
      "num_input_tokens_seen": 18775120,
      "step": 32350
    },
    {
      "epoch": 4.819034852546917,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.028220929810751933,
      "loss": 0.8071,
      "num_input_tokens_seen": 18778032,
      "step": 32355
    },
    {
      "epoch": 4.819779565087876,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.028220008723241458,
      "loss": 0.8202,
      "num_input_tokens_seen": 18780624,
      "step": 32360
    },
    {
      "epoch": 4.820524277628835,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.028219087412392374,
      "loss": 0.7639,
      "num_input_tokens_seen": 18783344,
      "step": 32365
    },
    {
      "epoch": 4.821268990169795,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.028218165878220243,
      "loss": 0.8264,
      "num_input_tokens_seen": 18786128,
      "step": 32370
    },
    {
      "epoch": 4.822013702710754,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.028217244120740632,
      "loss": 0.7867,
      "num_input_tokens_seen": 18789008,
      "step": 32375
    },
    {
      "epoch": 4.8227584152517124,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.028216322139969115,
      "loss": 0.8322,
      "num_input_tokens_seen": 18792080,
      "step": 32380
    },
    {
      "epoch": 4.823503127792672,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02821539993592127,
      "loss": 0.7997,
      "num_input_tokens_seen": 18795056,
      "step": 32385
    },
    {
      "epoch": 4.824247840333631,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.028214477508612675,
      "loss": 0.8168,
      "num_input_tokens_seen": 18797648,
      "step": 32390
    },
    {
      "epoch": 4.8249925528745905,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.02821355485805891,
      "loss": 0.7977,
      "num_input_tokens_seen": 18800528,
      "step": 32395
    },
    {
      "epoch": 4.825737265415549,
      "grad_norm": 0.01544189453125,
      "learning_rate": 0.02821263198427557,
      "loss": 0.8006,
      "num_input_tokens_seen": 18803760,
      "step": 32400
    },
    {
      "epoch": 4.826481977956509,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02821170888727824,
      "loss": 0.7975,
      "num_input_tokens_seen": 18806928,
      "step": 32405
    },
    {
      "epoch": 4.827226690497468,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.028210785567082515,
      "loss": 0.813,
      "num_input_tokens_seen": 18810096,
      "step": 32410
    },
    {
      "epoch": 4.827971403038427,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.028209862023703996,
      "loss": 0.8121,
      "num_input_tokens_seen": 18813264,
      "step": 32415
    },
    {
      "epoch": 4.828716115579386,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.028208938257158277,
      "loss": 0.8017,
      "num_input_tokens_seen": 18816240,
      "step": 32420
    },
    {
      "epoch": 4.829460828120346,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02820801426746098,
      "loss": 0.809,
      "num_input_tokens_seen": 18819056,
      "step": 32425
    },
    {
      "epoch": 4.8302055406613045,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.0282070900546277,
      "loss": 0.7784,
      "num_input_tokens_seen": 18822096,
      "step": 32430
    },
    {
      "epoch": 4.830950253202264,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.028206165618674058,
      "loss": 0.7875,
      "num_input_tokens_seen": 18825008,
      "step": 32435
    },
    {
      "epoch": 4.831694965743223,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.028205240959615675,
      "loss": 0.8094,
      "num_input_tokens_seen": 18827664,
      "step": 32440
    },
    {
      "epoch": 4.8324396782841825,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.028204316077468164,
      "loss": 0.8,
      "num_input_tokens_seen": 18830832,
      "step": 32445
    },
    {
      "epoch": 4.833184390825141,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.02820339097224715,
      "loss": 0.8197,
      "num_input_tokens_seen": 18833680,
      "step": 32450
    },
    {
      "epoch": 4.833929103366101,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.028202465643968267,
      "loss": 0.8028,
      "num_input_tokens_seen": 18836752,
      "step": 32455
    },
    {
      "epoch": 4.83467381590706,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.028201540092647148,
      "loss": 0.7981,
      "num_input_tokens_seen": 18839568,
      "step": 32460
    },
    {
      "epoch": 4.835418528448019,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.02820061431829942,
      "loss": 0.8101,
      "num_input_tokens_seen": 18842608,
      "step": 32465
    },
    {
      "epoch": 4.836163240988978,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.028199688320940736,
      "loss": 0.7891,
      "num_input_tokens_seen": 18845904,
      "step": 32470
    },
    {
      "epoch": 4.836907953529938,
      "grad_norm": 0.1328125,
      "learning_rate": 0.028198762100586732,
      "loss": 0.7731,
      "num_input_tokens_seen": 18848880,
      "step": 32475
    },
    {
      "epoch": 4.8376526660708965,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.028197835657253055,
      "loss": 0.8964,
      "num_input_tokens_seen": 18851824,
      "step": 32480
    },
    {
      "epoch": 4.838397378611856,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02819690899095536,
      "loss": 0.8163,
      "num_input_tokens_seen": 18854672,
      "step": 32485
    },
    {
      "epoch": 4.839142091152815,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.028195982101709302,
      "loss": 0.8086,
      "num_input_tokens_seen": 18857712,
      "step": 32490
    },
    {
      "epoch": 4.8398868036937746,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.028195054989530535,
      "loss": 0.7728,
      "num_input_tokens_seen": 18860528,
      "step": 32495
    },
    {
      "epoch": 4.840631516234733,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.028194127654434725,
      "loss": 0.7847,
      "num_input_tokens_seen": 18863312,
      "step": 32500
    },
    {
      "epoch": 4.841376228775693,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.028193200096437537,
      "loss": 0.7742,
      "num_input_tokens_seen": 18866192,
      "step": 32505
    },
    {
      "epoch": 4.842120941316652,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.028192272315554642,
      "loss": 0.7712,
      "num_input_tokens_seen": 18869072,
      "step": 32510
    },
    {
      "epoch": 4.842865653857611,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02819134431180172,
      "loss": 0.8122,
      "num_input_tokens_seen": 18872176,
      "step": 32515
    },
    {
      "epoch": 4.84361036639857,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02819041608519444,
      "loss": 0.7936,
      "num_input_tokens_seen": 18874896,
      "step": 32520
    },
    {
      "epoch": 4.844355078939529,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.028189487635748484,
      "loss": 0.7932,
      "num_input_tokens_seen": 18877552,
      "step": 32525
    },
    {
      "epoch": 4.8450997914804885,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.02818855896347954,
      "loss": 0.815,
      "num_input_tokens_seen": 18880592,
      "step": 32530
    },
    {
      "epoch": 4.845844504021448,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.028187630068403302,
      "loss": 0.8214,
      "num_input_tokens_seen": 18883344,
      "step": 32535
    },
    {
      "epoch": 4.846589216562407,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.028186700950535454,
      "loss": 0.7966,
      "num_input_tokens_seen": 18886512,
      "step": 32540
    },
    {
      "epoch": 4.847333929103366,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.028185771609891695,
      "loss": 0.789,
      "num_input_tokens_seen": 18889616,
      "step": 32545
    },
    {
      "epoch": 4.848078641644325,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02818484204648772,
      "loss": 0.7863,
      "num_input_tokens_seen": 18892400,
      "step": 32550
    },
    {
      "epoch": 4.848823354185284,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02818391226033925,
      "loss": 0.789,
      "num_input_tokens_seen": 18895184,
      "step": 32555
    },
    {
      "epoch": 4.849568066726244,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.02818298225146198,
      "loss": 0.7879,
      "num_input_tokens_seen": 18898000,
      "step": 32560
    },
    {
      "epoch": 4.8503127792672025,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.028182052019871617,
      "loss": 0.8253,
      "num_input_tokens_seen": 18900944,
      "step": 32565
    },
    {
      "epoch": 4.851057491808162,
      "grad_norm": 0.021484375,
      "learning_rate": 0.028181121565583885,
      "loss": 0.7948,
      "num_input_tokens_seen": 18904112,
      "step": 32570
    },
    {
      "epoch": 4.851802204349121,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.028180190888614506,
      "loss": 0.7969,
      "num_input_tokens_seen": 18907024,
      "step": 32575
    },
    {
      "epoch": 4.8525469168900806,
      "grad_norm": 0.01953125,
      "learning_rate": 0.028179259988979192,
      "loss": 0.8088,
      "num_input_tokens_seen": 18909968,
      "step": 32580
    },
    {
      "epoch": 4.853291629431039,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.028178328866693678,
      "loss": 0.8195,
      "num_input_tokens_seen": 18912848,
      "step": 32585
    },
    {
      "epoch": 4.854036341971999,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02817739752177369,
      "loss": 0.7934,
      "num_input_tokens_seen": 18915504,
      "step": 32590
    },
    {
      "epoch": 4.854781054512958,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.028176465954234963,
      "loss": 0.8145,
      "num_input_tokens_seen": 18918064,
      "step": 32595
    },
    {
      "epoch": 4.855525767053917,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.028175534164093238,
      "loss": 0.7842,
      "num_input_tokens_seen": 18921072,
      "step": 32600
    },
    {
      "epoch": 4.856270479594876,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.028174602151364254,
      "loss": 0.7709,
      "num_input_tokens_seen": 18924016,
      "step": 32605
    },
    {
      "epoch": 4.857015192135836,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.028173669916063755,
      "loss": 0.8172,
      "num_input_tokens_seen": 18926768,
      "step": 32610
    },
    {
      "epoch": 4.8577599046767945,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.028172737458207493,
      "loss": 0.8304,
      "num_input_tokens_seen": 18929296,
      "step": 32615
    },
    {
      "epoch": 4.858504617217754,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028171804777811223,
      "loss": 0.7872,
      "num_input_tokens_seen": 18932624,
      "step": 32620
    },
    {
      "epoch": 4.859249329758713,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.028170871874890695,
      "loss": 0.8162,
      "num_input_tokens_seen": 18935536,
      "step": 32625
    },
    {
      "epoch": 4.859994042299673,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.028169938749461672,
      "loss": 0.8199,
      "num_input_tokens_seen": 18938480,
      "step": 32630
    },
    {
      "epoch": 4.860738754840631,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028169005401539923,
      "loss": 0.8083,
      "num_input_tokens_seen": 18941328,
      "step": 32635
    },
    {
      "epoch": 4.861483467381591,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02816807183114121,
      "loss": 0.8076,
      "num_input_tokens_seen": 18944432,
      "step": 32640
    },
    {
      "epoch": 4.86222817992255,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.028167138038281306,
      "loss": 0.8022,
      "num_input_tokens_seen": 18947088,
      "step": 32645
    },
    {
      "epoch": 4.862972892463509,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02816620402297599,
      "loss": 0.804,
      "num_input_tokens_seen": 18949968,
      "step": 32650
    },
    {
      "epoch": 4.863717605004468,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.028165269785241035,
      "loss": 0.782,
      "num_input_tokens_seen": 18952880,
      "step": 32655
    },
    {
      "epoch": 4.864462317545428,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02816433532509223,
      "loss": 0.8237,
      "num_input_tokens_seen": 18955824,
      "step": 32660
    },
    {
      "epoch": 4.8652070300863866,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.028163400642545355,
      "loss": 0.8142,
      "num_input_tokens_seen": 18959088,
      "step": 32665
    },
    {
      "epoch": 4.865951742627346,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02816246573761621,
      "loss": 0.8109,
      "num_input_tokens_seen": 18961808,
      "step": 32670
    },
    {
      "epoch": 4.866696455168305,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.028161530610320583,
      "loss": 0.7953,
      "num_input_tokens_seen": 18964528,
      "step": 32675
    },
    {
      "epoch": 4.867441167709265,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02816059526067427,
      "loss": 0.7945,
      "num_input_tokens_seen": 18967280,
      "step": 32680
    },
    {
      "epoch": 4.868185880250223,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.028159659688693082,
      "loss": 0.7788,
      "num_input_tokens_seen": 18969936,
      "step": 32685
    },
    {
      "epoch": 4.868930592791182,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02815872389439281,
      "loss": 0.8082,
      "num_input_tokens_seen": 18972752,
      "step": 32690
    },
    {
      "epoch": 4.869675305332142,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.02815778787778928,
      "loss": 0.7887,
      "num_input_tokens_seen": 18975536,
      "step": 32695
    },
    {
      "epoch": 4.870420017873101,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.028156851638898292,
      "loss": 0.8109,
      "num_input_tokens_seen": 18978192,
      "step": 32700
    },
    {
      "epoch": 4.87116473041406,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02815591517773567,
      "loss": 0.8005,
      "num_input_tokens_seen": 18981136,
      "step": 32705
    },
    {
      "epoch": 4.871909442955019,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02815497849431723,
      "loss": 0.8,
      "num_input_tokens_seen": 18984048,
      "step": 32710
    },
    {
      "epoch": 4.872654155495979,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.0281540415886588,
      "loss": 0.8125,
      "num_input_tokens_seen": 18986768,
      "step": 32715
    },
    {
      "epoch": 4.873398868036938,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.028153104460776204,
      "loss": 0.8098,
      "num_input_tokens_seen": 18989456,
      "step": 32720
    },
    {
      "epoch": 4.874143580577897,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02815216711068528,
      "loss": 0.7891,
      "num_input_tokens_seen": 18992464,
      "step": 32725
    },
    {
      "epoch": 4.874888293118856,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.028151229538401858,
      "loss": 0.8087,
      "num_input_tokens_seen": 18995312,
      "step": 32730
    },
    {
      "epoch": 4.875633005659815,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02815029174394178,
      "loss": 0.7936,
      "num_input_tokens_seen": 18998384,
      "step": 32735
    },
    {
      "epoch": 4.876377718200774,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.028149353727320886,
      "loss": 0.8027,
      "num_input_tokens_seen": 19001488,
      "step": 32740
    },
    {
      "epoch": 4.877122430741734,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02814841548855503,
      "loss": 0.8034,
      "num_input_tokens_seen": 19004560,
      "step": 32745
    },
    {
      "epoch": 4.8778671432826926,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.028147477027660052,
      "loss": 0.8192,
      "num_input_tokens_seen": 19007504,
      "step": 32750
    },
    {
      "epoch": 4.878611855823652,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.028146538344651816,
      "loss": 0.7962,
      "num_input_tokens_seen": 19010224,
      "step": 32755
    },
    {
      "epoch": 4.879356568364611,
      "grad_norm": 0.021484375,
      "learning_rate": 0.028145599439546178,
      "loss": 0.7859,
      "num_input_tokens_seen": 19012752,
      "step": 32760
    },
    {
      "epoch": 4.880101280905571,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.028144660312358995,
      "loss": 0.8155,
      "num_input_tokens_seen": 19015952,
      "step": 32765
    },
    {
      "epoch": 4.880845993446529,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.028143720963106134,
      "loss": 0.7949,
      "num_input_tokens_seen": 19018768,
      "step": 32770
    },
    {
      "epoch": 4.881590705987489,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02814278139180347,
      "loss": 0.8108,
      "num_input_tokens_seen": 19021360,
      "step": 32775
    },
    {
      "epoch": 4.882335418528448,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028141841598466867,
      "loss": 0.7756,
      "num_input_tokens_seen": 19024080,
      "step": 32780
    },
    {
      "epoch": 4.883080131069407,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02814090158311221,
      "loss": 0.8099,
      "num_input_tokens_seen": 19026864,
      "step": 32785
    },
    {
      "epoch": 4.883824843610366,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.028139961345755377,
      "loss": 0.8072,
      "num_input_tokens_seen": 19029648,
      "step": 32790
    },
    {
      "epoch": 4.884569556151326,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02813902088641225,
      "loss": 0.7946,
      "num_input_tokens_seen": 19032592,
      "step": 32795
    },
    {
      "epoch": 4.885314268692285,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.028138080205098722,
      "loss": 0.8092,
      "num_input_tokens_seen": 19035568,
      "step": 32800
    },
    {
      "epoch": 4.886058981233244,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02813713930183068,
      "loss": 0.8013,
      "num_input_tokens_seen": 19038384,
      "step": 32805
    },
    {
      "epoch": 4.886803693774203,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028136198176624022,
      "loss": 0.7963,
      "num_input_tokens_seen": 19041264,
      "step": 32810
    },
    {
      "epoch": 4.887548406315163,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02813525682949465,
      "loss": 0.7954,
      "num_input_tokens_seen": 19044208,
      "step": 32815
    },
    {
      "epoch": 4.888293118856121,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028134315260458457,
      "loss": 0.7975,
      "num_input_tokens_seen": 19047408,
      "step": 32820
    },
    {
      "epoch": 4.889037831397081,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02813337346953136,
      "loss": 0.7973,
      "num_input_tokens_seen": 19050032,
      "step": 32825
    },
    {
      "epoch": 4.88978254393804,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02813243145672927,
      "loss": 0.7887,
      "num_input_tokens_seen": 19053104,
      "step": 32830
    },
    {
      "epoch": 4.890527256478999,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.028131489222068096,
      "loss": 0.828,
      "num_input_tokens_seen": 19055792,
      "step": 32835
    },
    {
      "epoch": 4.891271969019958,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.028130546765563754,
      "loss": 0.7873,
      "num_input_tokens_seen": 19058800,
      "step": 32840
    },
    {
      "epoch": 4.892016681560918,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.028129604087232172,
      "loss": 0.8199,
      "num_input_tokens_seen": 19061488,
      "step": 32845
    },
    {
      "epoch": 4.892761394101877,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02812866118708927,
      "loss": 0.7845,
      "num_input_tokens_seen": 19064656,
      "step": 32850
    },
    {
      "epoch": 4.893506106642836,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.028127718065150985,
      "loss": 0.7891,
      "num_input_tokens_seen": 19067632,
      "step": 32855
    },
    {
      "epoch": 4.894250819183795,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.028126774721433243,
      "loss": 0.8083,
      "num_input_tokens_seen": 19070320,
      "step": 32860
    },
    {
      "epoch": 4.894995531724755,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.028125831155951988,
      "loss": 0.8111,
      "num_input_tokens_seen": 19073424,
      "step": 32865
    },
    {
      "epoch": 4.895740244265713,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.028124887368723154,
      "loss": 0.7903,
      "num_input_tokens_seen": 19076400,
      "step": 32870
    },
    {
      "epoch": 4.896484956806672,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.028123943359762687,
      "loss": 0.8127,
      "num_input_tokens_seen": 19079152,
      "step": 32875
    },
    {
      "epoch": 4.897229669347632,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02812299912908654,
      "loss": 0.7906,
      "num_input_tokens_seen": 19081936,
      "step": 32880
    },
    {
      "epoch": 4.8979743818885915,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.028122054676710653,
      "loss": 0.8124,
      "num_input_tokens_seen": 19084656,
      "step": 32885
    },
    {
      "epoch": 4.89871909442955,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028121110002650992,
      "loss": 0.806,
      "num_input_tokens_seen": 19087696,
      "step": 32890
    },
    {
      "epoch": 4.899463806970509,
      "grad_norm": 0.0234375,
      "learning_rate": 0.028120165106923514,
      "loss": 0.7971,
      "num_input_tokens_seen": 19090672,
      "step": 32895
    },
    {
      "epoch": 4.900208519511469,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.028119219989544177,
      "loss": 0.8185,
      "num_input_tokens_seen": 19093392,
      "step": 32900
    },
    {
      "epoch": 4.900953232052427,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.028118274650528956,
      "loss": 0.8013,
      "num_input_tokens_seen": 19096112,
      "step": 32905
    },
    {
      "epoch": 4.901697944593387,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.028117329089893815,
      "loss": 0.7918,
      "num_input_tokens_seen": 19098960,
      "step": 32910
    },
    {
      "epoch": 4.902442657134346,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.028116383307654733,
      "loss": 0.8329,
      "num_input_tokens_seen": 19101680,
      "step": 32915
    },
    {
      "epoch": 4.903187369675305,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.028115437303827685,
      "loss": 0.8178,
      "num_input_tokens_seen": 19104272,
      "step": 32920
    },
    {
      "epoch": 4.903932082216264,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.028114491078428654,
      "loss": 0.7893,
      "num_input_tokens_seen": 19107216,
      "step": 32925
    },
    {
      "epoch": 4.904676794757224,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.028113544631473628,
      "loss": 0.8108,
      "num_input_tokens_seen": 19110064,
      "step": 32930
    },
    {
      "epoch": 4.905421507298183,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02811259796297859,
      "loss": 0.7889,
      "num_input_tokens_seen": 19113072,
      "step": 32935
    },
    {
      "epoch": 4.906166219839142,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.028111651072959536,
      "loss": 0.8209,
      "num_input_tokens_seen": 19116048,
      "step": 32940
    },
    {
      "epoch": 4.906910932380101,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.028110703961432466,
      "loss": 0.812,
      "num_input_tokens_seen": 19119152,
      "step": 32945
    },
    {
      "epoch": 4.907655644921061,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.028109756628413377,
      "loss": 0.8114,
      "num_input_tokens_seen": 19121968,
      "step": 32950
    },
    {
      "epoch": 4.908400357462019,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02810880907391827,
      "loss": 0.8108,
      "num_input_tokens_seen": 19124784,
      "step": 32955
    },
    {
      "epoch": 4.909145070002979,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028107861297963162,
      "loss": 0.8127,
      "num_input_tokens_seen": 19127888,
      "step": 32960
    },
    {
      "epoch": 4.909889782543938,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028106913300564057,
      "loss": 0.7876,
      "num_input_tokens_seen": 19130960,
      "step": 32965
    },
    {
      "epoch": 4.9106344950848975,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.028105965081736975,
      "loss": 0.7925,
      "num_input_tokens_seen": 19133872,
      "step": 32970
    },
    {
      "epoch": 4.911379207625856,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02810501664149793,
      "loss": 0.8045,
      "num_input_tokens_seen": 19136752,
      "step": 32975
    },
    {
      "epoch": 4.912123920166816,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02810406797986295,
      "loss": 0.8046,
      "num_input_tokens_seen": 19139440,
      "step": 32980
    },
    {
      "epoch": 4.912868632707775,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02810311909684806,
      "loss": 0.8032,
      "num_input_tokens_seen": 19142544,
      "step": 32985
    },
    {
      "epoch": 4.913613345248734,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.028102169992469294,
      "loss": 0.7853,
      "num_input_tokens_seen": 19145360,
      "step": 32990
    },
    {
      "epoch": 4.914358057789693,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.028101220666742678,
      "loss": 0.8063,
      "num_input_tokens_seen": 19149008,
      "step": 32995
    },
    {
      "epoch": 4.915102770330653,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02810027111968425,
      "loss": 0.7978,
      "num_input_tokens_seen": 19151920,
      "step": 33000
    },
    {
      "epoch": 4.915847482871611,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.028099321351310064,
      "loss": 0.7905,
      "num_input_tokens_seen": 19154864,
      "step": 33005
    },
    {
      "epoch": 4.916592195412571,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.028098371361636156,
      "loss": 0.8088,
      "num_input_tokens_seen": 19157744,
      "step": 33010
    },
    {
      "epoch": 4.91733690795353,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.028097421150678575,
      "loss": 0.8192,
      "num_input_tokens_seen": 19160752,
      "step": 33015
    },
    {
      "epoch": 4.9180816204944895,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02809647071845337,
      "loss": 0.8322,
      "num_input_tokens_seen": 19163824,
      "step": 33020
    },
    {
      "epoch": 4.918826333035448,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.028095520064976613,
      "loss": 0.8019,
      "num_input_tokens_seen": 19166672,
      "step": 33025
    },
    {
      "epoch": 4.919571045576408,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.028094569190264346,
      "loss": 0.828,
      "num_input_tokens_seen": 19169584,
      "step": 33030
    },
    {
      "epoch": 4.920315758117367,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.028093618094332647,
      "loss": 0.7896,
      "num_input_tokens_seen": 19172720,
      "step": 33035
    },
    {
      "epoch": 4.921060470658325,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.028092666777197576,
      "loss": 0.7961,
      "num_input_tokens_seen": 19175568,
      "step": 33040
    },
    {
      "epoch": 4.921805183199285,
      "grad_norm": 0.03125,
      "learning_rate": 0.028091715238875204,
      "loss": 0.8196,
      "num_input_tokens_seen": 19178704,
      "step": 33045
    },
    {
      "epoch": 4.922549895740245,
      "grad_norm": 0.011474609375,
      "learning_rate": 0.028090763479381612,
      "loss": 0.797,
      "num_input_tokens_seen": 19181776,
      "step": 33050
    },
    {
      "epoch": 4.9232946082812035,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.028089811498732874,
      "loss": 0.787,
      "num_input_tokens_seen": 19184592,
      "step": 33055
    },
    {
      "epoch": 4.924039320822162,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.028088859296945077,
      "loss": 0.796,
      "num_input_tokens_seen": 19187600,
      "step": 33060
    },
    {
      "epoch": 4.924784033363122,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.028087906874034303,
      "loss": 0.8184,
      "num_input_tokens_seen": 19190224,
      "step": 33065
    },
    {
      "epoch": 4.9255287459040815,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.028086954230016642,
      "loss": 0.788,
      "num_input_tokens_seen": 19193392,
      "step": 33070
    },
    {
      "epoch": 4.92627345844504,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.028086001364908197,
      "loss": 0.7804,
      "num_input_tokens_seen": 19196336,
      "step": 33075
    },
    {
      "epoch": 4.927018170985999,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.028085048278725055,
      "loss": 0.7906,
      "num_input_tokens_seen": 19199184,
      "step": 33080
    },
    {
      "epoch": 4.927762883526959,
      "grad_norm": 0.013671875,
      "learning_rate": 0.02808409497148332,
      "loss": 0.7896,
      "num_input_tokens_seen": 19202064,
      "step": 33085
    },
    {
      "epoch": 4.928507596067917,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.028083141443199098,
      "loss": 0.7769,
      "num_input_tokens_seen": 19205072,
      "step": 33090
    },
    {
      "epoch": 4.929252308608877,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.0280821876938885,
      "loss": 0.8146,
      "num_input_tokens_seen": 19207792,
      "step": 33095
    },
    {
      "epoch": 4.929997021149836,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.028081233723567637,
      "loss": 0.8148,
      "num_input_tokens_seen": 19210480,
      "step": 33100
    },
    {
      "epoch": 4.9307417336907955,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02808027953225263,
      "loss": 0.807,
      "num_input_tokens_seen": 19213360,
      "step": 33105
    },
    {
      "epoch": 4.931486446231754,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02807932511995959,
      "loss": 0.7982,
      "num_input_tokens_seen": 19216112,
      "step": 33110
    },
    {
      "epoch": 4.932231158772714,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028078370486704644,
      "loss": 0.788,
      "num_input_tokens_seen": 19219024,
      "step": 33115
    },
    {
      "epoch": 4.932975871313673,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.028077415632503922,
      "loss": 0.8194,
      "num_input_tokens_seen": 19221680,
      "step": 33120
    },
    {
      "epoch": 4.933720583854632,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.028076460557373555,
      "loss": 0.7916,
      "num_input_tokens_seen": 19224336,
      "step": 33125
    },
    {
      "epoch": 4.934465296395591,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.028075505261329677,
      "loss": 0.8203,
      "num_input_tokens_seen": 19227120,
      "step": 33130
    },
    {
      "epoch": 4.935210008936551,
      "grad_norm": 0.029296875,
      "learning_rate": 0.028074549744388427,
      "loss": 0.7985,
      "num_input_tokens_seen": 19230128,
      "step": 33135
    },
    {
      "epoch": 4.9359547214775095,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.028073594006565946,
      "loss": 0.7838,
      "num_input_tokens_seen": 19232944,
      "step": 33140
    },
    {
      "epoch": 4.936699434018469,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.028072638047878382,
      "loss": 0.8176,
      "num_input_tokens_seen": 19235664,
      "step": 33145
    },
    {
      "epoch": 4.937444146559428,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02807168186834189,
      "loss": 0.8101,
      "num_input_tokens_seen": 19238608,
      "step": 33150
    },
    {
      "epoch": 4.9381888591003875,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.028070725467972613,
      "loss": 0.7973,
      "num_input_tokens_seen": 19241584,
      "step": 33155
    },
    {
      "epoch": 4.938933571641346,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.028069768846786716,
      "loss": 0.8143,
      "num_input_tokens_seen": 19244496,
      "step": 33160
    },
    {
      "epoch": 4.939678284182306,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.028068812004800353,
      "loss": 0.8304,
      "num_input_tokens_seen": 19247664,
      "step": 33165
    },
    {
      "epoch": 4.940422996723265,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.028067854942029697,
      "loss": 0.8133,
      "num_input_tokens_seen": 19250608,
      "step": 33170
    },
    {
      "epoch": 4.941167709264224,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.028066897658490915,
      "loss": 0.797,
      "num_input_tokens_seen": 19253776,
      "step": 33175
    },
    {
      "epoch": 4.941912421805183,
      "grad_norm": 0.029296875,
      "learning_rate": 0.028065940154200178,
      "loss": 0.7896,
      "num_input_tokens_seen": 19256400,
      "step": 33180
    },
    {
      "epoch": 4.942657134346143,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.028064982429173657,
      "loss": 0.7978,
      "num_input_tokens_seen": 19259152,
      "step": 33185
    },
    {
      "epoch": 4.9434018468871015,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.028064024483427537,
      "loss": 0.7795,
      "num_input_tokens_seen": 19261968,
      "step": 33190
    },
    {
      "epoch": 4.944146559428061,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.028063066316978007,
      "loss": 0.8044,
      "num_input_tokens_seen": 19265136,
      "step": 33195
    },
    {
      "epoch": 4.94489127196902,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.028062107929841245,
      "loss": 0.8221,
      "num_input_tokens_seen": 19268080,
      "step": 33200
    },
    {
      "epoch": 4.945635984509979,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.028061149322033446,
      "loss": 0.8113,
      "num_input_tokens_seen": 19270960,
      "step": 33205
    },
    {
      "epoch": 4.946380697050938,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.0280601904935708,
      "loss": 0.8036,
      "num_input_tokens_seen": 19273616,
      "step": 33210
    },
    {
      "epoch": 4.947125409591898,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028059231444469517,
      "loss": 0.79,
      "num_input_tokens_seen": 19276400,
      "step": 33215
    },
    {
      "epoch": 4.947870122132857,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.028058272174745783,
      "loss": 0.8052,
      "num_input_tokens_seen": 19279056,
      "step": 33220
    },
    {
      "epoch": 4.9486148346738155,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02805731268441582,
      "loss": 0.7877,
      "num_input_tokens_seen": 19281904,
      "step": 33225
    },
    {
      "epoch": 4.949359547214775,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.028056352973495825,
      "loss": 0.7759,
      "num_input_tokens_seen": 19284784,
      "step": 33230
    },
    {
      "epoch": 4.950104259755735,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.028055393042002018,
      "loss": 0.8038,
      "num_input_tokens_seen": 19287952,
      "step": 33235
    },
    {
      "epoch": 4.9508489722966935,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.028054432889950614,
      "loss": 0.7976,
      "num_input_tokens_seen": 19290896,
      "step": 33240
    },
    {
      "epoch": 4.951593684837652,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.028053472517357837,
      "loss": 0.8234,
      "num_input_tokens_seen": 19293648,
      "step": 33245
    },
    {
      "epoch": 4.952338397378612,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.028052511924239902,
      "loss": 0.7856,
      "num_input_tokens_seen": 19296592,
      "step": 33250
    },
    {
      "epoch": 4.953083109919571,
      "grad_norm": 0.025390625,
      "learning_rate": 0.028051551110613047,
      "loss": 0.7999,
      "num_input_tokens_seen": 19299504,
      "step": 33255
    },
    {
      "epoch": 4.95382782246053,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.028050590076493503,
      "loss": 0.7952,
      "num_input_tokens_seen": 19302704,
      "step": 33260
    },
    {
      "epoch": 4.954572535001489,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.028049628821897505,
      "loss": 0.8183,
      "num_input_tokens_seen": 19305936,
      "step": 33265
    },
    {
      "epoch": 4.955317247542449,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.028048667346841288,
      "loss": 0.8173,
      "num_input_tokens_seen": 19308720,
      "step": 33270
    },
    {
      "epoch": 4.9560619600834075,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.028047705651341102,
      "loss": 0.7924,
      "num_input_tokens_seen": 19311440,
      "step": 33275
    },
    {
      "epoch": 4.956806672624367,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.028046743735413184,
      "loss": 0.8261,
      "num_input_tokens_seen": 19314768,
      "step": 33280
    },
    {
      "epoch": 4.957551385165326,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.028045781599073796,
      "loss": 0.808,
      "num_input_tokens_seen": 19317584,
      "step": 33285
    },
    {
      "epoch": 4.9582960977062855,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.028044819242339182,
      "loss": 0.8093,
      "num_input_tokens_seen": 19320496,
      "step": 33290
    },
    {
      "epoch": 4.959040810247244,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.028043856665225613,
      "loss": 0.799,
      "num_input_tokens_seen": 19323760,
      "step": 33295
    },
    {
      "epoch": 4.959785522788204,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.028042893867749333,
      "loss": 0.8097,
      "num_input_tokens_seen": 19326512,
      "step": 33300
    },
    {
      "epoch": 4.960530235329163,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.028041930849926625,
      "loss": 0.8062,
      "num_input_tokens_seen": 19329200,
      "step": 33305
    },
    {
      "epoch": 4.961274947870122,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.028040967611773746,
      "loss": 0.8096,
      "num_input_tokens_seen": 19331920,
      "step": 33310
    },
    {
      "epoch": 4.962019660411081,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02804000415330698,
      "loss": 0.8074,
      "num_input_tokens_seen": 19334608,
      "step": 33315
    },
    {
      "epoch": 4.962764372952041,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02803904047454259,
      "loss": 0.7969,
      "num_input_tokens_seen": 19337456,
      "step": 33320
    },
    {
      "epoch": 4.9635090854929995,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.028038076575496864,
      "loss": 0.8008,
      "num_input_tokens_seen": 19340528,
      "step": 33325
    },
    {
      "epoch": 4.964253798033959,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02803711245618609,
      "loss": 0.7927,
      "num_input_tokens_seen": 19343248,
      "step": 33330
    },
    {
      "epoch": 4.964998510574918,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.028036148116626546,
      "loss": 0.8157,
      "num_input_tokens_seen": 19345776,
      "step": 33335
    },
    {
      "epoch": 4.965743223115878,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.028035183556834535,
      "loss": 0.8192,
      "num_input_tokens_seen": 19348624,
      "step": 33340
    },
    {
      "epoch": 4.966487935656836,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.028034218776826342,
      "loss": 0.783,
      "num_input_tokens_seen": 19351664,
      "step": 33345
    },
    {
      "epoch": 4.967232648197796,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02803325377661827,
      "loss": 0.8119,
      "num_input_tokens_seen": 19354128,
      "step": 33350
    },
    {
      "epoch": 4.967977360738755,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.028032288556226623,
      "loss": 0.8249,
      "num_input_tokens_seen": 19357008,
      "step": 33355
    },
    {
      "epoch": 4.968722073279714,
      "grad_norm": 0.0126953125,
      "learning_rate": 0.028031323115667705,
      "loss": 0.7983,
      "num_input_tokens_seen": 19359952,
      "step": 33360
    },
    {
      "epoch": 4.969466785820673,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02803035745495783,
      "loss": 0.8037,
      "num_input_tokens_seen": 19363088,
      "step": 33365
    },
    {
      "epoch": 4.970211498361633,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02802939157411331,
      "loss": 0.7887,
      "num_input_tokens_seen": 19365840,
      "step": 33370
    },
    {
      "epoch": 4.9709562109025915,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02802842547315046,
      "loss": 0.8115,
      "num_input_tokens_seen": 19368912,
      "step": 33375
    },
    {
      "epoch": 4.971700923443551,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.028027459152085602,
      "loss": 0.8273,
      "num_input_tokens_seen": 19371792,
      "step": 33380
    },
    {
      "epoch": 4.97244563598451,
      "grad_norm": 0.037109375,
      "learning_rate": 0.02802649261093506,
      "loss": 0.7886,
      "num_input_tokens_seen": 19374704,
      "step": 33385
    },
    {
      "epoch": 4.973190348525469,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.02802552584971517,
      "loss": 0.8136,
      "num_input_tokens_seen": 19377456,
      "step": 33390
    },
    {
      "epoch": 4.973935061066428,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.028024558868442257,
      "loss": 0.8006,
      "num_input_tokens_seen": 19380496,
      "step": 33395
    },
    {
      "epoch": 4.974679773607388,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.02802359166713266,
      "loss": 0.8157,
      "num_input_tokens_seen": 19383504,
      "step": 33400
    },
    {
      "epoch": 4.975424486148347,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.02802262424580272,
      "loss": 0.8142,
      "num_input_tokens_seen": 19386288,
      "step": 33405
    },
    {
      "epoch": 4.9761691986893055,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.028021656604468773,
      "loss": 0.7967,
      "num_input_tokens_seen": 19388912,
      "step": 33410
    },
    {
      "epoch": 4.976913911230265,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.028020688743147178,
      "loss": 0.8071,
      "num_input_tokens_seen": 19392368,
      "step": 33415
    },
    {
      "epoch": 4.977658623771224,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02801972066185428,
      "loss": 0.8058,
      "num_input_tokens_seen": 19395184,
      "step": 33420
    },
    {
      "epoch": 4.978403336312184,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02801875236060643,
      "loss": 0.7896,
      "num_input_tokens_seen": 19398032,
      "step": 33425
    },
    {
      "epoch": 4.979148048853142,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.028017783839419996,
      "loss": 0.8035,
      "num_input_tokens_seen": 19400976,
      "step": 33430
    },
    {
      "epoch": 4.979892761394102,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.028016815098311332,
      "loss": 0.8204,
      "num_input_tokens_seen": 19403760,
      "step": 33435
    },
    {
      "epoch": 4.980637473935061,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.028015846137296806,
      "loss": 0.8054,
      "num_input_tokens_seen": 19406640,
      "step": 33440
    },
    {
      "epoch": 4.98138218647602,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02801487695639279,
      "loss": 0.7935,
      "num_input_tokens_seen": 19410032,
      "step": 33445
    },
    {
      "epoch": 4.982126899016979,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.028013907555615655,
      "loss": 0.8087,
      "num_input_tokens_seen": 19412752,
      "step": 33450
    },
    {
      "epoch": 4.982871611557939,
      "grad_norm": 0.0234375,
      "learning_rate": 0.028012937934981778,
      "loss": 0.8043,
      "num_input_tokens_seen": 19415728,
      "step": 33455
    },
    {
      "epoch": 4.9836163240988975,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.028011968094507535,
      "loss": 0.8085,
      "num_input_tokens_seen": 19418992,
      "step": 33460
    },
    {
      "epoch": 4.984361036639857,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.028010998034209325,
      "loss": 0.8268,
      "num_input_tokens_seen": 19422032,
      "step": 33465
    },
    {
      "epoch": 4.985105749180816,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.028010027754103523,
      "loss": 0.7991,
      "num_input_tokens_seen": 19425328,
      "step": 33470
    },
    {
      "epoch": 4.985850461721776,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.028009057254206524,
      "loss": 0.7892,
      "num_input_tokens_seen": 19428304,
      "step": 33475
    },
    {
      "epoch": 4.986595174262734,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.028008086534534727,
      "loss": 0.8062,
      "num_input_tokens_seen": 19431312,
      "step": 33480
    },
    {
      "epoch": 4.987339886803694,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.028007115595104527,
      "loss": 0.8032,
      "num_input_tokens_seen": 19433968,
      "step": 33485
    },
    {
      "epoch": 4.988084599344653,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02800614443593233,
      "loss": 0.8002,
      "num_input_tokens_seen": 19436528,
      "step": 33490
    },
    {
      "epoch": 4.988829311885612,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.02800517305703454,
      "loss": 0.7929,
      "num_input_tokens_seen": 19439248,
      "step": 33495
    },
    {
      "epoch": 4.989574024426571,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.028004201458427573,
      "loss": 0.8063,
      "num_input_tokens_seen": 19442096,
      "step": 33500
    },
    {
      "epoch": 4.990318736967531,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.028003229640127834,
      "loss": 0.8126,
      "num_input_tokens_seen": 19445136,
      "step": 33505
    },
    {
      "epoch": 4.99106344950849,
      "grad_norm": 0.02734375,
      "learning_rate": 0.028002257602151748,
      "loss": 0.8089,
      "num_input_tokens_seen": 19448112,
      "step": 33510
    },
    {
      "epoch": 4.991808162049449,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.028001285344515736,
      "loss": 0.8148,
      "num_input_tokens_seen": 19451120,
      "step": 33515
    },
    {
      "epoch": 4.992552874590408,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.028000312867236216,
      "loss": 0.8033,
      "num_input_tokens_seen": 19454000,
      "step": 33520
    },
    {
      "epoch": 4.993297587131368,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02799934017032963,
      "loss": 0.8052,
      "num_input_tokens_seen": 19456848,
      "step": 33525
    },
    {
      "epoch": 4.994042299672326,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.027998367253812405,
      "loss": 0.8031,
      "num_input_tokens_seen": 19459536,
      "step": 33530
    },
    {
      "epoch": 4.994787012213286,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.027997394117700972,
      "loss": 0.8093,
      "num_input_tokens_seen": 19462384,
      "step": 33535
    },
    {
      "epoch": 4.995531724754245,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.027996420762011774,
      "loss": 0.8,
      "num_input_tokens_seen": 19465296,
      "step": 33540
    },
    {
      "epoch": 4.996276437295204,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.027995447186761262,
      "loss": 0.7924,
      "num_input_tokens_seen": 19468336,
      "step": 33545
    },
    {
      "epoch": 4.997021149836163,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.027994473391965875,
      "loss": 0.8032,
      "num_input_tokens_seen": 19471216,
      "step": 33550
    },
    {
      "epoch": 4.997765862377122,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.027993499377642065,
      "loss": 0.8091,
      "num_input_tokens_seen": 19474160,
      "step": 33555
    },
    {
      "epoch": 4.998510574918082,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.027992525143806287,
      "loss": 0.8196,
      "num_input_tokens_seen": 19476912,
      "step": 33560
    },
    {
      "epoch": 4.999255287459041,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.027991550690475007,
      "loss": 0.8019,
      "num_input_tokens_seen": 19479824,
      "step": 33565
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.027990576017664682,
      "loss": 0.8031,
      "num_input_tokens_seen": 19482128,
      "step": 33570
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.8017240166664124,
      "eval_runtime": 70.6032,
      "eval_samples_per_second": 42.264,
      "eval_steps_per_second": 10.566,
      "num_input_tokens_seen": 19482128,
      "step": 33570
    },
    {
      "epoch": 5.000744712540959,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.027989601125391773,
      "loss": 0.7931,
      "num_input_tokens_seen": 19484976,
      "step": 33575
    },
    {
      "epoch": 5.001489425081918,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02798862601367276,
      "loss": 0.7849,
      "num_input_tokens_seen": 19487792,
      "step": 33580
    },
    {
      "epoch": 5.002234137622877,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02798765068252411,
      "loss": 0.8129,
      "num_input_tokens_seen": 19490448,
      "step": 33585
    },
    {
      "epoch": 5.002978850163837,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.027986675131962305,
      "loss": 0.8053,
      "num_input_tokens_seen": 19493136,
      "step": 33590
    },
    {
      "epoch": 5.003723562704796,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.027985699362003815,
      "loss": 0.7911,
      "num_input_tokens_seen": 19496144,
      "step": 33595
    },
    {
      "epoch": 5.004468275245755,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.027984723372665138,
      "loss": 0.8019,
      "num_input_tokens_seen": 19499056,
      "step": 33600
    },
    {
      "epoch": 5.005212987786714,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.027983747163962753,
      "loss": 0.8098,
      "num_input_tokens_seen": 19502064,
      "step": 33605
    },
    {
      "epoch": 5.005957700327674,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02798277073591316,
      "loss": 0.8001,
      "num_input_tokens_seen": 19504752,
      "step": 33610
    },
    {
      "epoch": 5.006702412868632,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.027981794088532852,
      "loss": 0.8156,
      "num_input_tokens_seen": 19507696,
      "step": 33615
    },
    {
      "epoch": 5.007447125409592,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.027980817221838324,
      "loss": 0.7885,
      "num_input_tokens_seen": 19510672,
      "step": 33620
    },
    {
      "epoch": 5.008191837950551,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02797984013584608,
      "loss": 0.7814,
      "num_input_tokens_seen": 19513424,
      "step": 33625
    },
    {
      "epoch": 5.00893655049151,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02797886283057263,
      "loss": 0.813,
      "num_input_tokens_seen": 19516240,
      "step": 33630
    },
    {
      "epoch": 5.009681263032469,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.027977885306034483,
      "loss": 0.8179,
      "num_input_tokens_seen": 19518896,
      "step": 33635
    },
    {
      "epoch": 5.010425975573429,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.027976907562248157,
      "loss": 0.7961,
      "num_input_tokens_seen": 19521584,
      "step": 33640
    },
    {
      "epoch": 5.011170688114388,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02797592959923017,
      "loss": 0.8232,
      "num_input_tokens_seen": 19524880,
      "step": 33645
    },
    {
      "epoch": 5.011915400655347,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.027974951416997035,
      "loss": 0.7902,
      "num_input_tokens_seen": 19527536,
      "step": 33650
    },
    {
      "epoch": 5.012660113196306,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02797397301556528,
      "loss": 0.8124,
      "num_input_tokens_seen": 19530128,
      "step": 33655
    },
    {
      "epoch": 5.013404825737266,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02797299439495144,
      "loss": 0.804,
      "num_input_tokens_seen": 19533328,
      "step": 33660
    },
    {
      "epoch": 5.014149538278224,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.027972015555172047,
      "loss": 0.8053,
      "num_input_tokens_seen": 19536272,
      "step": 33665
    },
    {
      "epoch": 5.014894250819184,
      "grad_norm": 0.021484375,
      "learning_rate": 0.027971036496243638,
      "loss": 0.7858,
      "num_input_tokens_seen": 19539184,
      "step": 33670
    },
    {
      "epoch": 5.015638963360143,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.027970057218182742,
      "loss": 0.801,
      "num_input_tokens_seen": 19541904,
      "step": 33675
    },
    {
      "epoch": 5.0163836759011025,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.027969077721005917,
      "loss": 0.8027,
      "num_input_tokens_seen": 19545008,
      "step": 33680
    },
    {
      "epoch": 5.017128388442061,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.027968098004729703,
      "loss": 0.7936,
      "num_input_tokens_seen": 19547600,
      "step": 33685
    },
    {
      "epoch": 5.017873100983021,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.027967118069370658,
      "loss": 0.8085,
      "num_input_tokens_seen": 19550640,
      "step": 33690
    },
    {
      "epoch": 5.01861781352398,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.027966137914945326,
      "loss": 0.7859,
      "num_input_tokens_seen": 19553744,
      "step": 33695
    },
    {
      "epoch": 5.019362526064939,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.027965157541470276,
      "loss": 0.8065,
      "num_input_tokens_seen": 19556784,
      "step": 33700
    },
    {
      "epoch": 5.020107238605898,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.027964176948962065,
      "loss": 0.7913,
      "num_input_tokens_seen": 19559792,
      "step": 33705
    },
    {
      "epoch": 5.020851951146858,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.027963196137437262,
      "loss": 0.7921,
      "num_input_tokens_seen": 19562896,
      "step": 33710
    },
    {
      "epoch": 5.021596663687816,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.027962215106912436,
      "loss": 0.7724,
      "num_input_tokens_seen": 19565488,
      "step": 33715
    },
    {
      "epoch": 5.022341376228776,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02796123385740416,
      "loss": 0.7951,
      "num_input_tokens_seen": 19568176,
      "step": 33720
    },
    {
      "epoch": 5.023086088769735,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02796025238892901,
      "loss": 0.8217,
      "num_input_tokens_seen": 19570960,
      "step": 33725
    },
    {
      "epoch": 5.0238308013106945,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.02795927070150357,
      "loss": 0.7933,
      "num_input_tokens_seen": 19574320,
      "step": 33730
    },
    {
      "epoch": 5.024575513851653,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.027958288795144417,
      "loss": 0.8,
      "num_input_tokens_seen": 19577040,
      "step": 33735
    },
    {
      "epoch": 5.025320226392613,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.027957306669868152,
      "loss": 0.8001,
      "num_input_tokens_seen": 19580080,
      "step": 33740
    },
    {
      "epoch": 5.026064938933572,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.027956324325691354,
      "loss": 0.8345,
      "num_input_tokens_seen": 19582960,
      "step": 33745
    },
    {
      "epoch": 5.02680965147453,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.027955341762630627,
      "loss": 0.8043,
      "num_input_tokens_seen": 19585552,
      "step": 33750
    },
    {
      "epoch": 5.02755436401549,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02795435898070257,
      "loss": 0.8001,
      "num_input_tokens_seen": 19588560,
      "step": 33755
    },
    {
      "epoch": 5.028299076556449,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.027953375979923784,
      "loss": 0.812,
      "num_input_tokens_seen": 19591248,
      "step": 33760
    },
    {
      "epoch": 5.0290437890974085,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.027952392760310873,
      "loss": 0.8163,
      "num_input_tokens_seen": 19594160,
      "step": 33765
    },
    {
      "epoch": 5.029788501638367,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.027951409321880456,
      "loss": 0.7814,
      "num_input_tokens_seen": 19596912,
      "step": 33770
    },
    {
      "epoch": 5.030533214179327,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.027950425664649132,
      "loss": 0.7828,
      "num_input_tokens_seen": 19599664,
      "step": 33775
    },
    {
      "epoch": 5.031277926720286,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02794944178863354,
      "loss": 0.8075,
      "num_input_tokens_seen": 19602416,
      "step": 33780
    },
    {
      "epoch": 5.032022639261245,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.027948457693850286,
      "loss": 0.8106,
      "num_input_tokens_seen": 19605360,
      "step": 33785
    },
    {
      "epoch": 5.032767351802204,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.027947473380315998,
      "loss": 0.8008,
      "num_input_tokens_seen": 19608304,
      "step": 33790
    },
    {
      "epoch": 5.033512064343164,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.027946488848047304,
      "loss": 0.7911,
      "num_input_tokens_seen": 19611184,
      "step": 33795
    },
    {
      "epoch": 5.034256776884122,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02794550409706084,
      "loss": 0.7841,
      "num_input_tokens_seen": 19614224,
      "step": 33800
    },
    {
      "epoch": 5.035001489425082,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02794451912737325,
      "loss": 0.8261,
      "num_input_tokens_seen": 19617424,
      "step": 33805
    },
    {
      "epoch": 5.035746201966041,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.027943533939001154,
      "loss": 0.8123,
      "num_input_tokens_seen": 19620016,
      "step": 33810
    },
    {
      "epoch": 5.0364909145070005,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.027942548531961216,
      "loss": 0.7998,
      "num_input_tokens_seen": 19622992,
      "step": 33815
    },
    {
      "epoch": 5.037235627047959,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02794156290627007,
      "loss": 0.7908,
      "num_input_tokens_seen": 19625872,
      "step": 33820
    },
    {
      "epoch": 5.037980339588919,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027940577061944374,
      "loss": 0.812,
      "num_input_tokens_seen": 19628784,
      "step": 33825
    },
    {
      "epoch": 5.038725052129878,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.027939590999000778,
      "loss": 0.8011,
      "num_input_tokens_seen": 19631600,
      "step": 33830
    },
    {
      "epoch": 5.039469764670837,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.027938604717455948,
      "loss": 0.8197,
      "num_input_tokens_seen": 19634576,
      "step": 33835
    },
    {
      "epoch": 5.040214477211796,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.027937618217326536,
      "loss": 0.7883,
      "num_input_tokens_seen": 19637648,
      "step": 33840
    },
    {
      "epoch": 5.040959189752756,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.027936631498629214,
      "loss": 0.7914,
      "num_input_tokens_seen": 19640464,
      "step": 33845
    },
    {
      "epoch": 5.0417039022937145,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.027935644561380656,
      "loss": 0.8413,
      "num_input_tokens_seen": 19643056,
      "step": 33850
    },
    {
      "epoch": 5.042448614834674,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.027934657405597526,
      "loss": 0.8116,
      "num_input_tokens_seen": 19645936,
      "step": 33855
    },
    {
      "epoch": 5.043193327375633,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.027933670031296508,
      "loss": 0.804,
      "num_input_tokens_seen": 19648976,
      "step": 33860
    },
    {
      "epoch": 5.0439380399165925,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.027932682438494283,
      "loss": 0.8072,
      "num_input_tokens_seen": 19651984,
      "step": 33865
    },
    {
      "epoch": 5.044682752457551,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02793169462720753,
      "loss": 0.8135,
      "num_input_tokens_seen": 19654864,
      "step": 33870
    },
    {
      "epoch": 5.045427464998511,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.027930706597452937,
      "loss": 0.8117,
      "num_input_tokens_seen": 19657744,
      "step": 33875
    },
    {
      "epoch": 5.04617217753947,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.027929718349247193,
      "loss": 0.7927,
      "num_input_tokens_seen": 19660560,
      "step": 33880
    },
    {
      "epoch": 5.046916890080429,
      "grad_norm": 0.0123291015625,
      "learning_rate": 0.027928729882607008,
      "loss": 0.7796,
      "num_input_tokens_seen": 19663152,
      "step": 33885
    },
    {
      "epoch": 5.047661602621388,
      "grad_norm": 0.02734375,
      "learning_rate": 0.027927741197549066,
      "loss": 0.8082,
      "num_input_tokens_seen": 19665904,
      "step": 33890
    },
    {
      "epoch": 5.048406315162348,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02792675229409008,
      "loss": 0.8117,
      "num_input_tokens_seen": 19668624,
      "step": 33895
    },
    {
      "epoch": 5.0491510277033065,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.027925763172246754,
      "loss": 0.7907,
      "num_input_tokens_seen": 19671504,
      "step": 33900
    },
    {
      "epoch": 5.049895740244266,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02792477383203579,
      "loss": 0.8283,
      "num_input_tokens_seen": 19674416,
      "step": 33905
    },
    {
      "epoch": 5.050640452785225,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02792378427347391,
      "loss": 0.8055,
      "num_input_tokens_seen": 19677328,
      "step": 33910
    },
    {
      "epoch": 5.0513851653261845,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02792279449657783,
      "loss": 0.8148,
      "num_input_tokens_seen": 19680240,
      "step": 33915
    },
    {
      "epoch": 5.052129877867143,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.027921804501364274,
      "loss": 0.8257,
      "num_input_tokens_seen": 19683344,
      "step": 33920
    },
    {
      "epoch": 5.052874590408102,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.02792081428784996,
      "loss": 0.803,
      "num_input_tokens_seen": 19686576,
      "step": 33925
    },
    {
      "epoch": 5.053619302949062,
      "grad_norm": 0.033203125,
      "learning_rate": 0.027919823856051622,
      "loss": 0.7834,
      "num_input_tokens_seen": 19689616,
      "step": 33930
    },
    {
      "epoch": 5.0543640154900205,
      "grad_norm": 0.013671875,
      "learning_rate": 0.027918833205985996,
      "loss": 0.8148,
      "num_input_tokens_seen": 19692432,
      "step": 33935
    },
    {
      "epoch": 5.05510872803098,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.027917842337669804,
      "loss": 0.7892,
      "num_input_tokens_seen": 19695408,
      "step": 33940
    },
    {
      "epoch": 5.055853440571939,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.0279168512511198,
      "loss": 0.8113,
      "num_input_tokens_seen": 19698352,
      "step": 33945
    },
    {
      "epoch": 5.0565981531128985,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.027915859946352723,
      "loss": 0.8033,
      "num_input_tokens_seen": 19701712,
      "step": 33950
    },
    {
      "epoch": 5.057342865653857,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02791486842338532,
      "loss": 0.7891,
      "num_input_tokens_seen": 19704592,
      "step": 33955
    },
    {
      "epoch": 5.058087578194817,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.027913876682234336,
      "loss": 0.8102,
      "num_input_tokens_seen": 19707664,
      "step": 33960
    },
    {
      "epoch": 5.058832290735776,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.027912884722916538,
      "loss": 0.8213,
      "num_input_tokens_seen": 19710704,
      "step": 33965
    },
    {
      "epoch": 5.059577003276735,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.027911892545448673,
      "loss": 0.7907,
      "num_input_tokens_seen": 19713520,
      "step": 33970
    },
    {
      "epoch": 5.060321715817694,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.0279109001498475,
      "loss": 0.7813,
      "num_input_tokens_seen": 19716560,
      "step": 33975
    },
    {
      "epoch": 5.061066428358654,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.0279099075361298,
      "loss": 0.821,
      "num_input_tokens_seen": 19719504,
      "step": 33980
    },
    {
      "epoch": 5.0618111408996125,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02790891470431233,
      "loss": 0.8386,
      "num_input_tokens_seen": 19722352,
      "step": 33985
    },
    {
      "epoch": 5.062555853440572,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.027907921654411867,
      "loss": 0.7928,
      "num_input_tokens_seen": 19725104,
      "step": 33990
    },
    {
      "epoch": 5.063300565981531,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.027906928386445186,
      "loss": 0.7779,
      "num_input_tokens_seen": 19727952,
      "step": 33995
    },
    {
      "epoch": 5.0640452785224905,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02790593490042907,
      "loss": 0.7963,
      "num_input_tokens_seen": 19730992,
      "step": 34000
    },
    {
      "epoch": 5.064789991063449,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.027904941196380297,
      "loss": 0.8033,
      "num_input_tokens_seen": 19734288,
      "step": 34005
    },
    {
      "epoch": 5.065534703604409,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02790394727431566,
      "loss": 0.7951,
      "num_input_tokens_seen": 19737264,
      "step": 34010
    },
    {
      "epoch": 5.066279416145368,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.027902953134251948,
      "loss": 0.8253,
      "num_input_tokens_seen": 19740048,
      "step": 34015
    },
    {
      "epoch": 5.067024128686327,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.027901958776205958,
      "loss": 0.8234,
      "num_input_tokens_seen": 19742736,
      "step": 34020
    },
    {
      "epoch": 5.067768841227286,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02790096420019449,
      "loss": 0.806,
      "num_input_tokens_seen": 19745424,
      "step": 34025
    },
    {
      "epoch": 5.068513553768246,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.027899969406234338,
      "loss": 0.7949,
      "num_input_tokens_seen": 19748176,
      "step": 34030
    },
    {
      "epoch": 5.0692582663092045,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.02789897439434232,
      "loss": 0.7925,
      "num_input_tokens_seen": 19750928,
      "step": 34035
    },
    {
      "epoch": 5.070002978850164,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.027897979164535236,
      "loss": 0.8217,
      "num_input_tokens_seen": 19754000,
      "step": 34040
    },
    {
      "epoch": 5.070747691391123,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.027896983716829905,
      "loss": 0.8315,
      "num_input_tokens_seen": 19756816,
      "step": 34045
    },
    {
      "epoch": 5.071492403932083,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027895988051243138,
      "loss": 0.8,
      "num_input_tokens_seen": 19759888,
      "step": 34050
    },
    {
      "epoch": 5.072237116473041,
      "grad_norm": 0.01129150390625,
      "learning_rate": 0.027894992167791764,
      "loss": 0.8027,
      "num_input_tokens_seen": 19762608,
      "step": 34055
    },
    {
      "epoch": 5.072981829014001,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.027893996066492597,
      "loss": 0.8062,
      "num_input_tokens_seen": 19765392,
      "step": 34060
    },
    {
      "epoch": 5.07372654155496,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02789299974736248,
      "loss": 0.8053,
      "num_input_tokens_seen": 19768336,
      "step": 34065
    },
    {
      "epoch": 5.074471254095919,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02789200321041823,
      "loss": 0.8144,
      "num_input_tokens_seen": 19771152,
      "step": 34070
    },
    {
      "epoch": 5.075215966636878,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02789100645567669,
      "loss": 0.7894,
      "num_input_tokens_seen": 19773712,
      "step": 34075
    },
    {
      "epoch": 5.075960679177838,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.0278900094831547,
      "loss": 0.7844,
      "num_input_tokens_seen": 19776784,
      "step": 34080
    },
    {
      "epoch": 5.0767053917187965,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.027889012292869098,
      "loss": 0.8,
      "num_input_tokens_seen": 19779568,
      "step": 34085
    },
    {
      "epoch": 5.077450104259755,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.027888014884836734,
      "loss": 0.7854,
      "num_input_tokens_seen": 19782256,
      "step": 34090
    },
    {
      "epoch": 5.078194816800715,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.02788701725907446,
      "loss": 0.7847,
      "num_input_tokens_seen": 19784976,
      "step": 34095
    },
    {
      "epoch": 5.078939529341674,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02788601941559912,
      "loss": 0.8148,
      "num_input_tokens_seen": 19787824,
      "step": 34100
    },
    {
      "epoch": 5.079684241882633,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.027885021354427584,
      "loss": 0.8061,
      "num_input_tokens_seen": 19790640,
      "step": 34105
    },
    {
      "epoch": 5.080428954423592,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02788402307557671,
      "loss": 0.8355,
      "num_input_tokens_seen": 19793552,
      "step": 34110
    },
    {
      "epoch": 5.081173666964552,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.027883024579063356,
      "loss": 0.7816,
      "num_input_tokens_seen": 19796528,
      "step": 34115
    },
    {
      "epoch": 5.0819183795055105,
      "grad_norm": 0.01055908203125,
      "learning_rate": 0.027882025864904396,
      "loss": 0.808,
      "num_input_tokens_seen": 19799216,
      "step": 34120
    },
    {
      "epoch": 5.08266309204647,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.0278810269331167,
      "loss": 0.8138,
      "num_input_tokens_seen": 19801904,
      "step": 34125
    },
    {
      "epoch": 5.083407804587429,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.027880027783717152,
      "loss": 0.8011,
      "num_input_tokens_seen": 19804784,
      "step": 34130
    },
    {
      "epoch": 5.084152517128389,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.027879028416722625,
      "loss": 0.7846,
      "num_input_tokens_seen": 19807728,
      "step": 34135
    },
    {
      "epoch": 5.084897229669347,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02787802883215,
      "loss": 0.7965,
      "num_input_tokens_seen": 19810384,
      "step": 34140
    },
    {
      "epoch": 5.085641942210307,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.027877029030016164,
      "loss": 0.7893,
      "num_input_tokens_seen": 19813232,
      "step": 34145
    },
    {
      "epoch": 5.086386654751266,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.027876029010338012,
      "loss": 0.842,
      "num_input_tokens_seen": 19815984,
      "step": 34150
    },
    {
      "epoch": 5.087131367292225,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.027875028773132436,
      "loss": 0.8089,
      "num_input_tokens_seen": 19818992,
      "step": 34155
    },
    {
      "epoch": 5.087876079833184,
      "grad_norm": 0.03125,
      "learning_rate": 0.027874028318416337,
      "loss": 0.8161,
      "num_input_tokens_seen": 19821968,
      "step": 34160
    },
    {
      "epoch": 5.088620792374144,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02787302764620661,
      "loss": 0.8131,
      "num_input_tokens_seen": 19825040,
      "step": 34165
    },
    {
      "epoch": 5.0893655049151025,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.02787202675652017,
      "loss": 0.7983,
      "num_input_tokens_seen": 19828080,
      "step": 34170
    },
    {
      "epoch": 5.090110217456062,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.02787102564937392,
      "loss": 0.792,
      "num_input_tokens_seen": 19830896,
      "step": 34175
    },
    {
      "epoch": 5.090854929997021,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.027870024324784765,
      "loss": 0.8063,
      "num_input_tokens_seen": 19833712,
      "step": 34180
    },
    {
      "epoch": 5.091599642537981,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.027869022782769638,
      "loss": 0.8042,
      "num_input_tokens_seen": 19836592,
      "step": 34185
    },
    {
      "epoch": 5.092344355078939,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.027868021023345447,
      "loss": 0.8114,
      "num_input_tokens_seen": 19839856,
      "step": 34190
    },
    {
      "epoch": 5.093089067619899,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02786701904652912,
      "loss": 0.7958,
      "num_input_tokens_seen": 19842832,
      "step": 34195
    },
    {
      "epoch": 5.093833780160858,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.027866016852337586,
      "loss": 0.8024,
      "num_input_tokens_seen": 19845616,
      "step": 34200
    },
    {
      "epoch": 5.094578492701817,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02786501444078777,
      "loss": 0.812,
      "num_input_tokens_seen": 19848464,
      "step": 34205
    },
    {
      "epoch": 5.095323205242776,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.027864011811896613,
      "loss": 0.7853,
      "num_input_tokens_seen": 19851344,
      "step": 34210
    },
    {
      "epoch": 5.096067917783736,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.027863008965681047,
      "loss": 0.802,
      "num_input_tokens_seen": 19854512,
      "step": 34215
    },
    {
      "epoch": 5.096812630324695,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.027862005902158024,
      "loss": 0.8045,
      "num_input_tokens_seen": 19857424,
      "step": 34220
    },
    {
      "epoch": 5.097557342865654,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.027861002621344475,
      "loss": 0.7774,
      "num_input_tokens_seen": 19860144,
      "step": 34225
    },
    {
      "epoch": 5.098302055406613,
      "grad_norm": 0.029296875,
      "learning_rate": 0.027859999123257366,
      "loss": 0.7751,
      "num_input_tokens_seen": 19863312,
      "step": 34230
    },
    {
      "epoch": 5.099046767947573,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02785899540791364,
      "loss": 0.8063,
      "num_input_tokens_seen": 19866352,
      "step": 34235
    },
    {
      "epoch": 5.099791480488531,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.027857991475330256,
      "loss": 0.8,
      "num_input_tokens_seen": 19869168,
      "step": 34240
    },
    {
      "epoch": 5.100536193029491,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.027856987325524173,
      "loss": 0.8162,
      "num_input_tokens_seen": 19872528,
      "step": 34245
    },
    {
      "epoch": 5.10128090557045,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.027855982958512354,
      "loss": 0.8039,
      "num_input_tokens_seen": 19875248,
      "step": 34250
    },
    {
      "epoch": 5.102025618111409,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.027854978374311775,
      "loss": 0.7961,
      "num_input_tokens_seen": 19877936,
      "step": 34255
    },
    {
      "epoch": 5.102770330652368,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.027853973572939398,
      "loss": 0.8149,
      "num_input_tokens_seen": 19880912,
      "step": 34260
    },
    {
      "epoch": 5.103515043193327,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.027852968554412202,
      "loss": 0.831,
      "num_input_tokens_seen": 19883856,
      "step": 34265
    },
    {
      "epoch": 5.104259755734287,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.027851963318747168,
      "loss": 0.8175,
      "num_input_tokens_seen": 19886960,
      "step": 34270
    },
    {
      "epoch": 5.105004468275245,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02785095786596127,
      "loss": 0.7906,
      "num_input_tokens_seen": 19889840,
      "step": 34275
    },
    {
      "epoch": 5.105749180816205,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.027849952196071507,
      "loss": 0.8,
      "num_input_tokens_seen": 19892688,
      "step": 34280
    },
    {
      "epoch": 5.106493893357164,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.027848946309094857,
      "loss": 0.8071,
      "num_input_tokens_seen": 19895856,
      "step": 34285
    },
    {
      "epoch": 5.107238605898123,
      "grad_norm": 0.01171875,
      "learning_rate": 0.02784794020504832,
      "loss": 0.811,
      "num_input_tokens_seen": 19898576,
      "step": 34290
    },
    {
      "epoch": 5.107983318439082,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.027846933883948894,
      "loss": 0.8056,
      "num_input_tokens_seen": 19901328,
      "step": 34295
    },
    {
      "epoch": 5.108728030980042,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.027845927345813568,
      "loss": 0.821,
      "num_input_tokens_seen": 19903984,
      "step": 34300
    },
    {
      "epoch": 5.109472743521001,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.027844920590659364,
      "loss": 0.8,
      "num_input_tokens_seen": 19906864,
      "step": 34305
    },
    {
      "epoch": 5.11021745606196,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.02784391361850328,
      "loss": 0.8066,
      "num_input_tokens_seen": 19909680,
      "step": 34310
    },
    {
      "epoch": 5.110962168602919,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.027842906429362328,
      "loss": 0.8066,
      "num_input_tokens_seen": 19912464,
      "step": 34315
    },
    {
      "epoch": 5.111706881143879,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.027841899023253527,
      "loss": 0.8223,
      "num_input_tokens_seen": 19915280,
      "step": 34320
    },
    {
      "epoch": 5.112451593684837,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.027840891400193888,
      "loss": 0.7996,
      "num_input_tokens_seen": 19917904,
      "step": 34325
    },
    {
      "epoch": 5.113196306225797,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.027839883560200448,
      "loss": 0.8148,
      "num_input_tokens_seen": 19920560,
      "step": 34330
    },
    {
      "epoch": 5.113941018766756,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02783887550329022,
      "loss": 0.8111,
      "num_input_tokens_seen": 19923248,
      "step": 34335
    },
    {
      "epoch": 5.114685731307715,
      "grad_norm": 0.01031494140625,
      "learning_rate": 0.02783786722948024,
      "loss": 0.7833,
      "num_input_tokens_seen": 19926064,
      "step": 34340
    },
    {
      "epoch": 5.115430443848674,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.027836858738787544,
      "loss": 0.8003,
      "num_input_tokens_seen": 19929040,
      "step": 34345
    },
    {
      "epoch": 5.116175156389634,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.027835850031229164,
      "loss": 0.8014,
      "num_input_tokens_seen": 19931888,
      "step": 34350
    },
    {
      "epoch": 5.116919868930593,
      "grad_norm": 0.021484375,
      "learning_rate": 0.027834841106822144,
      "loss": 0.8031,
      "num_input_tokens_seen": 19934736,
      "step": 34355
    },
    {
      "epoch": 5.117664581471552,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.027833831965583523,
      "loss": 0.7977,
      "num_input_tokens_seen": 19937392,
      "step": 34360
    },
    {
      "epoch": 5.118409294012511,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02783282260753036,
      "loss": 0.7949,
      "num_input_tokens_seen": 19940112,
      "step": 34365
    },
    {
      "epoch": 5.119154006553471,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.0278318130326797,
      "loss": 0.7942,
      "num_input_tokens_seen": 19942992,
      "step": 34370
    },
    {
      "epoch": 5.119898719094429,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.027830803241048602,
      "loss": 0.8076,
      "num_input_tokens_seen": 19946256,
      "step": 34375
    },
    {
      "epoch": 5.120643431635389,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.027829793232654127,
      "loss": 0.7742,
      "num_input_tokens_seen": 19949232,
      "step": 34380
    },
    {
      "epoch": 5.121388144176348,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.027828783007513326,
      "loss": 0.8006,
      "num_input_tokens_seen": 19952208,
      "step": 34385
    },
    {
      "epoch": 5.1221328567173074,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.027827772565643283,
      "loss": 0.7991,
      "num_input_tokens_seen": 19954960,
      "step": 34390
    },
    {
      "epoch": 5.122877569258266,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.027826761907061053,
      "loss": 0.8013,
      "num_input_tokens_seen": 19957904,
      "step": 34395
    },
    {
      "epoch": 5.123622281799226,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02782575103178372,
      "loss": 0.7793,
      "num_input_tokens_seen": 19960720,
      "step": 34400
    },
    {
      "epoch": 5.124366994340185,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.027824739939828362,
      "loss": 0.8197,
      "num_input_tokens_seen": 19963312,
      "step": 34405
    },
    {
      "epoch": 5.125111706881144,
      "grad_norm": 0.01104736328125,
      "learning_rate": 0.027823728631212057,
      "loss": 0.7779,
      "num_input_tokens_seen": 19966192,
      "step": 34410
    },
    {
      "epoch": 5.125856419422103,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02782271710595189,
      "loss": 0.7914,
      "num_input_tokens_seen": 19969296,
      "step": 34415
    },
    {
      "epoch": 5.126601131963063,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.027821705364064947,
      "loss": 0.8089,
      "num_input_tokens_seen": 19972208,
      "step": 34420
    },
    {
      "epoch": 5.127345844504021,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.027820693405568322,
      "loss": 0.7996,
      "num_input_tokens_seen": 19975088,
      "step": 34425
    },
    {
      "epoch": 5.128090557044981,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.027819681230479117,
      "loss": 0.8318,
      "num_input_tokens_seen": 19977840,
      "step": 34430
    },
    {
      "epoch": 5.12883526958594,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.027818668838814425,
      "loss": 0.7898,
      "num_input_tokens_seen": 19980496,
      "step": 34435
    },
    {
      "epoch": 5.129579982126899,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02781765623059135,
      "loss": 0.7927,
      "num_input_tokens_seen": 19983696,
      "step": 34440
    },
    {
      "epoch": 5.130324694667858,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.027816643405827002,
      "loss": 0.8158,
      "num_input_tokens_seen": 19986576,
      "step": 34445
    },
    {
      "epoch": 5.131069407208817,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.027815630364538487,
      "loss": 0.7994,
      "num_input_tokens_seen": 19989200,
      "step": 34450
    },
    {
      "epoch": 5.131814119749777,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.027814617106742927,
      "loss": 0.8187,
      "num_input_tokens_seen": 19991920,
      "step": 34455
    },
    {
      "epoch": 5.132558832290735,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.027813603632457433,
      "loss": 0.8433,
      "num_input_tokens_seen": 19994800,
      "step": 34460
    },
    {
      "epoch": 5.133303544831695,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.027812589941699124,
      "loss": 0.8119,
      "num_input_tokens_seen": 19997680,
      "step": 34465
    },
    {
      "epoch": 5.134048257372654,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.027811576034485138,
      "loss": 0.7891,
      "num_input_tokens_seen": 20000368,
      "step": 34470
    },
    {
      "epoch": 5.1347929699136134,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02781056191083259,
      "loss": 0.8158,
      "num_input_tokens_seen": 20003152,
      "step": 34475
    },
    {
      "epoch": 5.135537682454572,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02780954757075862,
      "loss": 0.7888,
      "num_input_tokens_seen": 20006064,
      "step": 34480
    },
    {
      "epoch": 5.136282394995532,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02780853301428037,
      "loss": 0.7903,
      "num_input_tokens_seen": 20008432,
      "step": 34485
    },
    {
      "epoch": 5.137027107536491,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.027807518241414964,
      "loss": 0.7985,
      "num_input_tokens_seen": 20011440,
      "step": 34490
    },
    {
      "epoch": 5.13777182007745,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.027806503252179555,
      "loss": 0.819,
      "num_input_tokens_seen": 20014064,
      "step": 34495
    },
    {
      "epoch": 5.138516532618409,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.027805488046591293,
      "loss": 0.8089,
      "num_input_tokens_seen": 20016976,
      "step": 34500
    },
    {
      "epoch": 5.139261245159369,
      "grad_norm": 0.01171875,
      "learning_rate": 0.027804472624667325,
      "loss": 0.7856,
      "num_input_tokens_seen": 20020016,
      "step": 34505
    },
    {
      "epoch": 5.140005957700327,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.027803456986424804,
      "loss": 0.8108,
      "num_input_tokens_seen": 20022768,
      "step": 34510
    },
    {
      "epoch": 5.140750670241287,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.027802441131880892,
      "loss": 0.8197,
      "num_input_tokens_seen": 20025712,
      "step": 34515
    },
    {
      "epoch": 5.141495382782246,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.027801425061052747,
      "loss": 0.8128,
      "num_input_tokens_seen": 20028368,
      "step": 34520
    },
    {
      "epoch": 5.1422400953232055,
      "grad_norm": 0.010986328125,
      "learning_rate": 0.02780040877395754,
      "loss": 0.8111,
      "num_input_tokens_seen": 20031472,
      "step": 34525
    },
    {
      "epoch": 5.142984807864164,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02779939227061243,
      "loss": 0.7916,
      "num_input_tokens_seen": 20034192,
      "step": 34530
    },
    {
      "epoch": 5.143729520405124,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.0277983755510346,
      "loss": 0.8159,
      "num_input_tokens_seen": 20037264,
      "step": 34535
    },
    {
      "epoch": 5.144474232946083,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.027797358615241224,
      "loss": 0.8096,
      "num_input_tokens_seen": 20040304,
      "step": 34540
    },
    {
      "epoch": 5.145218945487042,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.02779634146324948,
      "loss": 0.8112,
      "num_input_tokens_seen": 20043248,
      "step": 34545
    },
    {
      "epoch": 5.145963658028001,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02779532409507655,
      "loss": 0.8008,
      "num_input_tokens_seen": 20046064,
      "step": 34550
    },
    {
      "epoch": 5.146708370568961,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027794306510739625,
      "loss": 0.8119,
      "num_input_tokens_seen": 20049392,
      "step": 34555
    },
    {
      "epoch": 5.1474530831099194,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.0277932887102559,
      "loss": 0.7972,
      "num_input_tokens_seen": 20052304,
      "step": 34560
    },
    {
      "epoch": 5.148197795650879,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.027792270693642557,
      "loss": 0.8057,
      "num_input_tokens_seen": 20054960,
      "step": 34565
    },
    {
      "epoch": 5.148942508191838,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.027791252460916807,
      "loss": 0.8091,
      "num_input_tokens_seen": 20058000,
      "step": 34570
    },
    {
      "epoch": 5.1496872207327975,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027790234012095846,
      "loss": 0.7995,
      "num_input_tokens_seen": 20061072,
      "step": 34575
    },
    {
      "epoch": 5.150431933273756,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.027789215347196885,
      "loss": 0.809,
      "num_input_tokens_seen": 20063824,
      "step": 34580
    },
    {
      "epoch": 5.151176645814716,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.027788196466237123,
      "loss": 0.7938,
      "num_input_tokens_seen": 20066896,
      "step": 34585
    },
    {
      "epoch": 5.151921358355675,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02778717736923378,
      "loss": 0.8054,
      "num_input_tokens_seen": 20069648,
      "step": 34590
    },
    {
      "epoch": 5.152666070896634,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02778615805620407,
      "loss": 0.7902,
      "num_input_tokens_seen": 20072528,
      "step": 34595
    },
    {
      "epoch": 5.153410783437593,
      "grad_norm": 0.02734375,
      "learning_rate": 0.027785138527165222,
      "loss": 0.8064,
      "num_input_tokens_seen": 20075504,
      "step": 34600
    },
    {
      "epoch": 5.154155495978552,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.027784118782134444,
      "loss": 0.8055,
      "num_input_tokens_seen": 20078672,
      "step": 34605
    },
    {
      "epoch": 5.1549002085195115,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02778309882112898,
      "loss": 0.7801,
      "num_input_tokens_seen": 20081296,
      "step": 34610
    },
    {
      "epoch": 5.15564492106047,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02778207864416605,
      "loss": 0.7981,
      "num_input_tokens_seen": 20084208,
      "step": 34615
    },
    {
      "epoch": 5.15638963360143,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02778105825126289,
      "loss": 0.8224,
      "num_input_tokens_seen": 20087088,
      "step": 34620
    },
    {
      "epoch": 5.157134346142389,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.027780037642436743,
      "loss": 0.7984,
      "num_input_tokens_seen": 20090032,
      "step": 34625
    },
    {
      "epoch": 5.157879058683348,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.027779016817704852,
      "loss": 0.7863,
      "num_input_tokens_seen": 20092848,
      "step": 34630
    },
    {
      "epoch": 5.158623771224307,
      "grad_norm": 0.0107421875,
      "learning_rate": 0.02777799577708446,
      "loss": 0.7858,
      "num_input_tokens_seen": 20095792,
      "step": 34635
    },
    {
      "epoch": 5.159368483765267,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.027776974520592815,
      "loss": 0.8018,
      "num_input_tokens_seen": 20099056,
      "step": 34640
    },
    {
      "epoch": 5.1601131963062254,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02777595304824717,
      "loss": 0.7876,
      "num_input_tokens_seen": 20101712,
      "step": 34645
    },
    {
      "epoch": 5.160857908847185,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.027774931360064783,
      "loss": 0.8089,
      "num_input_tokens_seen": 20104720,
      "step": 34650
    },
    {
      "epoch": 5.161602621388144,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.027773909456062913,
      "loss": 0.794,
      "num_input_tokens_seen": 20107632,
      "step": 34655
    },
    {
      "epoch": 5.1623473339291035,
      "grad_norm": 0.017578125,
      "learning_rate": 0.02777288733625883,
      "loss": 0.7871,
      "num_input_tokens_seen": 20110768,
      "step": 34660
    },
    {
      "epoch": 5.163092046470062,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.027771865000669794,
      "loss": 0.8353,
      "num_input_tokens_seen": 20113584,
      "step": 34665
    },
    {
      "epoch": 5.163836759011022,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.02777084244931308,
      "loss": 0.7958,
      "num_input_tokens_seen": 20116656,
      "step": 34670
    },
    {
      "epoch": 5.164581471551981,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.027769819682205963,
      "loss": 0.8249,
      "num_input_tokens_seen": 20119696,
      "step": 34675
    },
    {
      "epoch": 5.16532618409294,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02776879669936572,
      "loss": 0.7985,
      "num_input_tokens_seen": 20122448,
      "step": 34680
    },
    {
      "epoch": 5.166070896633899,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.027767773500809636,
      "loss": 0.8122,
      "num_input_tokens_seen": 20125328,
      "step": 34685
    },
    {
      "epoch": 5.166815609174859,
      "grad_norm": 0.03515625,
      "learning_rate": 0.027766750086554992,
      "loss": 0.8108,
      "num_input_tokens_seen": 20128272,
      "step": 34690
    },
    {
      "epoch": 5.1675603217158175,
      "grad_norm": 0.01092529296875,
      "learning_rate": 0.027765726456619084,
      "loss": 0.7984,
      "num_input_tokens_seen": 20131216,
      "step": 34695
    },
    {
      "epoch": 5.168305034256777,
      "grad_norm": 0.021484375,
      "learning_rate": 0.027764702611019203,
      "loss": 0.8163,
      "num_input_tokens_seen": 20134384,
      "step": 34700
    },
    {
      "epoch": 5.169049746797736,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.027763678549772644,
      "loss": 0.7871,
      "num_input_tokens_seen": 20137360,
      "step": 34705
    },
    {
      "epoch": 5.1697944593386955,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.027762654272896706,
      "loss": 0.8081,
      "num_input_tokens_seen": 20140432,
      "step": 34710
    },
    {
      "epoch": 5.170539171879654,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.0277616297804087,
      "loss": 0.7963,
      "num_input_tokens_seen": 20143344,
      "step": 34715
    },
    {
      "epoch": 5.171283884420614,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.027760605072325927,
      "loss": 0.807,
      "num_input_tokens_seen": 20146384,
      "step": 34720
    },
    {
      "epoch": 5.172028596961573,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.0277595801486657,
      "loss": 0.7934,
      "num_input_tokens_seen": 20149136,
      "step": 34725
    },
    {
      "epoch": 5.172773309502532,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.027758555009445334,
      "loss": 0.7933,
      "num_input_tokens_seen": 20151824,
      "step": 34730
    },
    {
      "epoch": 5.173518022043491,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02775752965468215,
      "loss": 0.7914,
      "num_input_tokens_seen": 20154544,
      "step": 34735
    },
    {
      "epoch": 5.174262734584451,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.027756504084393466,
      "loss": 0.8016,
      "num_input_tokens_seen": 20157488,
      "step": 34740
    },
    {
      "epoch": 5.1750074471254095,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.027755478298596616,
      "loss": 0.8058,
      "num_input_tokens_seen": 20159984,
      "step": 34745
    },
    {
      "epoch": 5.175752159666369,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.02775445229730892,
      "loss": 0.7876,
      "num_input_tokens_seen": 20162704,
      "step": 34750
    },
    {
      "epoch": 5.176496872207328,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.027753426080547716,
      "loss": 0.8135,
      "num_input_tokens_seen": 20165616,
      "step": 34755
    },
    {
      "epoch": 5.1772415847482876,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.027752399648330338,
      "loss": 0.7873,
      "num_input_tokens_seen": 20168432,
      "step": 34760
    },
    {
      "epoch": 5.177986297289246,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.02775137300067413,
      "loss": 0.787,
      "num_input_tokens_seen": 20171408,
      "step": 34765
    },
    {
      "epoch": 5.178731009830206,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02775034613759644,
      "loss": 0.7927,
      "num_input_tokens_seen": 20174224,
      "step": 34770
    },
    {
      "epoch": 5.179475722371165,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.027749319059114607,
      "loss": 0.808,
      "num_input_tokens_seen": 20177360,
      "step": 34775
    },
    {
      "epoch": 5.180220434912124,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.027748291765245982,
      "loss": 0.7904,
      "num_input_tokens_seen": 20180176,
      "step": 34780
    },
    {
      "epoch": 5.180965147453083,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02774726425600793,
      "loss": 0.7994,
      "num_input_tokens_seen": 20183024,
      "step": 34785
    },
    {
      "epoch": 5.181709859994042,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.027746236531417806,
      "loss": 0.8035,
      "num_input_tokens_seen": 20185872,
      "step": 34790
    },
    {
      "epoch": 5.1824545725350015,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.027745208591492966,
      "loss": 0.8122,
      "num_input_tokens_seen": 20188592,
      "step": 34795
    },
    {
      "epoch": 5.18319928507596,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.027744180436250782,
      "loss": 0.8001,
      "num_input_tokens_seen": 20191504,
      "step": 34800
    },
    {
      "epoch": 5.18394399761692,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.027743152065708623,
      "loss": 0.8265,
      "num_input_tokens_seen": 20194576,
      "step": 34805
    },
    {
      "epoch": 5.184688710157879,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02774212347988386,
      "loss": 0.8177,
      "num_input_tokens_seen": 20197456,
      "step": 34810
    },
    {
      "epoch": 5.185433422698838,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.027741094678793873,
      "loss": 0.8127,
      "num_input_tokens_seen": 20200368,
      "step": 34815
    },
    {
      "epoch": 5.186178135239797,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02774006566245604,
      "loss": 0.7817,
      "num_input_tokens_seen": 20203152,
      "step": 34820
    },
    {
      "epoch": 5.186922847780757,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.027739036430887747,
      "loss": 0.7913,
      "num_input_tokens_seen": 20205840,
      "step": 34825
    },
    {
      "epoch": 5.1876675603217155,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02773800698410638,
      "loss": 0.7899,
      "num_input_tokens_seen": 20208656,
      "step": 34830
    },
    {
      "epoch": 5.188412272862675,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.027736977322129326,
      "loss": 0.7792,
      "num_input_tokens_seen": 20211344,
      "step": 34835
    },
    {
      "epoch": 5.189156985403634,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.027735947444974,
      "loss": 0.8308,
      "num_input_tokens_seen": 20214288,
      "step": 34840
    },
    {
      "epoch": 5.1899016979445936,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.027734917352657772,
      "loss": 0.7925,
      "num_input_tokens_seen": 20216912,
      "step": 34845
    },
    {
      "epoch": 5.190646410485552,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.027733887045198063,
      "loss": 0.7965,
      "num_input_tokens_seen": 20219984,
      "step": 34850
    },
    {
      "epoch": 5.191391123026512,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.027732856522612277,
      "loss": 0.8058,
      "num_input_tokens_seen": 20222736,
      "step": 34855
    },
    {
      "epoch": 5.192135835567471,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.027731825784917816,
      "loss": 0.8065,
      "num_input_tokens_seen": 20225488,
      "step": 34860
    },
    {
      "epoch": 5.19288054810843,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.027730794832132102,
      "loss": 0.8214,
      "num_input_tokens_seen": 20228336,
      "step": 34865
    },
    {
      "epoch": 5.193625260649389,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.027729763664272548,
      "loss": 0.7902,
      "num_input_tokens_seen": 20231568,
      "step": 34870
    },
    {
      "epoch": 5.194369973190349,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.027728732281356576,
      "loss": 0.8249,
      "num_input_tokens_seen": 20234512,
      "step": 34875
    },
    {
      "epoch": 5.1951146857313075,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.027727700683401607,
      "loss": 0.8002,
      "num_input_tokens_seen": 20237456,
      "step": 34880
    },
    {
      "epoch": 5.195859398272267,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02772666887042507,
      "loss": 0.8033,
      "num_input_tokens_seen": 20240848,
      "step": 34885
    },
    {
      "epoch": 5.196604110813226,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.027725636842444397,
      "loss": 0.8123,
      "num_input_tokens_seen": 20243696,
      "step": 34890
    },
    {
      "epoch": 5.197348823354186,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.027724604599477027,
      "loss": 0.8279,
      "num_input_tokens_seen": 20246608,
      "step": 34895
    },
    {
      "epoch": 5.198093535895144,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.027723572141540392,
      "loss": 0.826,
      "num_input_tokens_seen": 20249360,
      "step": 34900
    },
    {
      "epoch": 5.198838248436104,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.027722539468651938,
      "loss": 0.7986,
      "num_input_tokens_seen": 20252464,
      "step": 34905
    },
    {
      "epoch": 5.199582960977063,
      "grad_norm": 0.01031494140625,
      "learning_rate": 0.027721506580829107,
      "loss": 0.7918,
      "num_input_tokens_seen": 20255056,
      "step": 34910
    },
    {
      "epoch": 5.200327673518022,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02772047347808936,
      "loss": 0.8068,
      "num_input_tokens_seen": 20258160,
      "step": 34915
    },
    {
      "epoch": 5.201072386058981,
      "grad_norm": 0.0111083984375,
      "learning_rate": 0.027719440160450137,
      "loss": 0.7967,
      "num_input_tokens_seen": 20261136,
      "step": 34920
    },
    {
      "epoch": 5.201817098599941,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.027718406627928897,
      "loss": 0.8118,
      "num_input_tokens_seen": 20264176,
      "step": 34925
    },
    {
      "epoch": 5.2025618111408996,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02771737288054311,
      "loss": 0.8223,
      "num_input_tokens_seen": 20266960,
      "step": 34930
    },
    {
      "epoch": 5.203306523681859,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.027716338918310227,
      "loss": 0.7983,
      "num_input_tokens_seen": 20269776,
      "step": 34935
    },
    {
      "epoch": 5.204051236222818,
      "grad_norm": 0.01068115234375,
      "learning_rate": 0.027715304741247726,
      "loss": 0.8092,
      "num_input_tokens_seen": 20272272,
      "step": 34940
    },
    {
      "epoch": 5.204795948763778,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.027714270349373078,
      "loss": 0.8036,
      "num_input_tokens_seen": 20275088,
      "step": 34945
    },
    {
      "epoch": 5.205540661304736,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02771323574270375,
      "loss": 0.8136,
      "num_input_tokens_seen": 20277808,
      "step": 34950
    },
    {
      "epoch": 5.206285373845695,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.027712200921257226,
      "loss": 0.8087,
      "num_input_tokens_seen": 20280496,
      "step": 34955
    },
    {
      "epoch": 5.207030086386655,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.027711165885050987,
      "loss": 0.8099,
      "num_input_tokens_seen": 20283824,
      "step": 34960
    },
    {
      "epoch": 5.2077747989276135,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02771013063410252,
      "loss": 0.8071,
      "num_input_tokens_seen": 20286864,
      "step": 34965
    },
    {
      "epoch": 5.208519511468573,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.027709095168429315,
      "loss": 0.7813,
      "num_input_tokens_seen": 20289648,
      "step": 34970
    },
    {
      "epoch": 5.209264224009532,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.027708059488048865,
      "loss": 0.808,
      "num_input_tokens_seen": 20293072,
      "step": 34975
    },
    {
      "epoch": 5.210008936550492,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.027707023592978663,
      "loss": 0.79,
      "num_input_tokens_seen": 20295920,
      "step": 34980
    },
    {
      "epoch": 5.21075364909145,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.027705987483236213,
      "loss": 0.7941,
      "num_input_tokens_seen": 20298768,
      "step": 34985
    },
    {
      "epoch": 5.21149836163241,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.027704951158839017,
      "loss": 0.8171,
      "num_input_tokens_seen": 20301584,
      "step": 34990
    },
    {
      "epoch": 5.212243074173369,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.027703914619804586,
      "loss": 0.8062,
      "num_input_tokens_seen": 20304624,
      "step": 34995
    },
    {
      "epoch": 5.212987786714328,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.027702877866150433,
      "loss": 0.7884,
      "num_input_tokens_seen": 20307664,
      "step": 35000
    },
    {
      "epoch": 5.213732499255287,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.027701840897894063,
      "loss": 0.8095,
      "num_input_tokens_seen": 20310640,
      "step": 35005
    },
    {
      "epoch": 5.214477211796247,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.027700803715053002,
      "loss": 0.8157,
      "num_input_tokens_seen": 20313328,
      "step": 35010
    },
    {
      "epoch": 5.2152219243372056,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02769976631764477,
      "loss": 0.8069,
      "num_input_tokens_seen": 20316080,
      "step": 35015
    },
    {
      "epoch": 5.215966636878165,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.027698728705686893,
      "loss": 0.8092,
      "num_input_tokens_seen": 20318960,
      "step": 35020
    },
    {
      "epoch": 5.216711349419124,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.027697690879196903,
      "loss": 0.8072,
      "num_input_tokens_seen": 20321936,
      "step": 35025
    },
    {
      "epoch": 5.217456061960084,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.02769665283819233,
      "loss": 0.8033,
      "num_input_tokens_seen": 20325072,
      "step": 35030
    },
    {
      "epoch": 5.218200774501042,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.027695614582690712,
      "loss": 0.8066,
      "num_input_tokens_seen": 20328272,
      "step": 35035
    },
    {
      "epoch": 5.218945487042002,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02769457611270959,
      "loss": 0.8119,
      "num_input_tokens_seen": 20331152,
      "step": 35040
    },
    {
      "epoch": 5.219690199582961,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.027693537428266506,
      "loss": 0.7936,
      "num_input_tokens_seen": 20334000,
      "step": 35045
    },
    {
      "epoch": 5.22043491212392,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02769249852937901,
      "loss": 0.8138,
      "num_input_tokens_seen": 20336976,
      "step": 35050
    },
    {
      "epoch": 5.221179624664879,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.027691459416064646,
      "loss": 0.8019,
      "num_input_tokens_seen": 20339952,
      "step": 35055
    },
    {
      "epoch": 5.221924337205839,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02769042008834098,
      "loss": 0.802,
      "num_input_tokens_seen": 20342832,
      "step": 35060
    },
    {
      "epoch": 5.222669049746798,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02768938054622556,
      "loss": 0.7866,
      "num_input_tokens_seen": 20345584,
      "step": 35065
    },
    {
      "epoch": 5.223413762287757,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02768834078973596,
      "loss": 0.809,
      "num_input_tokens_seen": 20348432,
      "step": 35070
    },
    {
      "epoch": 5.224158474828716,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.027687300818889732,
      "loss": 0.801,
      "num_input_tokens_seen": 20351632,
      "step": 35075
    },
    {
      "epoch": 5.224903187369676,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02768626063370445,
      "loss": 0.8058,
      "num_input_tokens_seen": 20354608,
      "step": 35080
    },
    {
      "epoch": 5.225647899910634,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.027685220234197692,
      "loss": 0.8178,
      "num_input_tokens_seen": 20357456,
      "step": 35085
    },
    {
      "epoch": 5.226392612451594,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02768417962038703,
      "loss": 0.8037,
      "num_input_tokens_seen": 20360784,
      "step": 35090
    },
    {
      "epoch": 5.227137324992553,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.027683138792290042,
      "loss": 0.7969,
      "num_input_tokens_seen": 20363600,
      "step": 35095
    },
    {
      "epoch": 5.227882037533512,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.027682097749924316,
      "loss": 0.8092,
      "num_input_tokens_seen": 20366608,
      "step": 35100
    },
    {
      "epoch": 5.228626750074471,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02768105649330744,
      "loss": 0.8005,
      "num_input_tokens_seen": 20369712,
      "step": 35105
    },
    {
      "epoch": 5.229371462615431,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.027680015022457007,
      "loss": 0.8035,
      "num_input_tokens_seen": 20372880,
      "step": 35110
    },
    {
      "epoch": 5.23011617515639,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0276789733373906,
      "loss": 0.8095,
      "num_input_tokens_seen": 20375760,
      "step": 35115
    },
    {
      "epoch": 5.230860887697349,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.027677931438125827,
      "loss": 0.7974,
      "num_input_tokens_seen": 20378704,
      "step": 35120
    },
    {
      "epoch": 5.231605600238308,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.02767688932468029,
      "loss": 0.7947,
      "num_input_tokens_seen": 20381360,
      "step": 35125
    },
    {
      "epoch": 5.232350312779268,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02767584699707159,
      "loss": 0.7916,
      "num_input_tokens_seen": 20384240,
      "step": 35130
    },
    {
      "epoch": 5.233095025320226,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.027674804455317337,
      "loss": 0.8184,
      "num_input_tokens_seen": 20387024,
      "step": 35135
    },
    {
      "epoch": 5.233839737861185,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.027673761699435147,
      "loss": 0.7906,
      "num_input_tokens_seen": 20390192,
      "step": 35140
    },
    {
      "epoch": 5.234584450402145,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.027672718729442634,
      "loss": 0.7877,
      "num_input_tokens_seen": 20393232,
      "step": 35145
    },
    {
      "epoch": 5.235329162943104,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.027671675545357415,
      "loss": 0.8172,
      "num_input_tokens_seen": 20396368,
      "step": 35150
    },
    {
      "epoch": 5.236073875484063,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02767063214719712,
      "loss": 0.8171,
      "num_input_tokens_seen": 20399280,
      "step": 35155
    },
    {
      "epoch": 5.236818588025022,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.027669588534979375,
      "loss": 0.7927,
      "num_input_tokens_seen": 20402256,
      "step": 35160
    },
    {
      "epoch": 5.237563300565982,
      "grad_norm": 0.01422119140625,
      "learning_rate": 0.027668544708721803,
      "loss": 0.8278,
      "num_input_tokens_seen": 20404880,
      "step": 35165
    },
    {
      "epoch": 5.23830801310694,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.02766750066844205,
      "loss": 0.7992,
      "num_input_tokens_seen": 20408016,
      "step": 35170
    },
    {
      "epoch": 5.2390527256479,
      "grad_norm": 0.01422119140625,
      "learning_rate": 0.027666456414157742,
      "loss": 0.8021,
      "num_input_tokens_seen": 20410832,
      "step": 35175
    },
    {
      "epoch": 5.239797438188859,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.027665411945886526,
      "loss": 0.7914,
      "num_input_tokens_seen": 20413520,
      "step": 35180
    },
    {
      "epoch": 5.240542150729818,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02766436726364605,
      "loss": 0.8042,
      "num_input_tokens_seen": 20416528,
      "step": 35185
    },
    {
      "epoch": 5.241286863270777,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02766332236745396,
      "loss": 0.8001,
      "num_input_tokens_seen": 20419504,
      "step": 35190
    },
    {
      "epoch": 5.242031575811737,
      "grad_norm": 0.011474609375,
      "learning_rate": 0.02766227725732791,
      "loss": 0.8096,
      "num_input_tokens_seen": 20422352,
      "step": 35195
    },
    {
      "epoch": 5.242776288352696,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.027661231933285554,
      "loss": 0.7905,
      "num_input_tokens_seen": 20425328,
      "step": 35200
    },
    {
      "epoch": 5.243521000893655,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.027660186395344556,
      "loss": 0.7958,
      "num_input_tokens_seen": 20428272,
      "step": 35205
    },
    {
      "epoch": 5.244265713434614,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.027659140643522574,
      "loss": 0.7801,
      "num_input_tokens_seen": 20431280,
      "step": 35210
    },
    {
      "epoch": 5.245010425975574,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.027658094677837277,
      "loss": 0.7914,
      "num_input_tokens_seen": 20433936,
      "step": 35215
    },
    {
      "epoch": 5.245755138516532,
      "grad_norm": 0.01544189453125,
      "learning_rate": 0.027657048498306334,
      "loss": 0.7739,
      "num_input_tokens_seen": 20437008,
      "step": 35220
    },
    {
      "epoch": 5.246499851057492,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.027656002104947425,
      "loss": 0.8152,
      "num_input_tokens_seen": 20439824,
      "step": 35225
    },
    {
      "epoch": 5.247244563598451,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02765495549777822,
      "loss": 0.8087,
      "num_input_tokens_seen": 20442640,
      "step": 35230
    },
    {
      "epoch": 5.2479892761394105,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.027653908676816403,
      "loss": 0.7873,
      "num_input_tokens_seen": 20445488,
      "step": 35235
    },
    {
      "epoch": 5.248733988680369,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.027652861642079665,
      "loss": 0.7886,
      "num_input_tokens_seen": 20448304,
      "step": 35240
    },
    {
      "epoch": 5.249478701221329,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.027651814393585678,
      "loss": 0.8002,
      "num_input_tokens_seen": 20451440,
      "step": 35245
    },
    {
      "epoch": 5.250223413762288,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.027650766931352155,
      "loss": 0.8342,
      "num_input_tokens_seen": 20454480,
      "step": 35250
    },
    {
      "epoch": 5.250968126303247,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.027649719255396783,
      "loss": 0.8277,
      "num_input_tokens_seen": 20457648,
      "step": 35255
    },
    {
      "epoch": 5.251712838844206,
      "grad_norm": 0.029296875,
      "learning_rate": 0.027648671365737256,
      "loss": 0.7961,
      "num_input_tokens_seen": 20460496,
      "step": 35260
    },
    {
      "epoch": 5.252457551385166,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.027647623262391285,
      "loss": 0.7956,
      "num_input_tokens_seen": 20463408,
      "step": 35265
    },
    {
      "epoch": 5.253202263926124,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.027646574945376572,
      "loss": 0.8054,
      "num_input_tokens_seen": 20466384,
      "step": 35270
    },
    {
      "epoch": 5.253946976467084,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02764552641471083,
      "loss": 0.7912,
      "num_input_tokens_seen": 20470064,
      "step": 35275
    },
    {
      "epoch": 5.254691689008043,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.027644477670411772,
      "loss": 0.82,
      "num_input_tokens_seen": 20473072,
      "step": 35280
    },
    {
      "epoch": 5.2554364015490025,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02764342871249711,
      "loss": 0.7996,
      "num_input_tokens_seen": 20476176,
      "step": 35285
    },
    {
      "epoch": 5.256181114089961,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.027642379540984575,
      "loss": 0.7955,
      "num_input_tokens_seen": 20479056,
      "step": 35290
    },
    {
      "epoch": 5.256925826630921,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.027641330155891885,
      "loss": 0.7786,
      "num_input_tokens_seen": 20481904,
      "step": 35295
    },
    {
      "epoch": 5.25767053917188,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02764028055723677,
      "loss": 0.8008,
      "num_input_tokens_seen": 20484656,
      "step": 35300
    },
    {
      "epoch": 5.258415251712838,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.027639230745036963,
      "loss": 0.7944,
      "num_input_tokens_seen": 20487632,
      "step": 35305
    },
    {
      "epoch": 5.259159964253798,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.0276381807193102,
      "loss": 0.7829,
      "num_input_tokens_seen": 20490448,
      "step": 35310
    },
    {
      "epoch": 5.259904676794757,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02763713048007422,
      "loss": 0.8146,
      "num_input_tokens_seen": 20493168,
      "step": 35315
    },
    {
      "epoch": 5.2606493893357165,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.027636080027346757,
      "loss": 0.7957,
      "num_input_tokens_seen": 20496016,
      "step": 35320
    },
    {
      "epoch": 5.261394101876675,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.027635029361145567,
      "loss": 0.7971,
      "num_input_tokens_seen": 20498992,
      "step": 35325
    },
    {
      "epoch": 5.262138814417635,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.0276339784814884,
      "loss": 0.7858,
      "num_input_tokens_seen": 20502256,
      "step": 35330
    },
    {
      "epoch": 5.262883526958594,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.027632927388393007,
      "loss": 0.7988,
      "num_input_tokens_seen": 20505328,
      "step": 35335
    },
    {
      "epoch": 5.263628239499553,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.027631876081877144,
      "loss": 0.8046,
      "num_input_tokens_seen": 20508272,
      "step": 35340
    },
    {
      "epoch": 5.264372952040512,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.027630824561958575,
      "loss": 0.8044,
      "num_input_tokens_seen": 20511280,
      "step": 35345
    },
    {
      "epoch": 5.265117664581472,
      "grad_norm": 0.01190185546875,
      "learning_rate": 0.02762977282865506,
      "loss": 0.7988,
      "num_input_tokens_seen": 20514256,
      "step": 35350
    },
    {
      "epoch": 5.26586237712243,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.027628720881984373,
      "loss": 0.7942,
      "num_input_tokens_seen": 20517296,
      "step": 35355
    },
    {
      "epoch": 5.26660708966339,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.027627668721964278,
      "loss": 0.8107,
      "num_input_tokens_seen": 20520304,
      "step": 35360
    },
    {
      "epoch": 5.267351802204349,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.027626616348612555,
      "loss": 0.8043,
      "num_input_tokens_seen": 20523248,
      "step": 35365
    },
    {
      "epoch": 5.2680965147453085,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.027625563761946987,
      "loss": 0.7961,
      "num_input_tokens_seen": 20526000,
      "step": 35370
    },
    {
      "epoch": 5.268841227286267,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.027624510961985344,
      "loss": 0.8264,
      "num_input_tokens_seen": 20528816,
      "step": 35375
    },
    {
      "epoch": 5.269585939827227,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027623457948745423,
      "loss": 0.7976,
      "num_input_tokens_seen": 20531440,
      "step": 35380
    },
    {
      "epoch": 5.270330652368186,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02762240472224501,
      "loss": 0.8096,
      "num_input_tokens_seen": 20534576,
      "step": 35385
    },
    {
      "epoch": 5.271075364909145,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.027621351282501896,
      "loss": 0.8029,
      "num_input_tokens_seen": 20537328,
      "step": 35390
    },
    {
      "epoch": 5.271820077450104,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02762029762953388,
      "loss": 0.8015,
      "num_input_tokens_seen": 20540272,
      "step": 35395
    },
    {
      "epoch": 5.272564789991064,
      "grad_norm": 0.025390625,
      "learning_rate": 0.027619243763358767,
      "loss": 0.8137,
      "num_input_tokens_seen": 20543088,
      "step": 35400
    },
    {
      "epoch": 5.2733095025320225,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.027618189683994352,
      "loss": 0.7914,
      "num_input_tokens_seen": 20546032,
      "step": 35405
    },
    {
      "epoch": 5.274054215072982,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.027617135391458447,
      "loss": 0.812,
      "num_input_tokens_seen": 20548784,
      "step": 35410
    },
    {
      "epoch": 5.274798927613941,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.027616080885768862,
      "loss": 0.7975,
      "num_input_tokens_seen": 20551856,
      "step": 35415
    },
    {
      "epoch": 5.2755436401549005,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.027615026166943415,
      "loss": 0.7968,
      "num_input_tokens_seen": 20554384,
      "step": 35420
    },
    {
      "epoch": 5.276288352695859,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.02761397123499992,
      "loss": 0.7827,
      "num_input_tokens_seen": 20557264,
      "step": 35425
    },
    {
      "epoch": 5.277033065236819,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.027612916089956203,
      "loss": 0.8072,
      "num_input_tokens_seen": 20560016,
      "step": 35430
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.027611860731830088,
      "loss": 0.8045,
      "num_input_tokens_seen": 20563216,
      "step": 35435
    },
    {
      "epoch": 5.278522490318737,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.027610805160639403,
      "loss": 0.8002,
      "num_input_tokens_seen": 20566000,
      "step": 35440
    },
    {
      "epoch": 5.279267202859696,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.027609749376401985,
      "loss": 0.7845,
      "num_input_tokens_seen": 20568752,
      "step": 35445
    },
    {
      "epoch": 5.280011915400656,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.027608693379135664,
      "loss": 0.8039,
      "num_input_tokens_seen": 20571664,
      "step": 35450
    },
    {
      "epoch": 5.2807566279416145,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.027607637168858286,
      "loss": 0.8119,
      "num_input_tokens_seen": 20574448,
      "step": 35455
    },
    {
      "epoch": 5.281501340482574,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.027606580745587687,
      "loss": 0.8328,
      "num_input_tokens_seen": 20577616,
      "step": 35460
    },
    {
      "epoch": 5.282246053023533,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.027605524109341723,
      "loss": 0.818,
      "num_input_tokens_seen": 20580848,
      "step": 35465
    },
    {
      "epoch": 5.282990765564492,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.027604467260138238,
      "loss": 0.795,
      "num_input_tokens_seen": 20583760,
      "step": 35470
    },
    {
      "epoch": 5.283735478105451,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.027603410197995094,
      "loss": 0.8034,
      "num_input_tokens_seen": 20586768,
      "step": 35475
    },
    {
      "epoch": 5.284480190646411,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.027602352922930135,
      "loss": 0.802,
      "num_input_tokens_seen": 20589712,
      "step": 35480
    },
    {
      "epoch": 5.28522490318737,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.027601295434961245,
      "loss": 0.7972,
      "num_input_tokens_seen": 20592368,
      "step": 35485
    },
    {
      "epoch": 5.2859696157283285,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.02760023773410627,
      "loss": 0.8105,
      "num_input_tokens_seen": 20595344,
      "step": 35490
    },
    {
      "epoch": 5.286714328269288,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.027599179820383082,
      "loss": 0.7914,
      "num_input_tokens_seen": 20598384,
      "step": 35495
    },
    {
      "epoch": 5.287459040810247,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.027598121693809558,
      "loss": 0.8035,
      "num_input_tokens_seen": 20601136,
      "step": 35500
    },
    {
      "epoch": 5.2882037533512065,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.027597063354403574,
      "loss": 0.8213,
      "num_input_tokens_seen": 20604176,
      "step": 35505
    },
    {
      "epoch": 5.288948465892165,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.027596004802183,
      "loss": 0.8166,
      "num_input_tokens_seen": 20607312,
      "step": 35510
    },
    {
      "epoch": 5.289693178433125,
      "grad_norm": 0.033203125,
      "learning_rate": 0.027594946037165738,
      "loss": 0.7933,
      "num_input_tokens_seen": 20610000,
      "step": 35515
    },
    {
      "epoch": 5.290437890974084,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.027593887059369656,
      "loss": 0.7901,
      "num_input_tokens_seen": 20612880,
      "step": 35520
    },
    {
      "epoch": 5.291182603515043,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.027592827868812654,
      "loss": 0.7897,
      "num_input_tokens_seen": 20616048,
      "step": 35525
    },
    {
      "epoch": 5.291927316056002,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.027591768465512624,
      "loss": 0.7916,
      "num_input_tokens_seen": 20618832,
      "step": 35530
    },
    {
      "epoch": 5.292672028596962,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.027590708849487465,
      "loss": 0.7951,
      "num_input_tokens_seen": 20621552,
      "step": 35535
    },
    {
      "epoch": 5.2934167411379205,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.027589649020755076,
      "loss": 0.7961,
      "num_input_tokens_seen": 20624624,
      "step": 35540
    },
    {
      "epoch": 5.29416145367888,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02758858897933336,
      "loss": 0.7829,
      "num_input_tokens_seen": 20627696,
      "step": 35545
    },
    {
      "epoch": 5.294906166219839,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.027587528725240233,
      "loss": 0.808,
      "num_input_tokens_seen": 20630704,
      "step": 35550
    },
    {
      "epoch": 5.2956508787607985,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.027586468258493597,
      "loss": 0.7798,
      "num_input_tokens_seen": 20633520,
      "step": 35555
    },
    {
      "epoch": 5.296395591301757,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.027585407579111373,
      "loss": 0.808,
      "num_input_tokens_seen": 20636336,
      "step": 35560
    },
    {
      "epoch": 5.297140303842717,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.027584346687111478,
      "loss": 0.8142,
      "num_input_tokens_seen": 20639120,
      "step": 35565
    },
    {
      "epoch": 5.297885016383676,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.027583285582511836,
      "loss": 0.8016,
      "num_input_tokens_seen": 20641904,
      "step": 35570
    },
    {
      "epoch": 5.298629728924635,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.027582224265330378,
      "loss": 0.7945,
      "num_input_tokens_seen": 20645008,
      "step": 35575
    },
    {
      "epoch": 5.299374441465594,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.027581162735585023,
      "loss": 0.8019,
      "num_input_tokens_seen": 20647984,
      "step": 35580
    },
    {
      "epoch": 5.300119154006554,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.027580100993293714,
      "loss": 0.7742,
      "num_input_tokens_seen": 20650928,
      "step": 35585
    },
    {
      "epoch": 5.3008638665475125,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.027579039038474387,
      "loss": 0.8067,
      "num_input_tokens_seen": 20653712,
      "step": 35590
    },
    {
      "epoch": 5.301608579088472,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.027577976871144976,
      "loss": 0.7782,
      "num_input_tokens_seen": 20656560,
      "step": 35595
    },
    {
      "epoch": 5.302353291629431,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.027576914491323427,
      "loss": 0.8178,
      "num_input_tokens_seen": 20659344,
      "step": 35600
    },
    {
      "epoch": 5.303098004170391,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.027575851899027694,
      "loss": 0.8241,
      "num_input_tokens_seen": 20662416,
      "step": 35605
    },
    {
      "epoch": 5.303842716711349,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.027574789094275725,
      "loss": 0.8115,
      "num_input_tokens_seen": 20665296,
      "step": 35610
    },
    {
      "epoch": 5.304587429252309,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.027573726077085473,
      "loss": 0.804,
      "num_input_tokens_seen": 20667920,
      "step": 35615
    },
    {
      "epoch": 5.305332141793268,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.0275726628474749,
      "loss": 0.8079,
      "num_input_tokens_seen": 20671120,
      "step": 35620
    },
    {
      "epoch": 5.306076854334227,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.027571599405461965,
      "loss": 0.8074,
      "num_input_tokens_seen": 20674096,
      "step": 35625
    },
    {
      "epoch": 5.306821566875186,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.027570535751064634,
      "loss": 0.7812,
      "num_input_tokens_seen": 20676784,
      "step": 35630
    },
    {
      "epoch": 5.307566279416146,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02756947188430088,
      "loss": 0.8056,
      "num_input_tokens_seen": 20679824,
      "step": 35635
    },
    {
      "epoch": 5.3083109919571045,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.027568407805188672,
      "loss": 0.7733,
      "num_input_tokens_seen": 20682544,
      "step": 35640
    },
    {
      "epoch": 5.309055704498064,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.027567343513745988,
      "loss": 0.7909,
      "num_input_tokens_seen": 20685488,
      "step": 35645
    },
    {
      "epoch": 5.309800417039023,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.027566279009990813,
      "loss": 0.7938,
      "num_input_tokens_seen": 20688304,
      "step": 35650
    },
    {
      "epoch": 5.310545129579982,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.027565214293941123,
      "loss": 0.8038,
      "num_input_tokens_seen": 20691024,
      "step": 35655
    },
    {
      "epoch": 5.311289842120941,
      "grad_norm": 0.00982666015625,
      "learning_rate": 0.027564149365614903,
      "loss": 0.7897,
      "num_input_tokens_seen": 20693840,
      "step": 35660
    },
    {
      "epoch": 5.3120345546619,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.02756308422503016,
      "loss": 0.7993,
      "num_input_tokens_seen": 20696816,
      "step": 35665
    },
    {
      "epoch": 5.31277926720286,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.027562018872204868,
      "loss": 0.756,
      "num_input_tokens_seen": 20699856,
      "step": 35670
    },
    {
      "epoch": 5.3135239797438185,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.027560953307157035,
      "loss": 0.7988,
      "num_input_tokens_seen": 20702544,
      "step": 35675
    },
    {
      "epoch": 5.314268692284778,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.027559887529904667,
      "loss": 0.8201,
      "num_input_tokens_seen": 20705264,
      "step": 35680
    },
    {
      "epoch": 5.315013404825737,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.027558821540465763,
      "loss": 0.7985,
      "num_input_tokens_seen": 20708080,
      "step": 35685
    },
    {
      "epoch": 5.315758117366697,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.027557755338858337,
      "loss": 0.7916,
      "num_input_tokens_seen": 20710864,
      "step": 35690
    },
    {
      "epoch": 5.316502829907655,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.027556688925100388,
      "loss": 0.8157,
      "num_input_tokens_seen": 20713840,
      "step": 35695
    },
    {
      "epoch": 5.317247542448615,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02755562229920995,
      "loss": 0.8263,
      "num_input_tokens_seen": 20716528,
      "step": 35700
    },
    {
      "epoch": 5.317992254989574,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02755455546120503,
      "loss": 0.8125,
      "num_input_tokens_seen": 20719216,
      "step": 35705
    },
    {
      "epoch": 5.318736967530533,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.027553488411103656,
      "loss": 0.7879,
      "num_input_tokens_seen": 20721904,
      "step": 35710
    },
    {
      "epoch": 5.319481680071492,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.027552421148923856,
      "loss": 0.8313,
      "num_input_tokens_seen": 20725072,
      "step": 35715
    },
    {
      "epoch": 5.320226392612452,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.027551353674683657,
      "loss": 0.8112,
      "num_input_tokens_seen": 20727760,
      "step": 35720
    },
    {
      "epoch": 5.3209711051534105,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.027550285988401094,
      "loss": 0.7944,
      "num_input_tokens_seen": 20730992,
      "step": 35725
    },
    {
      "epoch": 5.32171581769437,
      "grad_norm": 0.01190185546875,
      "learning_rate": 0.027549218090094207,
      "loss": 0.8135,
      "num_input_tokens_seen": 20733584,
      "step": 35730
    },
    {
      "epoch": 5.322460530235329,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02754814997978103,
      "loss": 0.8119,
      "num_input_tokens_seen": 20736784,
      "step": 35735
    },
    {
      "epoch": 5.323205242776289,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.027547081657479617,
      "loss": 0.7783,
      "num_input_tokens_seen": 20739856,
      "step": 35740
    },
    {
      "epoch": 5.323949955317247,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02754601312320801,
      "loss": 0.8063,
      "num_input_tokens_seen": 20742928,
      "step": 35745
    },
    {
      "epoch": 5.324694667858207,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02754494437698426,
      "loss": 0.7837,
      "num_input_tokens_seen": 20746096,
      "step": 35750
    },
    {
      "epoch": 5.325439380399166,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02754387541882643,
      "loss": 0.7864,
      "num_input_tokens_seen": 20749008,
      "step": 35755
    },
    {
      "epoch": 5.326184092940125,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02754280624875257,
      "loss": 0.7951,
      "num_input_tokens_seen": 20752016,
      "step": 35760
    },
    {
      "epoch": 5.326928805481084,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02754173686678075,
      "loss": 0.8162,
      "num_input_tokens_seen": 20755248,
      "step": 35765
    },
    {
      "epoch": 5.327673518022044,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.027540667272929028,
      "loss": 0.826,
      "num_input_tokens_seen": 20758224,
      "step": 35770
    },
    {
      "epoch": 5.328418230563003,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02753959746721548,
      "loss": 0.8009,
      "num_input_tokens_seen": 20761232,
      "step": 35775
    },
    {
      "epoch": 5.329162943103962,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02753852744965818,
      "loss": 0.8101,
      "num_input_tokens_seen": 20764272,
      "step": 35780
    },
    {
      "epoch": 5.329907655644921,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.0275374572202752,
      "loss": 0.8004,
      "num_input_tokens_seen": 20767024,
      "step": 35785
    },
    {
      "epoch": 5.330652368185881,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.027536386779084623,
      "loss": 0.8086,
      "num_input_tokens_seen": 20769968,
      "step": 35790
    },
    {
      "epoch": 5.331397080726839,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.02753531612610453,
      "loss": 0.7994,
      "num_input_tokens_seen": 20772816,
      "step": 35795
    },
    {
      "epoch": 5.332141793267799,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.027534245261353018,
      "loss": 0.7982,
      "num_input_tokens_seen": 20775568,
      "step": 35800
    },
    {
      "epoch": 5.332886505808758,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027533174184848165,
      "loss": 0.8039,
      "num_input_tokens_seen": 20778384,
      "step": 35805
    },
    {
      "epoch": 5.333631218349717,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02753210289660808,
      "loss": 0.7986,
      "num_input_tokens_seen": 20780944,
      "step": 35810
    },
    {
      "epoch": 5.334375930890676,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.027531031396650845,
      "loss": 0.785,
      "num_input_tokens_seen": 20783792,
      "step": 35815
    },
    {
      "epoch": 5.335120643431635,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.027529959684994578,
      "loss": 0.8152,
      "num_input_tokens_seen": 20786704,
      "step": 35820
    },
    {
      "epoch": 5.335865355972595,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02752888776165737,
      "loss": 0.8023,
      "num_input_tokens_seen": 20789904,
      "step": 35825
    },
    {
      "epoch": 5.336610068513554,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.027527815626657338,
      "loss": 0.7955,
      "num_input_tokens_seen": 20792912,
      "step": 35830
    },
    {
      "epoch": 5.337354781054513,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.027526743280012596,
      "loss": 0.8002,
      "num_input_tokens_seen": 20795728,
      "step": 35835
    },
    {
      "epoch": 5.338099493595472,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02752567072174126,
      "loss": 0.7873,
      "num_input_tokens_seen": 20798480,
      "step": 35840
    },
    {
      "epoch": 5.338844206136431,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.027524597951861445,
      "loss": 0.8023,
      "num_input_tokens_seen": 20801552,
      "step": 35845
    },
    {
      "epoch": 5.33958891867739,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.02752352497039128,
      "loss": 0.7978,
      "num_input_tokens_seen": 20804496,
      "step": 35850
    },
    {
      "epoch": 5.34033363121835,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.027522451777348883,
      "loss": 0.8413,
      "num_input_tokens_seen": 20807472,
      "step": 35855
    },
    {
      "epoch": 5.341078343759309,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.027521378372752394,
      "loss": 0.7997,
      "num_input_tokens_seen": 20810448,
      "step": 35860
    },
    {
      "epoch": 5.341823056300268,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.027520304756619946,
      "loss": 0.8135,
      "num_input_tokens_seen": 20813488,
      "step": 35865
    },
    {
      "epoch": 5.342567768841227,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.027519230928969674,
      "loss": 0.8294,
      "num_input_tokens_seen": 20816208,
      "step": 35870
    },
    {
      "epoch": 5.343312481382187,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.027518156889819716,
      "loss": 0.7882,
      "num_input_tokens_seen": 20819120,
      "step": 35875
    },
    {
      "epoch": 5.344057193923145,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.027517082639188218,
      "loss": 0.8223,
      "num_input_tokens_seen": 20822128,
      "step": 35880
    },
    {
      "epoch": 5.344801906464105,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.027516008177093335,
      "loss": 0.7983,
      "num_input_tokens_seen": 20825168,
      "step": 35885
    },
    {
      "epoch": 5.345546619005064,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.027514933503553215,
      "loss": 0.8238,
      "num_input_tokens_seen": 20827824,
      "step": 35890
    },
    {
      "epoch": 5.346291331546023,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02751385861858601,
      "loss": 0.8079,
      "num_input_tokens_seen": 20830576,
      "step": 35895
    },
    {
      "epoch": 5.347036044086982,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.027512783522209885,
      "loss": 0.7932,
      "num_input_tokens_seen": 20833328,
      "step": 35900
    },
    {
      "epoch": 5.347780756627942,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.027511708214442997,
      "loss": 0.8097,
      "num_input_tokens_seen": 20836208,
      "step": 35905
    },
    {
      "epoch": 5.348525469168901,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.027510632695303516,
      "loss": 0.8046,
      "num_input_tokens_seen": 20838896,
      "step": 35910
    },
    {
      "epoch": 5.34927018170986,
      "grad_norm": 0.01544189453125,
      "learning_rate": 0.02750955696480961,
      "loss": 0.7995,
      "num_input_tokens_seen": 20842128,
      "step": 35915
    },
    {
      "epoch": 5.350014894250819,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.027508481022979456,
      "loss": 0.8063,
      "num_input_tokens_seen": 20845104,
      "step": 35920
    },
    {
      "epoch": 5.350759606791779,
      "grad_norm": 0.0123291015625,
      "learning_rate": 0.027507404869831227,
      "loss": 0.8113,
      "num_input_tokens_seen": 20848144,
      "step": 35925
    },
    {
      "epoch": 5.351504319332737,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.027506328505383103,
      "loss": 0.7926,
      "num_input_tokens_seen": 20851088,
      "step": 35930
    },
    {
      "epoch": 5.352249031873697,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.027505251929653272,
      "loss": 0.7909,
      "num_input_tokens_seen": 20853968,
      "step": 35935
    },
    {
      "epoch": 5.352993744414656,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02750417514265992,
      "loss": 0.802,
      "num_input_tokens_seen": 20856944,
      "step": 35940
    },
    {
      "epoch": 5.3537384569556155,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02750309814442124,
      "loss": 0.7984,
      "num_input_tokens_seen": 20859856,
      "step": 35945
    },
    {
      "epoch": 5.354483169496574,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.027502020934955417,
      "loss": 0.7989,
      "num_input_tokens_seen": 20862608,
      "step": 35950
    },
    {
      "epoch": 5.355227882037534,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02750094351428066,
      "loss": 0.7986,
      "num_input_tokens_seen": 20865616,
      "step": 35955
    },
    {
      "epoch": 5.355972594578493,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02749986588241517,
      "loss": 0.8019,
      "num_input_tokens_seen": 20868368,
      "step": 35960
    },
    {
      "epoch": 5.356717307119452,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.027498788039377147,
      "loss": 0.7947,
      "num_input_tokens_seen": 20871568,
      "step": 35965
    },
    {
      "epoch": 5.357462019660411,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.027497709985184804,
      "loss": 0.7733,
      "num_input_tokens_seen": 20874320,
      "step": 35970
    },
    {
      "epoch": 5.358206732201371,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.027496631719856352,
      "loss": 0.83,
      "num_input_tokens_seen": 20877296,
      "step": 35975
    },
    {
      "epoch": 5.358951444742329,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.027495553243410013,
      "loss": 0.8156,
      "num_input_tokens_seen": 20879984,
      "step": 35980
    },
    {
      "epoch": 5.359696157283288,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.027494474555864002,
      "loss": 0.8026,
      "num_input_tokens_seen": 20882896,
      "step": 35985
    },
    {
      "epoch": 5.360440869824248,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.027493395657236536,
      "loss": 0.8004,
      "num_input_tokens_seen": 20885968,
      "step": 35990
    },
    {
      "epoch": 5.3611855823652075,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02749231654754585,
      "loss": 0.7979,
      "num_input_tokens_seen": 20888784,
      "step": 35995
    },
    {
      "epoch": 5.361930294906166,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027491237226810173,
      "loss": 0.8004,
      "num_input_tokens_seen": 20891888,
      "step": 36000
    },
    {
      "epoch": 5.362675007447125,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.027490157695047743,
      "loss": 0.8089,
      "num_input_tokens_seen": 20894672,
      "step": 36005
    },
    {
      "epoch": 5.363419719988085,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02748907795227679,
      "loss": 0.8141,
      "num_input_tokens_seen": 20897712,
      "step": 36010
    },
    {
      "epoch": 5.364164432529043,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.027487997998515555,
      "loss": 0.7844,
      "num_input_tokens_seen": 20900688,
      "step": 36015
    },
    {
      "epoch": 5.364909145070003,
      "grad_norm": 0.01123046875,
      "learning_rate": 0.027486917833782286,
      "loss": 0.7876,
      "num_input_tokens_seen": 20903632,
      "step": 36020
    },
    {
      "epoch": 5.365653857610962,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.027485837458095234,
      "loss": 0.8159,
      "num_input_tokens_seen": 20906448,
      "step": 36025
    },
    {
      "epoch": 5.3663985701519215,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.027484756871472654,
      "loss": 0.8023,
      "num_input_tokens_seen": 20909648,
      "step": 36030
    },
    {
      "epoch": 5.36714328269288,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.027483676073932786,
      "loss": 0.8054,
      "num_input_tokens_seen": 20912688,
      "step": 36035
    },
    {
      "epoch": 5.36788799523384,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.027482595065493904,
      "loss": 0.7901,
      "num_input_tokens_seen": 20915536,
      "step": 36040
    },
    {
      "epoch": 5.368632707774799,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02748151384617426,
      "loss": 0.8387,
      "num_input_tokens_seen": 20918416,
      "step": 36045
    },
    {
      "epoch": 5.369377420315758,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02748043241599213,
      "loss": 0.8174,
      "num_input_tokens_seen": 20921104,
      "step": 36050
    },
    {
      "epoch": 5.370122132856717,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02747935077496578,
      "loss": 0.8222,
      "num_input_tokens_seen": 20924400,
      "step": 36055
    },
    {
      "epoch": 5.370866845397677,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.027478268923113475,
      "loss": 0.804,
      "num_input_tokens_seen": 20927440,
      "step": 36060
    },
    {
      "epoch": 5.371611557938635,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.027477186860453508,
      "loss": 0.7977,
      "num_input_tokens_seen": 20930192,
      "step": 36065
    },
    {
      "epoch": 5.372356270479595,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.027476104587004144,
      "loss": 0.8328,
      "num_input_tokens_seen": 20933136,
      "step": 36070
    },
    {
      "epoch": 5.373100983020554,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.027475022102783676,
      "loss": 0.7978,
      "num_input_tokens_seen": 20936048,
      "step": 36075
    },
    {
      "epoch": 5.3738456955615135,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.027473939407810385,
      "loss": 0.7902,
      "num_input_tokens_seen": 20939120,
      "step": 36080
    },
    {
      "epoch": 5.374590408102472,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.027472856502102568,
      "loss": 0.7824,
      "num_input_tokens_seen": 20941808,
      "step": 36085
    },
    {
      "epoch": 5.375335120643432,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02747177338567852,
      "loss": 0.7999,
      "num_input_tokens_seen": 20944624,
      "step": 36090
    },
    {
      "epoch": 5.376079833184391,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.027470690058556534,
      "loss": 0.8008,
      "num_input_tokens_seen": 20947568,
      "step": 36095
    },
    {
      "epoch": 5.37682454572535,
      "grad_norm": 0.01092529296875,
      "learning_rate": 0.027469606520754912,
      "loss": 0.801,
      "num_input_tokens_seen": 20950416,
      "step": 36100
    },
    {
      "epoch": 5.377569258266309,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.027468522772291965,
      "loss": 0.8096,
      "num_input_tokens_seen": 20953360,
      "step": 36105
    },
    {
      "epoch": 5.378313970807269,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.027467438813185995,
      "loss": 0.7822,
      "num_input_tokens_seen": 20956144,
      "step": 36110
    },
    {
      "epoch": 5.3790586833482275,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.027466354643455322,
      "loss": 0.8018,
      "num_input_tokens_seen": 20958928,
      "step": 36115
    },
    {
      "epoch": 5.379803395889187,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02746527026311825,
      "loss": 0.7901,
      "num_input_tokens_seen": 20961904,
      "step": 36120
    },
    {
      "epoch": 5.380548108430146,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.027464185672193112,
      "loss": 0.8087,
      "num_input_tokens_seen": 20964592,
      "step": 36125
    },
    {
      "epoch": 5.3812928209711055,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.027463100870698226,
      "loss": 0.7937,
      "num_input_tokens_seen": 20967376,
      "step": 36130
    },
    {
      "epoch": 5.382037533512064,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.027462015858651913,
      "loss": 0.7957,
      "num_input_tokens_seen": 20970352,
      "step": 36135
    },
    {
      "epoch": 5.382782246053024,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.027460930636072512,
      "loss": 0.8044,
      "num_input_tokens_seen": 20973168,
      "step": 36140
    },
    {
      "epoch": 5.383526958593983,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.027459845202978353,
      "loss": 0.7801,
      "num_input_tokens_seen": 20976304,
      "step": 36145
    },
    {
      "epoch": 5.384271671134942,
      "grad_norm": 0.033203125,
      "learning_rate": 0.027458759559387768,
      "loss": 0.8004,
      "num_input_tokens_seen": 20979376,
      "step": 36150
    },
    {
      "epoch": 5.385016383675901,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.027457673705319104,
      "loss": 0.8046,
      "num_input_tokens_seen": 20982160,
      "step": 36155
    },
    {
      "epoch": 5.385761096216861,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.027456587640790706,
      "loss": 0.7977,
      "num_input_tokens_seen": 20984880,
      "step": 36160
    },
    {
      "epoch": 5.3865058087578195,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.027455501365820922,
      "loss": 0.7933,
      "num_input_tokens_seen": 20987696,
      "step": 36165
    },
    {
      "epoch": 5.387250521298778,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.0274544148804281,
      "loss": 0.7756,
      "num_input_tokens_seen": 20990704,
      "step": 36170
    },
    {
      "epoch": 5.387995233839738,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027453328184630598,
      "loss": 0.8094,
      "num_input_tokens_seen": 20993680,
      "step": 36175
    },
    {
      "epoch": 5.388739946380697,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.027452241278446768,
      "loss": 0.809,
      "num_input_tokens_seen": 20996816,
      "step": 36180
    },
    {
      "epoch": 5.389484658921656,
      "grad_norm": 0.010498046875,
      "learning_rate": 0.027451154161894982,
      "loss": 0.7828,
      "num_input_tokens_seen": 20999568,
      "step": 36185
    },
    {
      "epoch": 5.390229371462615,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.0274500668349936,
      "loss": 0.792,
      "num_input_tokens_seen": 21002256,
      "step": 36190
    },
    {
      "epoch": 5.390974084003575,
      "grad_norm": 0.02734375,
      "learning_rate": 0.027448979297760995,
      "loss": 0.781,
      "num_input_tokens_seen": 21005040,
      "step": 36195
    },
    {
      "epoch": 5.3917187965445335,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.027447891550215535,
      "loss": 0.8235,
      "num_input_tokens_seen": 21008208,
      "step": 36200
    },
    {
      "epoch": 5.392463509085493,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.027446803592375595,
      "loss": 0.8344,
      "num_input_tokens_seen": 21011312,
      "step": 36205
    },
    {
      "epoch": 5.393208221626452,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.027445715424259564,
      "loss": 0.8026,
      "num_input_tokens_seen": 21014192,
      "step": 36210
    },
    {
      "epoch": 5.3939529341674115,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.027444627045885822,
      "loss": 0.8084,
      "num_input_tokens_seen": 21017200,
      "step": 36215
    },
    {
      "epoch": 5.39469764670837,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02744353845727275,
      "loss": 0.7957,
      "num_input_tokens_seen": 21019856,
      "step": 36220
    },
    {
      "epoch": 5.39544235924933,
      "grad_norm": 0.013671875,
      "learning_rate": 0.027442449658438745,
      "loss": 0.8156,
      "num_input_tokens_seen": 21022800,
      "step": 36225
    },
    {
      "epoch": 5.396187071790289,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.027441360649402197,
      "loss": 0.8113,
      "num_input_tokens_seen": 21025840,
      "step": 36230
    },
    {
      "epoch": 5.396931784331248,
      "grad_norm": 0.01104736328125,
      "learning_rate": 0.027440271430181507,
      "loss": 0.7809,
      "num_input_tokens_seen": 21028752,
      "step": 36235
    },
    {
      "epoch": 5.397676496872207,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02743918200079507,
      "loss": 0.7978,
      "num_input_tokens_seen": 21031760,
      "step": 36240
    },
    {
      "epoch": 5.398421209413167,
      "grad_norm": 0.0234375,
      "learning_rate": 0.0274380923612613,
      "loss": 0.8114,
      "num_input_tokens_seen": 21034800,
      "step": 36245
    },
    {
      "epoch": 5.3991659219541255,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.027437002511598602,
      "loss": 0.8076,
      "num_input_tokens_seen": 21037360,
      "step": 36250
    },
    {
      "epoch": 5.399910634495085,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.027435912451825386,
      "loss": 0.8101,
      "num_input_tokens_seen": 21040336,
      "step": 36255
    },
    {
      "epoch": 5.400655347036044,
      "grad_norm": 0.017578125,
      "learning_rate": 0.02743482218196007,
      "loss": 0.7969,
      "num_input_tokens_seen": 21043184,
      "step": 36260
    },
    {
      "epoch": 5.4014000595770035,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.02743373170202107,
      "loss": 0.7912,
      "num_input_tokens_seen": 21046096,
      "step": 36265
    },
    {
      "epoch": 5.402144772117962,
      "grad_norm": 0.0234375,
      "learning_rate": 0.027432641012026807,
      "loss": 0.831,
      "num_input_tokens_seen": 21048784,
      "step": 36270
    },
    {
      "epoch": 5.402889484658922,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.027431550111995714,
      "loss": 0.8051,
      "num_input_tokens_seen": 21051376,
      "step": 36275
    },
    {
      "epoch": 5.403634197199881,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.02743045900194622,
      "loss": 0.7932,
      "num_input_tokens_seen": 21054128,
      "step": 36280
    },
    {
      "epoch": 5.40437890974084,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.027429367681896747,
      "loss": 0.8182,
      "num_input_tokens_seen": 21056944,
      "step": 36285
    },
    {
      "epoch": 5.405123622281799,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.027428276151865744,
      "loss": 0.8041,
      "num_input_tokens_seen": 21059696,
      "step": 36290
    },
    {
      "epoch": 5.405868334822759,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.027427184411871645,
      "loss": 0.7953,
      "num_input_tokens_seen": 21062736,
      "step": 36295
    },
    {
      "epoch": 5.4066130473637175,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.0274260924619329,
      "loss": 0.7992,
      "num_input_tokens_seen": 21066160,
      "step": 36300
    },
    {
      "epoch": 5.407357759904677,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02742500030206795,
      "loss": 0.7837,
      "num_input_tokens_seen": 21068944,
      "step": 36305
    },
    {
      "epoch": 5.408102472445636,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.027423907932295242,
      "loss": 0.823,
      "num_input_tokens_seen": 21071856,
      "step": 36310
    },
    {
      "epoch": 5.408847184986596,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.027422815352633245,
      "loss": 0.782,
      "num_input_tokens_seen": 21074992,
      "step": 36315
    },
    {
      "epoch": 5.409591897527554,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.027421722563100402,
      "loss": 0.7902,
      "num_input_tokens_seen": 21077744,
      "step": 36320
    },
    {
      "epoch": 5.410336610068514,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.027420629563715188,
      "loss": 0.7943,
      "num_input_tokens_seen": 21080272,
      "step": 36325
    },
    {
      "epoch": 5.411081322609473,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.027419536354496055,
      "loss": 0.8095,
      "num_input_tokens_seen": 21083376,
      "step": 36330
    },
    {
      "epoch": 5.4118260351504315,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.027418442935461487,
      "loss": 0.7943,
      "num_input_tokens_seen": 21086096,
      "step": 36335
    },
    {
      "epoch": 5.412570747691391,
      "grad_norm": 0.01019287109375,
      "learning_rate": 0.027417349306629937,
      "loss": 0.7785,
      "num_input_tokens_seen": 21089072,
      "step": 36340
    },
    {
      "epoch": 5.413315460232351,
      "grad_norm": 0.01190185546875,
      "learning_rate": 0.027416255468019894,
      "loss": 0.8222,
      "num_input_tokens_seen": 21092112,
      "step": 36345
    },
    {
      "epoch": 5.4140601727733095,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.027415161419649836,
      "loss": 0.7984,
      "num_input_tokens_seen": 21094768,
      "step": 36350
    },
    {
      "epoch": 5.414804885314268,
      "grad_norm": 0.010986328125,
      "learning_rate": 0.02741406716153824,
      "loss": 0.8374,
      "num_input_tokens_seen": 21097744,
      "step": 36355
    },
    {
      "epoch": 5.415549597855228,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.0274129726937036,
      "loss": 0.8241,
      "num_input_tokens_seen": 21100848,
      "step": 36360
    },
    {
      "epoch": 5.416294310396187,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.027411878016164404,
      "loss": 0.7861,
      "num_input_tokens_seen": 21103952,
      "step": 36365
    },
    {
      "epoch": 5.417039022937146,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02741078312893914,
      "loss": 0.7923,
      "num_input_tokens_seen": 21107088,
      "step": 36370
    },
    {
      "epoch": 5.417783735478105,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.027409688032046315,
      "loss": 0.7821,
      "num_input_tokens_seen": 21109680,
      "step": 36375
    },
    {
      "epoch": 5.418528448019065,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.027408592725504416,
      "loss": 0.7822,
      "num_input_tokens_seen": 21112432,
      "step": 36380
    },
    {
      "epoch": 5.4192731605600235,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.027407497209331957,
      "loss": 0.8202,
      "num_input_tokens_seen": 21115408,
      "step": 36385
    },
    {
      "epoch": 5.420017873100983,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.027406401483547446,
      "loss": 0.8363,
      "num_input_tokens_seen": 21118416,
      "step": 36390
    },
    {
      "epoch": 5.420762585641942,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.027405305548169383,
      "loss": 0.7899,
      "num_input_tokens_seen": 21121328,
      "step": 36395
    },
    {
      "epoch": 5.421507298182902,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.0274042094032163,
      "loss": 0.7963,
      "num_input_tokens_seen": 21124144,
      "step": 36400
    },
    {
      "epoch": 5.42225201072386,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.0274031130487067,
      "loss": 0.791,
      "num_input_tokens_seen": 21126928,
      "step": 36405
    },
    {
      "epoch": 5.42299672326482,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.027402016484659116,
      "loss": 0.8134,
      "num_input_tokens_seen": 21129360,
      "step": 36410
    },
    {
      "epoch": 5.423741435805779,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.027400919711092066,
      "loss": 0.7955,
      "num_input_tokens_seen": 21132144,
      "step": 36415
    },
    {
      "epoch": 5.424486148346738,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.02739982272802408,
      "loss": 0.7894,
      "num_input_tokens_seen": 21135088,
      "step": 36420
    },
    {
      "epoch": 5.425230860887697,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.027398725535473692,
      "loss": 0.8251,
      "num_input_tokens_seen": 21137840,
      "step": 36425
    },
    {
      "epoch": 5.425975573428657,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.027397628133459438,
      "loss": 0.7925,
      "num_input_tokens_seen": 21140528,
      "step": 36430
    },
    {
      "epoch": 5.4267202859696155,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02739653052199986,
      "loss": 0.8001,
      "num_input_tokens_seen": 21143248,
      "step": 36435
    },
    {
      "epoch": 5.427464998510575,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.027395432701113494,
      "loss": 0.7994,
      "num_input_tokens_seen": 21146320,
      "step": 36440
    },
    {
      "epoch": 5.428209711051534,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.027394334670818893,
      "loss": 0.8048,
      "num_input_tokens_seen": 21148880,
      "step": 36445
    },
    {
      "epoch": 5.428954423592494,
      "grad_norm": 0.01422119140625,
      "learning_rate": 0.027393236431134605,
      "loss": 0.81,
      "num_input_tokens_seen": 21151792,
      "step": 36450
    },
    {
      "epoch": 5.429699136133452,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.027392137982079182,
      "loss": 0.8029,
      "num_input_tokens_seen": 21154672,
      "step": 36455
    },
    {
      "epoch": 5.430443848674412,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.027391039323671183,
      "loss": 0.8208,
      "num_input_tokens_seen": 21157552,
      "step": 36460
    },
    {
      "epoch": 5.431188561215371,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.02738994045592917,
      "loss": 0.8067,
      "num_input_tokens_seen": 21160400,
      "step": 36465
    },
    {
      "epoch": 5.43193327375633,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.027388841378871702,
      "loss": 0.8056,
      "num_input_tokens_seen": 21163312,
      "step": 36470
    },
    {
      "epoch": 5.432677986297289,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.027387742092517355,
      "loss": 0.8021,
      "num_input_tokens_seen": 21166064,
      "step": 36475
    },
    {
      "epoch": 5.433422698838249,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02738664259688469,
      "loss": 0.7906,
      "num_input_tokens_seen": 21168784,
      "step": 36480
    },
    {
      "epoch": 5.434167411379208,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02738554289199229,
      "loss": 0.803,
      "num_input_tokens_seen": 21171824,
      "step": 36485
    },
    {
      "epoch": 5.434912123920167,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.027384442977858733,
      "loss": 0.8191,
      "num_input_tokens_seen": 21175440,
      "step": 36490
    },
    {
      "epoch": 5.435656836461126,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.027383342854502595,
      "loss": 0.805,
      "num_input_tokens_seen": 21178320,
      "step": 36495
    },
    {
      "epoch": 5.436401549002086,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02738224252194247,
      "loss": 0.8045,
      "num_input_tokens_seen": 21181136,
      "step": 36500
    },
    {
      "epoch": 5.437146261543044,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02738114198019694,
      "loss": 0.8038,
      "num_input_tokens_seen": 21183920,
      "step": 36505
    },
    {
      "epoch": 5.437890974084004,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.027380041229284603,
      "loss": 0.8004,
      "num_input_tokens_seen": 21186704,
      "step": 36510
    },
    {
      "epoch": 5.438635686624963,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.027378940269224047,
      "loss": 0.7922,
      "num_input_tokens_seen": 21189552,
      "step": 36515
    },
    {
      "epoch": 5.4393803991659215,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.027377839100033878,
      "loss": 0.797,
      "num_input_tokens_seen": 21192336,
      "step": 36520
    },
    {
      "epoch": 5.440125111706881,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.027376737721732696,
      "loss": 0.7996,
      "num_input_tokens_seen": 21195376,
      "step": 36525
    },
    {
      "epoch": 5.44086982424784,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.027375636134339115,
      "loss": 0.7731,
      "num_input_tokens_seen": 21198192,
      "step": 36530
    },
    {
      "epoch": 5.4416145367888,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.027374534337871733,
      "loss": 0.8034,
      "num_input_tokens_seen": 21201168,
      "step": 36535
    },
    {
      "epoch": 5.442359249329758,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.027373432332349176,
      "loss": 0.7873,
      "num_input_tokens_seen": 21203984,
      "step": 36540
    },
    {
      "epoch": 5.443103961870718,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.027372330117790052,
      "loss": 0.7972,
      "num_input_tokens_seen": 21206608,
      "step": 36545
    },
    {
      "epoch": 5.443848674411677,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.027371227694212987,
      "loss": 0.795,
      "num_input_tokens_seen": 21209680,
      "step": 36550
    },
    {
      "epoch": 5.444593386952636,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.0273701250616366,
      "loss": 0.8087,
      "num_input_tokens_seen": 21212752,
      "step": 36555
    },
    {
      "epoch": 5.445338099493595,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.027369022220079524,
      "loss": 0.8481,
      "num_input_tokens_seen": 21215536,
      "step": 36560
    },
    {
      "epoch": 5.446082812034555,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.027367919169560394,
      "loss": 0.8001,
      "num_input_tokens_seen": 21218224,
      "step": 36565
    },
    {
      "epoch": 5.446827524575514,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.027366815910097833,
      "loss": 0.8045,
      "num_input_tokens_seen": 21221168,
      "step": 36570
    },
    {
      "epoch": 5.447572237116473,
      "grad_norm": 0.017578125,
      "learning_rate": 0.027365712441710488,
      "loss": 0.7862,
      "num_input_tokens_seen": 21223984,
      "step": 36575
    },
    {
      "epoch": 5.448316949657432,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.027364608764417004,
      "loss": 0.8001,
      "num_input_tokens_seen": 21227088,
      "step": 36580
    },
    {
      "epoch": 5.449061662198392,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.027363504878236015,
      "loss": 0.7711,
      "num_input_tokens_seen": 21230032,
      "step": 36585
    },
    {
      "epoch": 5.44980637473935,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02736240078318618,
      "loss": 0.8055,
      "num_input_tokens_seen": 21233072,
      "step": 36590
    },
    {
      "epoch": 5.45055108728031,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.027361296479286153,
      "loss": 0.8268,
      "num_input_tokens_seen": 21236080,
      "step": 36595
    },
    {
      "epoch": 5.451295799821269,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02736019196655458,
      "loss": 0.7913,
      "num_input_tokens_seen": 21239056,
      "step": 36600
    },
    {
      "epoch": 5.452040512362228,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.027359087245010127,
      "loss": 0.8051,
      "num_input_tokens_seen": 21241872,
      "step": 36605
    },
    {
      "epoch": 5.452785224903187,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.027357982314671455,
      "loss": 0.812,
      "num_input_tokens_seen": 21244752,
      "step": 36610
    },
    {
      "epoch": 5.453529937444147,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.027356877175557234,
      "loss": 0.8132,
      "num_input_tokens_seen": 21247408,
      "step": 36615
    },
    {
      "epoch": 5.454274649985106,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02735577182768613,
      "loss": 0.8109,
      "num_input_tokens_seen": 21250032,
      "step": 36620
    },
    {
      "epoch": 5.455019362526065,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02735466627107682,
      "loss": 0.8009,
      "num_input_tokens_seen": 21253136,
      "step": 36625
    },
    {
      "epoch": 5.455764075067024,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02735356050574798,
      "loss": 0.8025,
      "num_input_tokens_seen": 21256016,
      "step": 36630
    },
    {
      "epoch": 5.456508787607984,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.027352454531718293,
      "loss": 0.8116,
      "num_input_tokens_seen": 21259248,
      "step": 36635
    },
    {
      "epoch": 5.457253500148942,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02735134834900644,
      "loss": 0.7941,
      "num_input_tokens_seen": 21262192,
      "step": 36640
    },
    {
      "epoch": 5.457998212689902,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.027350241957631104,
      "loss": 0.8149,
      "num_input_tokens_seen": 21264912,
      "step": 36645
    },
    {
      "epoch": 5.458742925230861,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.027349135357610986,
      "loss": 0.7904,
      "num_input_tokens_seen": 21267824,
      "step": 36650
    },
    {
      "epoch": 5.4594876377718204,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.027348028548964783,
      "loss": 0.8027,
      "num_input_tokens_seen": 21270672,
      "step": 36655
    },
    {
      "epoch": 5.460232350312779,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.027346921531711182,
      "loss": 0.7849,
      "num_input_tokens_seen": 21273488,
      "step": 36660
    },
    {
      "epoch": 5.460977062853739,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.027345814305868892,
      "loss": 0.8132,
      "num_input_tokens_seen": 21276688,
      "step": 36665
    },
    {
      "epoch": 5.461721775394698,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.027344706871456615,
      "loss": 0.8118,
      "num_input_tokens_seen": 21279696,
      "step": 36670
    },
    {
      "epoch": 5.462466487935657,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.027343599228493064,
      "loss": 0.8083,
      "num_input_tokens_seen": 21282704,
      "step": 36675
    },
    {
      "epoch": 5.463211200476616,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02734249137699695,
      "loss": 0.8003,
      "num_input_tokens_seen": 21285648,
      "step": 36680
    },
    {
      "epoch": 5.463955913017575,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.027341383316986986,
      "loss": 0.8311,
      "num_input_tokens_seen": 21288880,
      "step": 36685
    },
    {
      "epoch": 5.464700625558534,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.027340275048481896,
      "loss": 0.819,
      "num_input_tokens_seen": 21291888,
      "step": 36690
    },
    {
      "epoch": 5.465445338099494,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.027339166571500404,
      "loss": 0.8137,
      "num_input_tokens_seen": 21294832,
      "step": 36695
    },
    {
      "epoch": 5.466190050640453,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02733805788606123,
      "loss": 0.7976,
      "num_input_tokens_seen": 21297392,
      "step": 36700
    },
    {
      "epoch": 5.466934763181412,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02733694899218311,
      "loss": 0.7955,
      "num_input_tokens_seen": 21300496,
      "step": 36705
    },
    {
      "epoch": 5.467679475722371,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.027335839889884774,
      "loss": 0.7977,
      "num_input_tokens_seen": 21303376,
      "step": 36710
    },
    {
      "epoch": 5.46842418826333,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.027334730579184962,
      "loss": 0.8184,
      "num_input_tokens_seen": 21306288,
      "step": 36715
    },
    {
      "epoch": 5.46916890080429,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.027333621060102408,
      "loss": 0.8087,
      "num_input_tokens_seen": 21309232,
      "step": 36720
    },
    {
      "epoch": 5.469913613345248,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.02733251133265587,
      "loss": 0.8014,
      "num_input_tokens_seen": 21312176,
      "step": 36725
    },
    {
      "epoch": 5.470658325886208,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02733140139686408,
      "loss": 0.7915,
      "num_input_tokens_seen": 21315184,
      "step": 36730
    },
    {
      "epoch": 5.471403038427167,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0273302912527458,
      "loss": 0.7885,
      "num_input_tokens_seen": 21318352,
      "step": 36735
    },
    {
      "epoch": 5.4721477509681264,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.027329180900319782,
      "loss": 0.8084,
      "num_input_tokens_seen": 21321584,
      "step": 36740
    },
    {
      "epoch": 5.472892463509085,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02732807033960478,
      "loss": 0.7889,
      "num_input_tokens_seen": 21324400,
      "step": 36745
    },
    {
      "epoch": 5.473637176050045,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.027326959570619563,
      "loss": 0.8181,
      "num_input_tokens_seen": 21327312,
      "step": 36750
    },
    {
      "epoch": 5.474381888591004,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02732584859338289,
      "loss": 0.7881,
      "num_input_tokens_seen": 21330160,
      "step": 36755
    },
    {
      "epoch": 5.475126601131963,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02732473740791353,
      "loss": 0.7917,
      "num_input_tokens_seen": 21332944,
      "step": 36760
    },
    {
      "epoch": 5.475871313672922,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.027323626014230263,
      "loss": 0.7898,
      "num_input_tokens_seen": 21335920,
      "step": 36765
    },
    {
      "epoch": 5.476616026213882,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02732251441235186,
      "loss": 0.8043,
      "num_input_tokens_seen": 21338864,
      "step": 36770
    },
    {
      "epoch": 5.47736073875484,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.0273214026022971,
      "loss": 0.7903,
      "num_input_tokens_seen": 21341872,
      "step": 36775
    },
    {
      "epoch": 5.4781054512958,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.027320290584084764,
      "loss": 0.7893,
      "num_input_tokens_seen": 21344432,
      "step": 36780
    },
    {
      "epoch": 5.478850163836759,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.027319178357733637,
      "loss": 0.7816,
      "num_input_tokens_seen": 21347248,
      "step": 36785
    },
    {
      "epoch": 5.4795948763777185,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.027318065923262518,
      "loss": 0.8245,
      "num_input_tokens_seen": 21349968,
      "step": 36790
    },
    {
      "epoch": 5.480339588918677,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.027316953280690195,
      "loss": 0.7865,
      "num_input_tokens_seen": 21352784,
      "step": 36795
    },
    {
      "epoch": 5.481084301459637,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.02731584043003546,
      "loss": 0.786,
      "num_input_tokens_seen": 21355632,
      "step": 36800
    },
    {
      "epoch": 5.481829014000596,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.02731472737131712,
      "loss": 0.8014,
      "num_input_tokens_seen": 21358288,
      "step": 36805
    },
    {
      "epoch": 5.482573726541555,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.027313614104553977,
      "loss": 0.7978,
      "num_input_tokens_seen": 21361232,
      "step": 36810
    },
    {
      "epoch": 5.483318439082514,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.027312500629764837,
      "loss": 0.7901,
      "num_input_tokens_seen": 21364080,
      "step": 36815
    },
    {
      "epoch": 5.484063151623474,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.027311386946968514,
      "loss": 0.8136,
      "num_input_tokens_seen": 21366928,
      "step": 36820
    },
    {
      "epoch": 5.4848078641644324,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.027310273056183824,
      "loss": 0.7917,
      "num_input_tokens_seen": 21369840,
      "step": 36825
    },
    {
      "epoch": 5.485552576705392,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02730915895742958,
      "loss": 0.8118,
      "num_input_tokens_seen": 21372624,
      "step": 36830
    },
    {
      "epoch": 5.486297289246351,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.027308044650724606,
      "loss": 0.7967,
      "num_input_tokens_seen": 21375536,
      "step": 36835
    },
    {
      "epoch": 5.4870420017873105,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027306930136087728,
      "loss": 0.7961,
      "num_input_tokens_seen": 21378544,
      "step": 36840
    },
    {
      "epoch": 5.487786714328269,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.027305815413537774,
      "loss": 0.8264,
      "num_input_tokens_seen": 21381200,
      "step": 36845
    },
    {
      "epoch": 5.488531426869228,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.027304700483093573,
      "loss": 0.8063,
      "num_input_tokens_seen": 21384656,
      "step": 36850
    },
    {
      "epoch": 5.489276139410188,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.027303585344773968,
      "loss": 0.7639,
      "num_input_tokens_seen": 21387440,
      "step": 36855
    },
    {
      "epoch": 5.490020851951147,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.02730246999859779,
      "loss": 0.7706,
      "num_input_tokens_seen": 21390128,
      "step": 36860
    },
    {
      "epoch": 5.490765564492106,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.027301354444583883,
      "loss": 0.8071,
      "num_input_tokens_seen": 21392912,
      "step": 36865
    },
    {
      "epoch": 5.491510277033065,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.027300238682751096,
      "loss": 0.8288,
      "num_input_tokens_seen": 21396048,
      "step": 36870
    },
    {
      "epoch": 5.4922549895740245,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.027299122713118285,
      "loss": 0.7908,
      "num_input_tokens_seen": 21398832,
      "step": 36875
    },
    {
      "epoch": 5.492999702114983,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02729800653570429,
      "loss": 0.8128,
      "num_input_tokens_seen": 21401616,
      "step": 36880
    },
    {
      "epoch": 5.493744414655943,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.027296890150527973,
      "loss": 0.817,
      "num_input_tokens_seen": 21404464,
      "step": 36885
    },
    {
      "epoch": 5.494489127196902,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.027295773557608195,
      "loss": 0.7882,
      "num_input_tokens_seen": 21407536,
      "step": 36890
    },
    {
      "epoch": 5.495233839737861,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02729465675696382,
      "loss": 0.8156,
      "num_input_tokens_seen": 21410448,
      "step": 36895
    },
    {
      "epoch": 5.49597855227882,
      "grad_norm": 0.025390625,
      "learning_rate": 0.027293539748613715,
      "loss": 0.8001,
      "num_input_tokens_seen": 21413200,
      "step": 36900
    },
    {
      "epoch": 5.49672326481978,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02729242253257675,
      "loss": 0.8057,
      "num_input_tokens_seen": 21415952,
      "step": 36905
    },
    {
      "epoch": 5.4974679773607384,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.027291305108871802,
      "loss": 0.7785,
      "num_input_tokens_seen": 21419312,
      "step": 36910
    },
    {
      "epoch": 5.498212689901698,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.027290187477517745,
      "loss": 0.7891,
      "num_input_tokens_seen": 21422320,
      "step": 36915
    },
    {
      "epoch": 5.498957402442657,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.027289069638533465,
      "loss": 0.8058,
      "num_input_tokens_seen": 21425392,
      "step": 36920
    },
    {
      "epoch": 5.4997021149836165,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.02728795159193784,
      "loss": 0.8052,
      "num_input_tokens_seen": 21428176,
      "step": 36925
    },
    {
      "epoch": 5.500446827524575,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02728683333774976,
      "loss": 0.7826,
      "num_input_tokens_seen": 21431344,
      "step": 36930
    },
    {
      "epoch": 5.501191540065535,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.027285714875988118,
      "loss": 0.7795,
      "num_input_tokens_seen": 21434288,
      "step": 36935
    },
    {
      "epoch": 5.501936252606494,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.027284596206671816,
      "loss": 0.7827,
      "num_input_tokens_seen": 21437104,
      "step": 36940
    },
    {
      "epoch": 5.502680965147453,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02728347732981974,
      "loss": 0.8112,
      "num_input_tokens_seen": 21440272,
      "step": 36945
    },
    {
      "epoch": 5.503425677688412,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.0272823582454508,
      "loss": 0.8063,
      "num_input_tokens_seen": 21442896,
      "step": 36950
    },
    {
      "epoch": 5.504170390229372,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.027281238953583906,
      "loss": 0.8011,
      "num_input_tokens_seen": 21445584,
      "step": 36955
    },
    {
      "epoch": 5.5049151027703305,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.02728011945423796,
      "loss": 0.8104,
      "num_input_tokens_seen": 21448400,
      "step": 36960
    },
    {
      "epoch": 5.50565981531129,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.027278999747431875,
      "loss": 0.7772,
      "num_input_tokens_seen": 21451088,
      "step": 36965
    },
    {
      "epoch": 5.506404527852249,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.027277879833184568,
      "loss": 0.8105,
      "num_input_tokens_seen": 21454160,
      "step": 36970
    },
    {
      "epoch": 5.5071492403932085,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.027276759711514963,
      "loss": 0.822,
      "num_input_tokens_seen": 21456752,
      "step": 36975
    },
    {
      "epoch": 5.507893952934167,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.027275639382441978,
      "loss": 0.8018,
      "num_input_tokens_seen": 21459920,
      "step": 36980
    },
    {
      "epoch": 5.508638665475127,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.027274518845984544,
      "loss": 0.8111,
      "num_input_tokens_seen": 21462864,
      "step": 36985
    },
    {
      "epoch": 5.509383378016086,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02727339810216159,
      "loss": 0.7922,
      "num_input_tokens_seen": 21465616,
      "step": 36990
    },
    {
      "epoch": 5.510128090557045,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.027272277150992046,
      "loss": 0.8005,
      "num_input_tokens_seen": 21468784,
      "step": 36995
    },
    {
      "epoch": 5.510872803098004,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.027271155992494855,
      "loss": 0.7755,
      "num_input_tokens_seen": 21471600,
      "step": 37000
    },
    {
      "epoch": 5.511617515638964,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.027270034626688953,
      "loss": 0.7848,
      "num_input_tokens_seen": 21474192,
      "step": 37005
    },
    {
      "epoch": 5.5123622281799225,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.027268913053593285,
      "loss": 0.8138,
      "num_input_tokens_seen": 21476752,
      "step": 37010
    },
    {
      "epoch": 5.513106940720881,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02726779127322681,
      "loss": 0.8282,
      "num_input_tokens_seen": 21479856,
      "step": 37015
    },
    {
      "epoch": 5.513851653261841,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02726666928560846,
      "loss": 0.7815,
      "num_input_tokens_seen": 21482640,
      "step": 37020
    },
    {
      "epoch": 5.5145963658028005,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.027265547090757205,
      "loss": 0.7897,
      "num_input_tokens_seen": 21485424,
      "step": 37025
    },
    {
      "epoch": 5.515341078343759,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.027264424688691995,
      "loss": 0.8037,
      "num_input_tokens_seen": 21488368,
      "step": 37030
    },
    {
      "epoch": 5.516085790884718,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.027263302079431798,
      "loss": 0.8143,
      "num_input_tokens_seen": 21491088,
      "step": 37035
    },
    {
      "epoch": 5.516830503425678,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.027262179262995567,
      "loss": 0.834,
      "num_input_tokens_seen": 21494288,
      "step": 37040
    },
    {
      "epoch": 5.517575215966637,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02726105623940229,
      "loss": 0.8133,
      "num_input_tokens_seen": 21497328,
      "step": 37045
    },
    {
      "epoch": 5.518319928507596,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02725993300867093,
      "loss": 0.8135,
      "num_input_tokens_seen": 21500240,
      "step": 37050
    },
    {
      "epoch": 5.519064641048555,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.027258809570820455,
      "loss": 0.8047,
      "num_input_tokens_seen": 21503632,
      "step": 37055
    },
    {
      "epoch": 5.5198093535895145,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.027257685925869857,
      "loss": 0.8111,
      "num_input_tokens_seen": 21506640,
      "step": 37060
    },
    {
      "epoch": 5.520554066130473,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.027256562073838114,
      "loss": 0.7845,
      "num_input_tokens_seen": 21509296,
      "step": 37065
    },
    {
      "epoch": 5.521298778671433,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.027255438014744206,
      "loss": 0.8005,
      "num_input_tokens_seen": 21512080,
      "step": 37070
    },
    {
      "epoch": 5.522043491212392,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.027254313748607133,
      "loss": 0.8237,
      "num_input_tokens_seen": 21515376,
      "step": 37075
    },
    {
      "epoch": 5.522788203753351,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.027253189275445885,
      "loss": 0.7949,
      "num_input_tokens_seen": 21518096,
      "step": 37080
    },
    {
      "epoch": 5.52353291629431,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.027252064595279454,
      "loss": 0.8099,
      "num_input_tokens_seen": 21521232,
      "step": 37085
    },
    {
      "epoch": 5.52427762883527,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.027250939708126847,
      "loss": 0.7772,
      "num_input_tokens_seen": 21524080,
      "step": 37090
    },
    {
      "epoch": 5.5250223413762285,
      "grad_norm": 0.017578125,
      "learning_rate": 0.027249814614007065,
      "loss": 0.7948,
      "num_input_tokens_seen": 21527248,
      "step": 37095
    },
    {
      "epoch": 5.525767053917188,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.027248689312939115,
      "loss": 0.7811,
      "num_input_tokens_seen": 21530128,
      "step": 37100
    },
    {
      "epoch": 5.526511766458147,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.027247563804942004,
      "loss": 0.8042,
      "num_input_tokens_seen": 21532720,
      "step": 37105
    },
    {
      "epoch": 5.5272564789991065,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02724643809003476,
      "loss": 0.7757,
      "num_input_tokens_seen": 21535984,
      "step": 37110
    },
    {
      "epoch": 5.528001191540065,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02724531216823638,
      "loss": 0.7839,
      "num_input_tokens_seen": 21538672,
      "step": 37115
    },
    {
      "epoch": 5.528745904081025,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.027244186039565903,
      "loss": 0.7914,
      "num_input_tokens_seen": 21541328,
      "step": 37120
    },
    {
      "epoch": 5.529490616621984,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.027243059704042345,
      "loss": 0.7986,
      "num_input_tokens_seen": 21544176,
      "step": 37125
    },
    {
      "epoch": 5.530235329162943,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.027241933161684738,
      "loss": 0.8328,
      "num_input_tokens_seen": 21546928,
      "step": 37130
    },
    {
      "epoch": 5.530980041703902,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02724080641251211,
      "loss": 0.8132,
      "num_input_tokens_seen": 21549904,
      "step": 37135
    },
    {
      "epoch": 5.531724754244862,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.027239679456543504,
      "loss": 0.8083,
      "num_input_tokens_seen": 21553136,
      "step": 37140
    },
    {
      "epoch": 5.5324694667858205,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.027238552293797947,
      "loss": 0.8204,
      "num_input_tokens_seen": 21556144,
      "step": 37145
    },
    {
      "epoch": 5.53321417932678,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02723742492429449,
      "loss": 0.7684,
      "num_input_tokens_seen": 21559024,
      "step": 37150
    },
    {
      "epoch": 5.533958891867739,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.027236297348052177,
      "loss": 0.7954,
      "num_input_tokens_seen": 21562224,
      "step": 37155
    },
    {
      "epoch": 5.534703604408699,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.027235169565090058,
      "loss": 0.8345,
      "num_input_tokens_seen": 21565264,
      "step": 37160
    },
    {
      "epoch": 5.535448316949657,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.027234041575427186,
      "loss": 0.808,
      "num_input_tokens_seen": 21568400,
      "step": 37165
    },
    {
      "epoch": 5.536193029490617,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.027232913379082613,
      "loss": 0.8198,
      "num_input_tokens_seen": 21571216,
      "step": 37170
    },
    {
      "epoch": 5.536937742031576,
      "grad_norm": 0.0234375,
      "learning_rate": 0.0272317849760754,
      "loss": 0.8017,
      "num_input_tokens_seen": 21574192,
      "step": 37175
    },
    {
      "epoch": 5.537682454572535,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02723065636642461,
      "loss": 0.8163,
      "num_input_tokens_seen": 21577040,
      "step": 37180
    },
    {
      "epoch": 5.538427167113494,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.027229527550149315,
      "loss": 0.8103,
      "num_input_tokens_seen": 21579920,
      "step": 37185
    },
    {
      "epoch": 5.539171879654454,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027228398527268584,
      "loss": 0.8029,
      "num_input_tokens_seen": 21582608,
      "step": 37190
    },
    {
      "epoch": 5.5399165921954125,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.027227269297801483,
      "loss": 0.8073,
      "num_input_tokens_seen": 21585520,
      "step": 37195
    },
    {
      "epoch": 5.540661304736371,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.0272261398617671,
      "loss": 0.7954,
      "num_input_tokens_seen": 21588496,
      "step": 37200
    },
    {
      "epoch": 5.541406017277331,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.027225010219184505,
      "loss": 0.8123,
      "num_input_tokens_seen": 21591536,
      "step": 37205
    },
    {
      "epoch": 5.542150729818291,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.027223880370072787,
      "loss": 0.7905,
      "num_input_tokens_seen": 21594192,
      "step": 37210
    },
    {
      "epoch": 5.542895442359249,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.027222750314451035,
      "loss": 0.8044,
      "num_input_tokens_seen": 21597136,
      "step": 37215
    },
    {
      "epoch": 5.543640154900208,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.027221620052338344,
      "loss": 0.8105,
      "num_input_tokens_seen": 21599984,
      "step": 37220
    },
    {
      "epoch": 5.544384867441168,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.027220489583753796,
      "loss": 0.7953,
      "num_input_tokens_seen": 21602736,
      "step": 37225
    },
    {
      "epoch": 5.5451295799821265,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.027219358908716504,
      "loss": 0.802,
      "num_input_tokens_seen": 21605616,
      "step": 37230
    },
    {
      "epoch": 5.545874292523086,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.027218228027245557,
      "loss": 0.8201,
      "num_input_tokens_seen": 21608912,
      "step": 37235
    },
    {
      "epoch": 5.546619005064045,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02721709693936007,
      "loss": 0.7976,
      "num_input_tokens_seen": 21611920,
      "step": 37240
    },
    {
      "epoch": 5.547363717605005,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.027215965645079145,
      "loss": 0.7928,
      "num_input_tokens_seen": 21614704,
      "step": 37245
    },
    {
      "epoch": 5.548108430145963,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.027214834144421895,
      "loss": 0.8271,
      "num_input_tokens_seen": 21617648,
      "step": 37250
    },
    {
      "epoch": 5.548853142686923,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.027213702437407433,
      "loss": 0.7873,
      "num_input_tokens_seen": 21620432,
      "step": 37255
    },
    {
      "epoch": 5.549597855227882,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02721257052405489,
      "loss": 0.8105,
      "num_input_tokens_seen": 21623216,
      "step": 37260
    },
    {
      "epoch": 5.550342567768841,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.027211438404383375,
      "loss": 0.7938,
      "num_input_tokens_seen": 21626064,
      "step": 37265
    },
    {
      "epoch": 5.5510872803098,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.027210306078412023,
      "loss": 0.8094,
      "num_input_tokens_seen": 21629328,
      "step": 37270
    },
    {
      "epoch": 5.55183199285076,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.027209173546159957,
      "loss": 0.8078,
      "num_input_tokens_seen": 21632240,
      "step": 37275
    },
    {
      "epoch": 5.5525767053917185,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.027208040807646317,
      "loss": 0.7845,
      "num_input_tokens_seen": 21635024,
      "step": 37280
    },
    {
      "epoch": 5.553321417932678,
      "grad_norm": 0.025390625,
      "learning_rate": 0.027206907862890227,
      "loss": 0.7913,
      "num_input_tokens_seen": 21637872,
      "step": 37285
    },
    {
      "epoch": 5.554066130473637,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02720577471191084,
      "loss": 0.7957,
      "num_input_tokens_seen": 21640944,
      "step": 37290
    },
    {
      "epoch": 5.554810843014597,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.027204641354727298,
      "loss": 0.7838,
      "num_input_tokens_seen": 21644048,
      "step": 37295
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02720350779135874,
      "loss": 0.7839,
      "num_input_tokens_seen": 21646832,
      "step": 37300
    },
    {
      "epoch": 5.556300268096515,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02720237402182432,
      "loss": 0.8184,
      "num_input_tokens_seen": 21649680,
      "step": 37305
    },
    {
      "epoch": 5.557044980637474,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0272012400461432,
      "loss": 0.7917,
      "num_input_tokens_seen": 21652464,
      "step": 37310
    },
    {
      "epoch": 5.557789693178433,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.027200105864334523,
      "loss": 0.8407,
      "num_input_tokens_seen": 21655344,
      "step": 37315
    },
    {
      "epoch": 5.558534405719392,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02719897147641746,
      "loss": 0.8127,
      "num_input_tokens_seen": 21658480,
      "step": 37320
    },
    {
      "epoch": 5.559279118260352,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.02719783688241117,
      "loss": 0.7741,
      "num_input_tokens_seen": 21661232,
      "step": 37325
    },
    {
      "epoch": 5.560023830801311,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.027196702082334823,
      "loss": 0.8365,
      "num_input_tokens_seen": 21664112,
      "step": 37330
    },
    {
      "epoch": 5.56076854334227,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.027195567076207593,
      "loss": 0.7886,
      "num_input_tokens_seen": 21666640,
      "step": 37335
    },
    {
      "epoch": 5.561513255883229,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.027194431864048654,
      "loss": 0.7904,
      "num_input_tokens_seen": 21669968,
      "step": 37340
    },
    {
      "epoch": 5.562257968424189,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02719329644587718,
      "loss": 0.798,
      "num_input_tokens_seen": 21672912,
      "step": 37345
    },
    {
      "epoch": 5.563002680965147,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.027192160821712354,
      "loss": 0.811,
      "num_input_tokens_seen": 21675760,
      "step": 37350
    },
    {
      "epoch": 5.563747393506107,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.027191024991573366,
      "loss": 0.7917,
      "num_input_tokens_seen": 21678480,
      "step": 37355
    },
    {
      "epoch": 5.564492106047066,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.0271898889554794,
      "loss": 0.837,
      "num_input_tokens_seen": 21681360,
      "step": 37360
    },
    {
      "epoch": 5.5652368185880245,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.027188752713449647,
      "loss": 0.8118,
      "num_input_tokens_seen": 21684496,
      "step": 37365
    },
    {
      "epoch": 5.565981531128984,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.027187616265503308,
      "loss": 0.8327,
      "num_input_tokens_seen": 21687152,
      "step": 37370
    },
    {
      "epoch": 5.566726243669944,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027186479611659577,
      "loss": 0.7972,
      "num_input_tokens_seen": 21689936,
      "step": 37375
    },
    {
      "epoch": 5.567470956210903,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.027185342751937663,
      "loss": 0.8217,
      "num_input_tokens_seen": 21692592,
      "step": 37380
    },
    {
      "epoch": 5.568215668751861,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02718420568635676,
      "loss": 0.8071,
      "num_input_tokens_seen": 21695440,
      "step": 37385
    },
    {
      "epoch": 5.568960381292821,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.027183068414936094,
      "loss": 0.7851,
      "num_input_tokens_seen": 21698640,
      "step": 37390
    },
    {
      "epoch": 5.569705093833781,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.027181930937694863,
      "loss": 0.7886,
      "num_input_tokens_seen": 21701328,
      "step": 37395
    },
    {
      "epoch": 5.570449806374739,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.027180793254652287,
      "loss": 0.8214,
      "num_input_tokens_seen": 21704112,
      "step": 37400
    },
    {
      "epoch": 5.571194518915698,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.027179655365827598,
      "loss": 0.8004,
      "num_input_tokens_seen": 21706736,
      "step": 37405
    },
    {
      "epoch": 5.571939231456658,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.027178517271240005,
      "loss": 0.7882,
      "num_input_tokens_seen": 21709360,
      "step": 37410
    },
    {
      "epoch": 5.572683943997617,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02717737897090874,
      "loss": 0.8204,
      "num_input_tokens_seen": 21711920,
      "step": 37415
    },
    {
      "epoch": 5.573428656538576,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02717624046485304,
      "loss": 0.7926,
      "num_input_tokens_seen": 21714928,
      "step": 37420
    },
    {
      "epoch": 5.574173369079535,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02717510175309213,
      "loss": 0.8038,
      "num_input_tokens_seen": 21717872,
      "step": 37425
    },
    {
      "epoch": 5.574918081620495,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.027173962835645243,
      "loss": 0.8018,
      "num_input_tokens_seen": 21721072,
      "step": 37430
    },
    {
      "epoch": 5.575662794161453,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.027172823712531633,
      "loss": 0.8177,
      "num_input_tokens_seen": 21724208,
      "step": 37435
    },
    {
      "epoch": 5.576407506702413,
      "grad_norm": 0.025390625,
      "learning_rate": 0.027171684383770538,
      "loss": 0.8049,
      "num_input_tokens_seen": 21727248,
      "step": 37440
    },
    {
      "epoch": 5.577152219243372,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.027170544849381206,
      "loss": 0.7916,
      "num_input_tokens_seen": 21730448,
      "step": 37445
    },
    {
      "epoch": 5.577896931784331,
      "grad_norm": 0.013671875,
      "learning_rate": 0.02716940510938289,
      "loss": 0.8149,
      "num_input_tokens_seen": 21733488,
      "step": 37450
    },
    {
      "epoch": 5.57864164432529,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02716826516379484,
      "loss": 0.7991,
      "num_input_tokens_seen": 21736432,
      "step": 37455
    },
    {
      "epoch": 5.57938635686625,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02716712501263632,
      "loss": 0.8101,
      "num_input_tokens_seen": 21739184,
      "step": 37460
    },
    {
      "epoch": 5.580131069407209,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.027165984655926585,
      "loss": 0.7882,
      "num_input_tokens_seen": 21741776,
      "step": 37465
    },
    {
      "epoch": 5.580875781948168,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.027164844093684908,
      "loss": 0.7825,
      "num_input_tokens_seen": 21744528,
      "step": 37470
    },
    {
      "epoch": 5.581620494489127,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.027163703325930554,
      "loss": 0.8138,
      "num_input_tokens_seen": 21747312,
      "step": 37475
    },
    {
      "epoch": 5.582365207030087,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02716256235268279,
      "loss": 0.8057,
      "num_input_tokens_seen": 21750096,
      "step": 37480
    },
    {
      "epoch": 5.583109919571045,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.027161421173960903,
      "loss": 0.8083,
      "num_input_tokens_seen": 21753104,
      "step": 37485
    },
    {
      "epoch": 5.583854632112005,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.027160279789784164,
      "loss": 0.8168,
      "num_input_tokens_seen": 21755888,
      "step": 37490
    },
    {
      "epoch": 5.584599344652964,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.02715913820017185,
      "loss": 0.7994,
      "num_input_tokens_seen": 21758736,
      "step": 37495
    },
    {
      "epoch": 5.5853440571939235,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.027157996405143263,
      "loss": 0.7984,
      "num_input_tokens_seen": 21761616,
      "step": 37500
    },
    {
      "epoch": 5.586088769734882,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02715685440471768,
      "loss": 0.8104,
      "num_input_tokens_seen": 21764336,
      "step": 37505
    },
    {
      "epoch": 5.586833482275842,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.0271557121989144,
      "loss": 0.8003,
      "num_input_tokens_seen": 21767280,
      "step": 37510
    },
    {
      "epoch": 5.587578194816801,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.027154569787752714,
      "loss": 0.8151,
      "num_input_tokens_seen": 21770224,
      "step": 37515
    },
    {
      "epoch": 5.58832290735776,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.027153427171251923,
      "loss": 0.8117,
      "num_input_tokens_seen": 21773200,
      "step": 37520
    },
    {
      "epoch": 5.589067619898719,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02715228434943134,
      "loss": 0.7792,
      "num_input_tokens_seen": 21776176,
      "step": 37525
    },
    {
      "epoch": 5.589812332439678,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02715114132231026,
      "loss": 0.7798,
      "num_input_tokens_seen": 21778928,
      "step": 37530
    },
    {
      "epoch": 5.590557044980637,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.027149998089907992,
      "loss": 0.8047,
      "num_input_tokens_seen": 21782192,
      "step": 37535
    },
    {
      "epoch": 5.591301757521597,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02714885465224386,
      "loss": 0.798,
      "num_input_tokens_seen": 21785040,
      "step": 37540
    },
    {
      "epoch": 5.592046470062556,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.027147711009337175,
      "loss": 0.7923,
      "num_input_tokens_seen": 21787792,
      "step": 37545
    },
    {
      "epoch": 5.592791182603515,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.02714656716120726,
      "loss": 0.8058,
      "num_input_tokens_seen": 21790576,
      "step": 37550
    },
    {
      "epoch": 5.593535895144474,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.027145423107873438,
      "loss": 0.8065,
      "num_input_tokens_seen": 21793360,
      "step": 37555
    },
    {
      "epoch": 5.594280607685434,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.027144278849355035,
      "loss": 0.793,
      "num_input_tokens_seen": 21796304,
      "step": 37560
    },
    {
      "epoch": 5.595025320226393,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.027143134385671387,
      "loss": 0.8038,
      "num_input_tokens_seen": 21799568,
      "step": 37565
    },
    {
      "epoch": 5.595770032767351,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02714198971684182,
      "loss": 0.8035,
      "num_input_tokens_seen": 21802384,
      "step": 37570
    },
    {
      "epoch": 5.596514745308311,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.02714084484288568,
      "loss": 0.801,
      "num_input_tokens_seen": 21805264,
      "step": 37575
    },
    {
      "epoch": 5.59725945784927,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.027139699763822304,
      "loss": 0.7944,
      "num_input_tokens_seen": 21808048,
      "step": 37580
    },
    {
      "epoch": 5.5980041703902295,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.027138554479671038,
      "loss": 0.8161,
      "num_input_tokens_seen": 21810992,
      "step": 37585
    },
    {
      "epoch": 5.598748882931188,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.027137408990451237,
      "loss": 0.7767,
      "num_input_tokens_seen": 21813648,
      "step": 37590
    },
    {
      "epoch": 5.599493595472148,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.027136263296182246,
      "loss": 0.8009,
      "num_input_tokens_seen": 21816624,
      "step": 37595
    },
    {
      "epoch": 5.600238308013107,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.027135117396883415,
      "loss": 0.7897,
      "num_input_tokens_seen": 21819600,
      "step": 37600
    },
    {
      "epoch": 5.600983020554066,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.027133971292574114,
      "loss": 0.8459,
      "num_input_tokens_seen": 21822480,
      "step": 37605
    },
    {
      "epoch": 5.601727733095025,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.027132824983273702,
      "loss": 0.8148,
      "num_input_tokens_seen": 21825328,
      "step": 37610
    },
    {
      "epoch": 5.602472445635985,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02713167846900154,
      "loss": 0.7789,
      "num_input_tokens_seen": 21827952,
      "step": 37615
    },
    {
      "epoch": 5.603217158176943,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.027130531749777,
      "loss": 0.799,
      "num_input_tokens_seen": 21830640,
      "step": 37620
    },
    {
      "epoch": 5.603961870717903,
      "grad_norm": 0.01953125,
      "learning_rate": 0.027129384825619458,
      "loss": 0.8033,
      "num_input_tokens_seen": 21833584,
      "step": 37625
    },
    {
      "epoch": 5.604706583258862,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.027128237696548287,
      "loss": 0.8223,
      "num_input_tokens_seen": 21836336,
      "step": 37630
    },
    {
      "epoch": 5.6054512957998215,
      "grad_norm": 0.017578125,
      "learning_rate": 0.027127090362582863,
      "loss": 0.8038,
      "num_input_tokens_seen": 21839248,
      "step": 37635
    },
    {
      "epoch": 5.60619600834078,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.027125942823742574,
      "loss": 0.7945,
      "num_input_tokens_seen": 21842096,
      "step": 37640
    },
    {
      "epoch": 5.60694072088174,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.027124795080046806,
      "loss": 0.8301,
      "num_input_tokens_seen": 21845040,
      "step": 37645
    },
    {
      "epoch": 5.607685433422699,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02712364713151495,
      "loss": 0.7913,
      "num_input_tokens_seen": 21847760,
      "step": 37650
    },
    {
      "epoch": 5.608430145963658,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.027122498978166393,
      "loss": 0.7924,
      "num_input_tokens_seen": 21851088,
      "step": 37655
    },
    {
      "epoch": 5.609174858504617,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.02712135062002054,
      "loss": 0.7976,
      "num_input_tokens_seen": 21854096,
      "step": 37660
    },
    {
      "epoch": 5.609919571045577,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.02712020205709679,
      "loss": 0.8052,
      "num_input_tokens_seen": 21856944,
      "step": 37665
    },
    {
      "epoch": 5.6106642835865355,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.027119053289414538,
      "loss": 0.7808,
      "num_input_tokens_seen": 21859792,
      "step": 37670
    },
    {
      "epoch": 5.611408996127495,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.027117904316993204,
      "loss": 0.8224,
      "num_input_tokens_seen": 21862896,
      "step": 37675
    },
    {
      "epoch": 5.612153708668454,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.027116755139852188,
      "loss": 0.7913,
      "num_input_tokens_seen": 21865808,
      "step": 37680
    },
    {
      "epoch": 5.6128984212094135,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.027115605758010915,
      "loss": 0.7965,
      "num_input_tokens_seen": 21868720,
      "step": 37685
    },
    {
      "epoch": 5.613643133750372,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.027114456171488794,
      "loss": 0.8107,
      "num_input_tokens_seen": 21871472,
      "step": 37690
    },
    {
      "epoch": 5.614387846291332,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.027113306380305242,
      "loss": 0.7938,
      "num_input_tokens_seen": 21874320,
      "step": 37695
    },
    {
      "epoch": 5.615132558832291,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.027112156384479696,
      "loss": 0.8,
      "num_input_tokens_seen": 21877168,
      "step": 37700
    },
    {
      "epoch": 5.61587727137325,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.027111006184031578,
      "loss": 0.8005,
      "num_input_tokens_seen": 21880016,
      "step": 37705
    },
    {
      "epoch": 5.616621983914209,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02710985577898032,
      "loss": 0.8005,
      "num_input_tokens_seen": 21882704,
      "step": 37710
    },
    {
      "epoch": 5.617366696455168,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.027108705169345354,
      "loss": 0.776,
      "num_input_tokens_seen": 21885744,
      "step": 37715
    },
    {
      "epoch": 5.6181114089961275,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.027107554355146127,
      "loss": 0.7884,
      "num_input_tokens_seen": 21888560,
      "step": 37720
    },
    {
      "epoch": 5.618856121537087,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.027106403336402067,
      "loss": 0.7984,
      "num_input_tokens_seen": 21891600,
      "step": 37725
    },
    {
      "epoch": 5.619600834078046,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02710525211313263,
      "loss": 0.7841,
      "num_input_tokens_seen": 21894416,
      "step": 37730
    },
    {
      "epoch": 5.620345546619005,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.027104100685357263,
      "loss": 0.7817,
      "num_input_tokens_seen": 21897712,
      "step": 37735
    },
    {
      "epoch": 5.621090259159964,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.027102949053095418,
      "loss": 0.807,
      "num_input_tokens_seen": 21900752,
      "step": 37740
    },
    {
      "epoch": 5.621834971700923,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.027101797216366547,
      "loss": 0.8034,
      "num_input_tokens_seen": 21903728,
      "step": 37745
    },
    {
      "epoch": 5.622579684241883,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.027100645175190112,
      "loss": 0.7893,
      "num_input_tokens_seen": 21906704,
      "step": 37750
    },
    {
      "epoch": 5.6233243967828415,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.027099492929585575,
      "loss": 0.7774,
      "num_input_tokens_seen": 21909552,
      "step": 37755
    },
    {
      "epoch": 5.624069109323801,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.027098340479572405,
      "loss": 0.8234,
      "num_input_tokens_seen": 21912432,
      "step": 37760
    },
    {
      "epoch": 5.62481382186476,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02709718782517007,
      "loss": 0.8051,
      "num_input_tokens_seen": 21915408,
      "step": 37765
    },
    {
      "epoch": 5.6255585344057195,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.027096034966398036,
      "loss": 0.8221,
      "num_input_tokens_seen": 21918256,
      "step": 37770
    },
    {
      "epoch": 5.626303246946678,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.027094881903275785,
      "loss": 0.8138,
      "num_input_tokens_seen": 21921392,
      "step": 37775
    },
    {
      "epoch": 5.627047959487638,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.027093728635822805,
      "loss": 0.8188,
      "num_input_tokens_seen": 21924240,
      "step": 37780
    },
    {
      "epoch": 5.627792672028597,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.027092575164058563,
      "loss": 0.8369,
      "num_input_tokens_seen": 21927184,
      "step": 37785
    },
    {
      "epoch": 5.628537384569556,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.02709142148800256,
      "loss": 0.8015,
      "num_input_tokens_seen": 21929936,
      "step": 37790
    },
    {
      "epoch": 5.629282097110515,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.027090267607674273,
      "loss": 0.7924,
      "num_input_tokens_seen": 21932912,
      "step": 37795
    },
    {
      "epoch": 5.630026809651475,
      "grad_norm": 0.0123291015625,
      "learning_rate": 0.027089113523093206,
      "loss": 0.7842,
      "num_input_tokens_seen": 21935888,
      "step": 37800
    },
    {
      "epoch": 5.6307715221924335,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.027087959234278855,
      "loss": 0.787,
      "num_input_tokens_seen": 21938672,
      "step": 37805
    },
    {
      "epoch": 5.631516234733393,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.027086804741250716,
      "loss": 0.8063,
      "num_input_tokens_seen": 21941552,
      "step": 37810
    },
    {
      "epoch": 5.632260947274352,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.027085650044028298,
      "loss": 0.7657,
      "num_input_tokens_seen": 21944336,
      "step": 37815
    },
    {
      "epoch": 5.6330056598153115,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.027084495142631105,
      "loss": 0.7941,
      "num_input_tokens_seen": 21947056,
      "step": 37820
    },
    {
      "epoch": 5.63375037235627,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.027083340037078647,
      "loss": 0.8151,
      "num_input_tokens_seen": 21949872,
      "step": 37825
    },
    {
      "epoch": 5.63449508489723,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.027082184727390442,
      "loss": 0.7876,
      "num_input_tokens_seen": 21952368,
      "step": 37830
    },
    {
      "epoch": 5.635239797438189,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.027081029213586006,
      "loss": 0.8079,
      "num_input_tokens_seen": 21955344,
      "step": 37835
    },
    {
      "epoch": 5.635984509979148,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.02707987349568486,
      "loss": 0.8174,
      "num_input_tokens_seen": 21958064,
      "step": 37840
    },
    {
      "epoch": 5.636729222520107,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.027078717573706525,
      "loss": 0.8062,
      "num_input_tokens_seen": 21961072,
      "step": 37845
    },
    {
      "epoch": 5.637473935061067,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.027077561447670536,
      "loss": 0.8035,
      "num_input_tokens_seen": 21964208,
      "step": 37850
    },
    {
      "epoch": 5.6382186476020255,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02707640511759642,
      "loss": 0.8254,
      "num_input_tokens_seen": 21967120,
      "step": 37855
    },
    {
      "epoch": 5.638963360142985,
      "grad_norm": 0.025390625,
      "learning_rate": 0.027075248583503717,
      "loss": 0.8054,
      "num_input_tokens_seen": 21970000,
      "step": 37860
    },
    {
      "epoch": 5.639708072683944,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.027074091845411957,
      "loss": 0.7953,
      "num_input_tokens_seen": 21972944,
      "step": 37865
    },
    {
      "epoch": 5.640452785224904,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02707293490334069,
      "loss": 0.7857,
      "num_input_tokens_seen": 21975792,
      "step": 37870
    },
    {
      "epoch": 5.641197497765862,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.027071777757309456,
      "loss": 0.8052,
      "num_input_tokens_seen": 21978448,
      "step": 37875
    },
    {
      "epoch": 5.641942210306821,
      "grad_norm": 0.021484375,
      "learning_rate": 0.027070620407337806,
      "loss": 0.7787,
      "num_input_tokens_seen": 21981552,
      "step": 37880
    },
    {
      "epoch": 5.642686922847781,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.027069462853445293,
      "loss": 0.8049,
      "num_input_tokens_seen": 21984304,
      "step": 37885
    },
    {
      "epoch": 5.64343163538874,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.027068305095651474,
      "loss": 0.7892,
      "num_input_tokens_seen": 21986928,
      "step": 37890
    },
    {
      "epoch": 5.644176347929699,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.027067147133975906,
      "loss": 0.7978,
      "num_input_tokens_seen": 21989680,
      "step": 37895
    },
    {
      "epoch": 5.644921060470658,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02706598896843815,
      "loss": 0.8105,
      "num_input_tokens_seen": 21992464,
      "step": 37900
    },
    {
      "epoch": 5.6456657730116175,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.027064830599057777,
      "loss": 0.7823,
      "num_input_tokens_seen": 21995216,
      "step": 37905
    },
    {
      "epoch": 5.646410485552577,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.027063672025854347,
      "loss": 0.8063,
      "num_input_tokens_seen": 21998032,
      "step": 37910
    },
    {
      "epoch": 5.647155198093536,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02706251324884744,
      "loss": 0.8248,
      "num_input_tokens_seen": 22001008,
      "step": 37915
    },
    {
      "epoch": 5.647899910634495,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.02706135426805663,
      "loss": 0.8044,
      "num_input_tokens_seen": 22004080,
      "step": 37920
    },
    {
      "epoch": 5.648644623175454,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02706019508350151,
      "loss": 0.8241,
      "num_input_tokens_seen": 22007248,
      "step": 37925
    },
    {
      "epoch": 5.649389335716413,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.027059035695201638,
      "loss": 0.7925,
      "num_input_tokens_seen": 22010128,
      "step": 37930
    },
    {
      "epoch": 5.650134048257373,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02705787610317662,
      "loss": 0.7696,
      "num_input_tokens_seen": 22012848,
      "step": 37935
    },
    {
      "epoch": 5.6508787607983315,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.027056716307446042,
      "loss": 0.7527,
      "num_input_tokens_seen": 22015888,
      "step": 37940
    },
    {
      "epoch": 5.651623473339291,
      "grad_norm": 0.02734375,
      "learning_rate": 0.027055556308029492,
      "loss": 0.8019,
      "num_input_tokens_seen": 22018704,
      "step": 37945
    },
    {
      "epoch": 5.65236818588025,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.027054396104946574,
      "loss": 0.8082,
      "num_input_tokens_seen": 22021264,
      "step": 37950
    },
    {
      "epoch": 5.65311289842121,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.027053235698216885,
      "loss": 0.8192,
      "num_input_tokens_seen": 22023984,
      "step": 37955
    },
    {
      "epoch": 5.653857610962168,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.027052075087860027,
      "loss": 0.7908,
      "num_input_tokens_seen": 22026960,
      "step": 37960
    },
    {
      "epoch": 5.654602323503128,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.02705091427389561,
      "loss": 0.7949,
      "num_input_tokens_seen": 22029616,
      "step": 37965
    },
    {
      "epoch": 5.655347036044087,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.027049753256343245,
      "loss": 0.7996,
      "num_input_tokens_seen": 22032720,
      "step": 37970
    },
    {
      "epoch": 5.656091748585046,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.027048592035222547,
      "loss": 0.8141,
      "num_input_tokens_seen": 22035888,
      "step": 37975
    },
    {
      "epoch": 5.656836461126005,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02704743061055313,
      "loss": 0.8035,
      "num_input_tokens_seen": 22038640,
      "step": 37980
    },
    {
      "epoch": 5.657581173666965,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02704626898235462,
      "loss": 0.8301,
      "num_input_tokens_seen": 22041264,
      "step": 37985
    },
    {
      "epoch": 5.6583258862079235,
      "grad_norm": 0.0390625,
      "learning_rate": 0.027045107150646636,
      "loss": 0.8494,
      "num_input_tokens_seen": 22044048,
      "step": 37990
    },
    {
      "epoch": 5.659070598748883,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.027043945115448807,
      "loss": 0.8247,
      "num_input_tokens_seen": 22046512,
      "step": 37995
    },
    {
      "epoch": 5.659815311289842,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02704278287678077,
      "loss": 0.8085,
      "num_input_tokens_seen": 22049424,
      "step": 38000
    },
    {
      "epoch": 5.660560023830802,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02704162043466216,
      "loss": 0.7797,
      "num_input_tokens_seen": 22052464,
      "step": 38005
    },
    {
      "epoch": 5.66130473637176,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.027040457789112602,
      "loss": 0.8001,
      "num_input_tokens_seen": 22055216,
      "step": 38010
    },
    {
      "epoch": 5.66204944891272,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.027039294940151753,
      "loss": 0.7852,
      "num_input_tokens_seen": 22058224,
      "step": 38015
    },
    {
      "epoch": 5.662794161453679,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02703813188779925,
      "loss": 0.8132,
      "num_input_tokens_seen": 22061392,
      "step": 38020
    },
    {
      "epoch": 5.663538873994638,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.027036968632074745,
      "loss": 0.7883,
      "num_input_tokens_seen": 22064336,
      "step": 38025
    },
    {
      "epoch": 5.664283586535597,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.027035805172997886,
      "loss": 0.8283,
      "num_input_tokens_seen": 22067536,
      "step": 38030
    },
    {
      "epoch": 5.665028299076557,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.027034641510588337,
      "loss": 0.8127,
      "num_input_tokens_seen": 22070352,
      "step": 38035
    },
    {
      "epoch": 5.665773011617516,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.027033477644865744,
      "loss": 0.8173,
      "num_input_tokens_seen": 22073360,
      "step": 38040
    },
    {
      "epoch": 5.666517724158475,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02703231357584978,
      "loss": 0.805,
      "num_input_tokens_seen": 22076048,
      "step": 38045
    },
    {
      "epoch": 5.667262436699434,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02703114930356011,
      "loss": 0.811,
      "num_input_tokens_seen": 22078992,
      "step": 38050
    },
    {
      "epoch": 5.668007149240394,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.0270299848280164,
      "loss": 0.7976,
      "num_input_tokens_seen": 22081488,
      "step": 38055
    },
    {
      "epoch": 5.668751861781352,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.027028820149238315,
      "loss": 0.8047,
      "num_input_tokens_seen": 22084368,
      "step": 38060
    },
    {
      "epoch": 5.669496574322311,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.027027655267245543,
      "loss": 0.7762,
      "num_input_tokens_seen": 22086992,
      "step": 38065
    },
    {
      "epoch": 5.670241286863271,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.027026490182057765,
      "loss": 0.8124,
      "num_input_tokens_seen": 22089840,
      "step": 38070
    },
    {
      "epoch": 5.67098599940423,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.027025324893694653,
      "loss": 0.8084,
      "num_input_tokens_seen": 22092752,
      "step": 38075
    },
    {
      "epoch": 5.671730711945189,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.027024159402175903,
      "loss": 0.8022,
      "num_input_tokens_seen": 22095408,
      "step": 38080
    },
    {
      "epoch": 5.672475424486148,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.0270229937075212,
      "loss": 0.788,
      "num_input_tokens_seen": 22098032,
      "step": 38085
    },
    {
      "epoch": 5.673220137027108,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.027021827809750234,
      "loss": 0.8192,
      "num_input_tokens_seen": 22100880,
      "step": 38090
    },
    {
      "epoch": 5.673964849568066,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.027020661708882707,
      "loss": 0.8005,
      "num_input_tokens_seen": 22103952,
      "step": 38095
    },
    {
      "epoch": 5.674709562109026,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.02701949540493832,
      "loss": 0.791,
      "num_input_tokens_seen": 22106864,
      "step": 38100
    },
    {
      "epoch": 5.675454274649985,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.027018328897936775,
      "loss": 0.8006,
      "num_input_tokens_seen": 22109840,
      "step": 38105
    },
    {
      "epoch": 5.676198987190944,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.027017162187897778,
      "loss": 0.8049,
      "num_input_tokens_seen": 22113008,
      "step": 38110
    },
    {
      "epoch": 5.676943699731903,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.027015995274841036,
      "loss": 0.7962,
      "num_input_tokens_seen": 22115824,
      "step": 38115
    },
    {
      "epoch": 5.677688412272863,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.02701482815878627,
      "loss": 0.7975,
      "num_input_tokens_seen": 22118928,
      "step": 38120
    },
    {
      "epoch": 5.678433124813822,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02701366083975319,
      "loss": 0.8126,
      "num_input_tokens_seen": 22122160,
      "step": 38125
    },
    {
      "epoch": 5.679177837354781,
      "grad_norm": 0.0103759765625,
      "learning_rate": 0.027012493317761524,
      "loss": 0.8057,
      "num_input_tokens_seen": 22125072,
      "step": 38130
    },
    {
      "epoch": 5.67992254989574,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02701132559283099,
      "loss": 0.8122,
      "num_input_tokens_seen": 22127984,
      "step": 38135
    },
    {
      "epoch": 5.6806672624367,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.027010157664981316,
      "loss": 0.7892,
      "num_input_tokens_seen": 22130896,
      "step": 38140
    },
    {
      "epoch": 5.681411974977658,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.027008989534232235,
      "loss": 0.8056,
      "num_input_tokens_seen": 22133808,
      "step": 38145
    },
    {
      "epoch": 5.682156687518618,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.027007821200603487,
      "loss": 0.8136,
      "num_input_tokens_seen": 22136784,
      "step": 38150
    },
    {
      "epoch": 5.682901400059577,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.0270066526641148,
      "loss": 0.8036,
      "num_input_tokens_seen": 22139824,
      "step": 38155
    },
    {
      "epoch": 5.683646112600536,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.027005483924785918,
      "loss": 0.8115,
      "num_input_tokens_seen": 22142832,
      "step": 38160
    },
    {
      "epoch": 5.684390825141495,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.027004314982636588,
      "loss": 0.8085,
      "num_input_tokens_seen": 22145520,
      "step": 38165
    },
    {
      "epoch": 5.685135537682455,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.027003145837686555,
      "loss": 0.7717,
      "num_input_tokens_seen": 22148560,
      "step": 38170
    },
    {
      "epoch": 5.685880250223414,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.027001976489955577,
      "loss": 0.8133,
      "num_input_tokens_seen": 22151600,
      "step": 38175
    },
    {
      "epoch": 5.686624962764373,
      "grad_norm": 0.0111083984375,
      "learning_rate": 0.0270008069394634,
      "loss": 0.7852,
      "num_input_tokens_seen": 22154864,
      "step": 38180
    },
    {
      "epoch": 5.687369675305332,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.026999637186229792,
      "loss": 0.7891,
      "num_input_tokens_seen": 22157648,
      "step": 38185
    },
    {
      "epoch": 5.688114387846292,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026998467230274506,
      "loss": 0.8026,
      "num_input_tokens_seen": 22160560,
      "step": 38190
    },
    {
      "epoch": 5.68885910038725,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.026997297071617313,
      "loss": 0.7908,
      "num_input_tokens_seen": 22163696,
      "step": 38195
    },
    {
      "epoch": 5.68960381292821,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.026996126710277978,
      "loss": 0.8,
      "num_input_tokens_seen": 22166448,
      "step": 38200
    },
    {
      "epoch": 5.690348525469169,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.026994956146276273,
      "loss": 0.8268,
      "num_input_tokens_seen": 22169232,
      "step": 38205
    },
    {
      "epoch": 5.6910932380101285,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.026993785379631978,
      "loss": 0.8057,
      "num_input_tokens_seen": 22172368,
      "step": 38210
    },
    {
      "epoch": 5.691837950551087,
      "grad_norm": 0.013671875,
      "learning_rate": 0.02699261441036487,
      "loss": 0.8282,
      "num_input_tokens_seen": 22175152,
      "step": 38215
    },
    {
      "epoch": 5.692582663092047,
      "grad_norm": 0.01953125,
      "learning_rate": 0.026991443238494724,
      "loss": 0.8017,
      "num_input_tokens_seen": 22177648,
      "step": 38220
    },
    {
      "epoch": 5.693327375633006,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.026990271864041338,
      "loss": 0.8044,
      "num_input_tokens_seen": 22180528,
      "step": 38225
    },
    {
      "epoch": 5.694072088173964,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.026989100287024493,
      "loss": 0.7848,
      "num_input_tokens_seen": 22183600,
      "step": 38230
    },
    {
      "epoch": 5.694816800714924,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.026987928507463985,
      "loss": 0.7885,
      "num_input_tokens_seen": 22186384,
      "step": 38235
    },
    {
      "epoch": 5.695561513255884,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.026986756525379605,
      "loss": 0.8291,
      "num_input_tokens_seen": 22189488,
      "step": 38240
    },
    {
      "epoch": 5.696306225796842,
      "grad_norm": 0.037109375,
      "learning_rate": 0.02698558434079116,
      "loss": 0.8011,
      "num_input_tokens_seen": 22192208,
      "step": 38245
    },
    {
      "epoch": 5.697050938337801,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.026984411953718448,
      "loss": 0.7845,
      "num_input_tokens_seen": 22195248,
      "step": 38250
    },
    {
      "epoch": 5.697795650878761,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02698323936418128,
      "loss": 0.8051,
      "num_input_tokens_seen": 22198384,
      "step": 38255
    },
    {
      "epoch": 5.6985403634197205,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.026982066572199457,
      "loss": 0.8207,
      "num_input_tokens_seen": 22201104,
      "step": 38260
    },
    {
      "epoch": 5.699285075960679,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.0269808935777928,
      "loss": 0.8125,
      "num_input_tokens_seen": 22203888,
      "step": 38265
    },
    {
      "epoch": 5.700029788501638,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.026979720380981124,
      "loss": 0.8118,
      "num_input_tokens_seen": 22207120,
      "step": 38270
    },
    {
      "epoch": 5.700774501042598,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.026978546981784247,
      "loss": 0.7984,
      "num_input_tokens_seen": 22210256,
      "step": 38275
    },
    {
      "epoch": 5.701519213583556,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.026977373380221985,
      "loss": 0.8031,
      "num_input_tokens_seen": 22213136,
      "step": 38280
    },
    {
      "epoch": 5.702263926124516,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.026976199576314183,
      "loss": 0.8178,
      "num_input_tokens_seen": 22216144,
      "step": 38285
    },
    {
      "epoch": 5.703008638665475,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02697502557008066,
      "loss": 0.7949,
      "num_input_tokens_seen": 22218704,
      "step": 38290
    },
    {
      "epoch": 5.7037533512064345,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.026973851361541252,
      "loss": 0.8109,
      "num_input_tokens_seen": 22221680,
      "step": 38295
    },
    {
      "epoch": 5.704498063747393,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02697267695071579,
      "loss": 0.8076,
      "num_input_tokens_seen": 22224368,
      "step": 38300
    },
    {
      "epoch": 5.705242776288353,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02697150233762412,
      "loss": 0.7889,
      "num_input_tokens_seen": 22227280,
      "step": 38305
    },
    {
      "epoch": 5.705987488829312,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.026970327522286088,
      "loss": 0.7899,
      "num_input_tokens_seen": 22230064,
      "step": 38310
    },
    {
      "epoch": 5.706732201370271,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02696915250472154,
      "loss": 0.7918,
      "num_input_tokens_seen": 22233008,
      "step": 38315
    },
    {
      "epoch": 5.70747691391123,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.026967977284950317,
      "loss": 0.8006,
      "num_input_tokens_seen": 22236144,
      "step": 38320
    },
    {
      "epoch": 5.70822162645219,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02696680186299229,
      "loss": 0.7975,
      "num_input_tokens_seen": 22239216,
      "step": 38325
    },
    {
      "epoch": 5.708966338993148,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026965626238867304,
      "loss": 0.7954,
      "num_input_tokens_seen": 22241808,
      "step": 38330
    },
    {
      "epoch": 5.709711051534108,
      "grad_norm": 0.01092529296875,
      "learning_rate": 0.026964450412595226,
      "loss": 0.7788,
      "num_input_tokens_seen": 22244752,
      "step": 38335
    },
    {
      "epoch": 5.710455764075067,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.026963274384195917,
      "loss": 0.8169,
      "num_input_tokens_seen": 22247920,
      "step": 38340
    },
    {
      "epoch": 5.7112004766160265,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02696209815368925,
      "loss": 0.7948,
      "num_input_tokens_seen": 22250768,
      "step": 38345
    },
    {
      "epoch": 5.711945189156985,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.02696092172109509,
      "loss": 0.8096,
      "num_input_tokens_seen": 22253616,
      "step": 38350
    },
    {
      "epoch": 5.712689901697945,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.026959745086433312,
      "loss": 0.7985,
      "num_input_tokens_seen": 22256208,
      "step": 38355
    },
    {
      "epoch": 5.713434614238904,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.026958568249723796,
      "loss": 0.807,
      "num_input_tokens_seen": 22259024,
      "step": 38360
    },
    {
      "epoch": 5.714179326779863,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.026957391210986427,
      "loss": 0.7908,
      "num_input_tokens_seen": 22262192,
      "step": 38365
    },
    {
      "epoch": 5.714924039320822,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026956213970241085,
      "loss": 0.782,
      "num_input_tokens_seen": 22264912,
      "step": 38370
    },
    {
      "epoch": 5.715668751861782,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02695503652750766,
      "loss": 0.8431,
      "num_input_tokens_seen": 22267824,
      "step": 38375
    },
    {
      "epoch": 5.7164134644027405,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026953858882806037,
      "loss": 0.8131,
      "num_input_tokens_seen": 22270960,
      "step": 38380
    },
    {
      "epoch": 5.7171581769437,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026952681036156126,
      "loss": 0.8302,
      "num_input_tokens_seen": 22274032,
      "step": 38385
    },
    {
      "epoch": 5.717902889484659,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.026951502987577817,
      "loss": 0.8079,
      "num_input_tokens_seen": 22276848,
      "step": 38390
    },
    {
      "epoch": 5.718647602025618,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.026950324737091007,
      "loss": 0.7985,
      "num_input_tokens_seen": 22279760,
      "step": 38395
    },
    {
      "epoch": 5.719392314566577,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02694914628471561,
      "loss": 0.7999,
      "num_input_tokens_seen": 22282608,
      "step": 38400
    },
    {
      "epoch": 5.720137027107537,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.02694796763047153,
      "loss": 0.8086,
      "num_input_tokens_seen": 22285584,
      "step": 38405
    },
    {
      "epoch": 5.720881739648496,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02694678877437868,
      "loss": 0.804,
      "num_input_tokens_seen": 22288144,
      "step": 38410
    },
    {
      "epoch": 5.721626452189454,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.026945609716456972,
      "loss": 0.7899,
      "num_input_tokens_seen": 22291056,
      "step": 38415
    },
    {
      "epoch": 5.722371164730414,
      "grad_norm": 0.01953125,
      "learning_rate": 0.026944430456726332,
      "loss": 0.8088,
      "num_input_tokens_seen": 22293808,
      "step": 38420
    },
    {
      "epoch": 5.723115877271374,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.026943250995206683,
      "loss": 0.8044,
      "num_input_tokens_seen": 22296688,
      "step": 38425
    },
    {
      "epoch": 5.7238605898123325,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.026942071331917943,
      "loss": 0.8192,
      "num_input_tokens_seen": 22299184,
      "step": 38430
    },
    {
      "epoch": 5.724605302353291,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.026940891466880047,
      "loss": 0.8064,
      "num_input_tokens_seen": 22302160,
      "step": 38435
    },
    {
      "epoch": 5.725350014894251,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.026939711400112924,
      "loss": 0.8054,
      "num_input_tokens_seen": 22305520,
      "step": 38440
    },
    {
      "epoch": 5.72609472743521,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.026938531131636516,
      "loss": 0.8062,
      "num_input_tokens_seen": 22308656,
      "step": 38445
    },
    {
      "epoch": 5.726839439976169,
      "grad_norm": 0.01953125,
      "learning_rate": 0.026937350661470752,
      "loss": 0.7839,
      "num_input_tokens_seen": 22311472,
      "step": 38450
    },
    {
      "epoch": 5.727584152517128,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.026936169989635585,
      "loss": 0.8006,
      "num_input_tokens_seen": 22314448,
      "step": 38455
    },
    {
      "epoch": 5.728328865058088,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.026934989116150956,
      "loss": 0.8123,
      "num_input_tokens_seen": 22317392,
      "step": 38460
    },
    {
      "epoch": 5.7290735775990465,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02693380804103682,
      "loss": 0.8033,
      "num_input_tokens_seen": 22320400,
      "step": 38465
    },
    {
      "epoch": 5.729818290140006,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02693262676431312,
      "loss": 0.7899,
      "num_input_tokens_seen": 22323376,
      "step": 38470
    },
    {
      "epoch": 5.730563002680965,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.026931445285999823,
      "loss": 0.8021,
      "num_input_tokens_seen": 22326320,
      "step": 38475
    },
    {
      "epoch": 5.7313077152219245,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.02693026360611688,
      "loss": 0.7972,
      "num_input_tokens_seen": 22328976,
      "step": 38480
    },
    {
      "epoch": 5.732052427762883,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.026929081724684265,
      "loss": 0.7955,
      "num_input_tokens_seen": 22332080,
      "step": 38485
    },
    {
      "epoch": 5.732797140303843,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.026927899641721934,
      "loss": 0.8289,
      "num_input_tokens_seen": 22335024,
      "step": 38490
    },
    {
      "epoch": 5.733541852844802,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02692671735724986,
      "loss": 0.8076,
      "num_input_tokens_seen": 22337712,
      "step": 38495
    },
    {
      "epoch": 5.734286565385761,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.026925534871288025,
      "loss": 0.8109,
      "num_input_tokens_seen": 22340752,
      "step": 38500
    },
    {
      "epoch": 5.73503127792672,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.026924352183856394,
      "loss": 0.7949,
      "num_input_tokens_seen": 22343312,
      "step": 38505
    },
    {
      "epoch": 5.73577599046768,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02692316929497495,
      "loss": 0.7793,
      "num_input_tokens_seen": 22346544,
      "step": 38510
    },
    {
      "epoch": 5.7365207030086385,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02692198620466368,
      "loss": 0.8171,
      "num_input_tokens_seen": 22349392,
      "step": 38515
    },
    {
      "epoch": 5.737265415549598,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.026920802912942573,
      "loss": 0.7861,
      "num_input_tokens_seen": 22352432,
      "step": 38520
    },
    {
      "epoch": 5.738010128090557,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02691961941983162,
      "loss": 0.8013,
      "num_input_tokens_seen": 22355536,
      "step": 38525
    },
    {
      "epoch": 5.7387548406315165,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.026918435725350805,
      "loss": 0.8092,
      "num_input_tokens_seen": 22358384,
      "step": 38530
    },
    {
      "epoch": 5.739499553172475,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.026917251829520133,
      "loss": 0.7799,
      "num_input_tokens_seen": 22361360,
      "step": 38535
    },
    {
      "epoch": 5.740244265713435,
      "grad_norm": 0.021484375,
      "learning_rate": 0.026916067732359602,
      "loss": 0.8152,
      "num_input_tokens_seen": 22364048,
      "step": 38540
    },
    {
      "epoch": 5.740988978254394,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02691488343388922,
      "loss": 0.7996,
      "num_input_tokens_seen": 22366800,
      "step": 38545
    },
    {
      "epoch": 5.741733690795353,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.026913698934128993,
      "loss": 0.7962,
      "num_input_tokens_seen": 22369552,
      "step": 38550
    },
    {
      "epoch": 5.742478403336312,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.026912514233098928,
      "loss": 0.8041,
      "num_input_tokens_seen": 22372496,
      "step": 38555
    },
    {
      "epoch": 5.743223115877272,
      "grad_norm": 0.013671875,
      "learning_rate": 0.026911329330819043,
      "loss": 0.8063,
      "num_input_tokens_seen": 22375376,
      "step": 38560
    },
    {
      "epoch": 5.7439678284182305,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.026910144227309363,
      "loss": 0.8047,
      "num_input_tokens_seen": 22378384,
      "step": 38565
    },
    {
      "epoch": 5.74471254095919,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.026908958922589894,
      "loss": 0.8211,
      "num_input_tokens_seen": 22380912,
      "step": 38570
    },
    {
      "epoch": 5.745457253500149,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.026907773416680665,
      "loss": 0.8079,
      "num_input_tokens_seen": 22383888,
      "step": 38575
    },
    {
      "epoch": 5.746201966041108,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02690658770960171,
      "loss": 0.8167,
      "num_input_tokens_seen": 22386640,
      "step": 38580
    },
    {
      "epoch": 5.746946678582067,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.026905401801373057,
      "loss": 0.8028,
      "num_input_tokens_seen": 22389552,
      "step": 38585
    },
    {
      "epoch": 5.747691391123027,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.026904215692014743,
      "loss": 0.7834,
      "num_input_tokens_seen": 22392464,
      "step": 38590
    },
    {
      "epoch": 5.748436103663986,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.026903029381546796,
      "loss": 0.8296,
      "num_input_tokens_seen": 22395248,
      "step": 38595
    },
    {
      "epoch": 5.7491808162049445,
      "grad_norm": 0.015625,
      "learning_rate": 0.02690184286998927,
      "loss": 0.8117,
      "num_input_tokens_seen": 22398384,
      "step": 38600
    },
    {
      "epoch": 5.749925528745904,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.02690065615736221,
      "loss": 0.8046,
      "num_input_tokens_seen": 22401072,
      "step": 38605
    },
    {
      "epoch": 5.750670241286863,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.026899469243685654,
      "loss": 0.8132,
      "num_input_tokens_seen": 22403888,
      "step": 38610
    },
    {
      "epoch": 5.7514149538278225,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02689828212897966,
      "loss": 0.8005,
      "num_input_tokens_seen": 22406832,
      "step": 38615
    },
    {
      "epoch": 5.752159666368781,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.026897094813264286,
      "loss": 0.8088,
      "num_input_tokens_seen": 22409616,
      "step": 38620
    },
    {
      "epoch": 5.752904378909741,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02689590729655958,
      "loss": 0.7857,
      "num_input_tokens_seen": 22412656,
      "step": 38625
    },
    {
      "epoch": 5.7536490914507,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.026894719578885614,
      "loss": 0.8004,
      "num_input_tokens_seen": 22415280,
      "step": 38630
    },
    {
      "epoch": 5.754393803991659,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.026893531660262454,
      "loss": 0.7874,
      "num_input_tokens_seen": 22417936,
      "step": 38635
    },
    {
      "epoch": 5.755138516532618,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02689234354071016,
      "loss": 0.7963,
      "num_input_tokens_seen": 22420880,
      "step": 38640
    },
    {
      "epoch": 5.755883229073578,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02689115522024881,
      "loss": 0.8055,
      "num_input_tokens_seen": 22423632,
      "step": 38645
    },
    {
      "epoch": 5.7566279416145365,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.026889966698898475,
      "loss": 0.8057,
      "num_input_tokens_seen": 22426992,
      "step": 38650
    },
    {
      "epoch": 5.757372654155496,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02688877797667924,
      "loss": 0.8163,
      "num_input_tokens_seen": 22430288,
      "step": 38655
    },
    {
      "epoch": 5.758117366696455,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.026887589053611185,
      "loss": 0.8094,
      "num_input_tokens_seen": 22433392,
      "step": 38660
    },
    {
      "epoch": 5.7588620792374146,
      "grad_norm": 0.029296875,
      "learning_rate": 0.026886399929714394,
      "loss": 0.7952,
      "num_input_tokens_seen": 22436368,
      "step": 38665
    },
    {
      "epoch": 5.759606791778373,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.02688521060500896,
      "loss": 0.8201,
      "num_input_tokens_seen": 22439376,
      "step": 38670
    },
    {
      "epoch": 5.760351504319333,
      "grad_norm": 0.01171875,
      "learning_rate": 0.02688402107951497,
      "loss": 0.8171,
      "num_input_tokens_seen": 22442544,
      "step": 38675
    },
    {
      "epoch": 5.761096216860292,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02688283135325252,
      "loss": 0.7996,
      "num_input_tokens_seen": 22445744,
      "step": 38680
    },
    {
      "epoch": 5.761840929401251,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.02688164142624171,
      "loss": 0.8131,
      "num_input_tokens_seen": 22448592,
      "step": 38685
    },
    {
      "epoch": 5.76258564194221,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.026880451298502653,
      "loss": 0.8019,
      "num_input_tokens_seen": 22451536,
      "step": 38690
    },
    {
      "epoch": 5.76333035448317,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.026879260970055434,
      "loss": 0.817,
      "num_input_tokens_seen": 22454736,
      "step": 38695
    },
    {
      "epoch": 5.7640750670241285,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.026878070440920183,
      "loss": 0.7934,
      "num_input_tokens_seen": 22457616,
      "step": 38700
    },
    {
      "epoch": 5.764819779565088,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.026876879711117,
      "loss": 0.7934,
      "num_input_tokens_seen": 22460464,
      "step": 38705
    },
    {
      "epoch": 5.765564492106047,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02687568878066601,
      "loss": 0.7999,
      "num_input_tokens_seen": 22463376,
      "step": 38710
    },
    {
      "epoch": 5.766309204647007,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.026874497649587326,
      "loss": 0.8065,
      "num_input_tokens_seen": 22466192,
      "step": 38715
    },
    {
      "epoch": 5.767053917187965,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02687330631790107,
      "loss": 0.8036,
      "num_input_tokens_seen": 22469520,
      "step": 38720
    },
    {
      "epoch": 5.767798629728925,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.026872114785627375,
      "loss": 0.8099,
      "num_input_tokens_seen": 22472272,
      "step": 38725
    },
    {
      "epoch": 5.768543342269884,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.026870923052786367,
      "loss": 0.7958,
      "num_input_tokens_seen": 22475216,
      "step": 38730
    },
    {
      "epoch": 5.769288054810843,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.026869731119398176,
      "loss": 0.7872,
      "num_input_tokens_seen": 22478000,
      "step": 38735
    },
    {
      "epoch": 5.770032767351802,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.026868538985482943,
      "loss": 0.7957,
      "num_input_tokens_seen": 22480912,
      "step": 38740
    },
    {
      "epoch": 5.770777479892761,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.026867346651060806,
      "loss": 0.795,
      "num_input_tokens_seen": 22483888,
      "step": 38745
    },
    {
      "epoch": 5.7715221924337206,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02686615411615191,
      "loss": 0.7951,
      "num_input_tokens_seen": 22486640,
      "step": 38750
    },
    {
      "epoch": 5.77226690497468,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.0268649613807764,
      "loss": 0.7854,
      "num_input_tokens_seen": 22489168,
      "step": 38755
    },
    {
      "epoch": 5.773011617515639,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.026863768444954427,
      "loss": 0.7997,
      "num_input_tokens_seen": 22491888,
      "step": 38760
    },
    {
      "epoch": 5.773756330056598,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.026862575308706147,
      "loss": 0.8128,
      "num_input_tokens_seen": 22494704,
      "step": 38765
    },
    {
      "epoch": 5.774501042597557,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.026861381972051707,
      "loss": 0.801,
      "num_input_tokens_seen": 22497744,
      "step": 38770
    },
    {
      "epoch": 5.775245755138517,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.026860188435011277,
      "loss": 0.8217,
      "num_input_tokens_seen": 22500816,
      "step": 38775
    },
    {
      "epoch": 5.775990467679476,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.026858994697605016,
      "loss": 0.8188,
      "num_input_tokens_seen": 22503568,
      "step": 38780
    },
    {
      "epoch": 5.7767351802204345,
      "grad_norm": 0.01953125,
      "learning_rate": 0.026857800759853095,
      "loss": 0.8118,
      "num_input_tokens_seen": 22506608,
      "step": 38785
    },
    {
      "epoch": 5.777479892761394,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02685660662177568,
      "loss": 0.7745,
      "num_input_tokens_seen": 22509360,
      "step": 38790
    },
    {
      "epoch": 5.778224605302353,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02685541228339295,
      "loss": 0.8117,
      "num_input_tokens_seen": 22512368,
      "step": 38795
    },
    {
      "epoch": 5.778969317843313,
      "grad_norm": 0.01953125,
      "learning_rate": 0.026854217744725076,
      "loss": 0.789,
      "num_input_tokens_seen": 22515376,
      "step": 38800
    },
    {
      "epoch": 5.779714030384271,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.026853023005792242,
      "loss": 0.806,
      "num_input_tokens_seen": 22518608,
      "step": 38805
    },
    {
      "epoch": 5.780458742925231,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02685182806661463,
      "loss": 0.806,
      "num_input_tokens_seen": 22521552,
      "step": 38810
    },
    {
      "epoch": 5.78120345546619,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02685063292721243,
      "loss": 0.7894,
      "num_input_tokens_seen": 22524464,
      "step": 38815
    },
    {
      "epoch": 5.781948168007149,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02684943758760583,
      "loss": 0.8003,
      "num_input_tokens_seen": 22527248,
      "step": 38820
    },
    {
      "epoch": 5.782692880548108,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.026848242047815024,
      "loss": 0.7978,
      "num_input_tokens_seen": 22530192,
      "step": 38825
    },
    {
      "epoch": 5.783437593089068,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.026847046307860212,
      "loss": 0.7888,
      "num_input_tokens_seen": 22533072,
      "step": 38830
    },
    {
      "epoch": 5.7841823056300266,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.026845850367761593,
      "loss": 0.7891,
      "num_input_tokens_seen": 22535984,
      "step": 38835
    },
    {
      "epoch": 5.784927018170986,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.026844654227539375,
      "loss": 0.7891,
      "num_input_tokens_seen": 22538768,
      "step": 38840
    },
    {
      "epoch": 5.785671730711945,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02684345788721376,
      "loss": 0.7886,
      "num_input_tokens_seen": 22541968,
      "step": 38845
    },
    {
      "epoch": 5.786416443252905,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.02684226134680496,
      "loss": 0.7931,
      "num_input_tokens_seen": 22544944,
      "step": 38850
    },
    {
      "epoch": 5.787161155793863,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02684106460633319,
      "loss": 0.8022,
      "num_input_tokens_seen": 22547984,
      "step": 38855
    },
    {
      "epoch": 5.787905868334823,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.026839867665818667,
      "loss": 0.8001,
      "num_input_tokens_seen": 22550992,
      "step": 38860
    },
    {
      "epoch": 5.788650580875782,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02683867052528162,
      "loss": 0.821,
      "num_input_tokens_seen": 22553936,
      "step": 38865
    },
    {
      "epoch": 5.789395293416741,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.026837473184742264,
      "loss": 0.7958,
      "num_input_tokens_seen": 22556624,
      "step": 38870
    },
    {
      "epoch": 5.7901400059577,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.026836275644220822,
      "loss": 0.8018,
      "num_input_tokens_seen": 22559600,
      "step": 38875
    },
    {
      "epoch": 5.79088471849866,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.026835077903737542,
      "loss": 0.7887,
      "num_input_tokens_seen": 22562512,
      "step": 38880
    },
    {
      "epoch": 5.791629431039619,
      "grad_norm": 0.021484375,
      "learning_rate": 0.026833879963312646,
      "loss": 0.8143,
      "num_input_tokens_seen": 22565680,
      "step": 38885
    },
    {
      "epoch": 5.792374143580578,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.026832681822966376,
      "loss": 0.7745,
      "num_input_tokens_seen": 22568528,
      "step": 38890
    },
    {
      "epoch": 5.793118856121537,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.026831483482718972,
      "loss": 0.842,
      "num_input_tokens_seen": 22571472,
      "step": 38895
    },
    {
      "epoch": 5.793863568662497,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.026830284942590682,
      "loss": 0.8089,
      "num_input_tokens_seen": 22574320,
      "step": 38900
    },
    {
      "epoch": 5.794608281203455,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02682908620260175,
      "loss": 0.7978,
      "num_input_tokens_seen": 22577232,
      "step": 38905
    },
    {
      "epoch": 5.795352993744415,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.026827887262772427,
      "loss": 0.8186,
      "num_input_tokens_seen": 22580560,
      "step": 38910
    },
    {
      "epoch": 5.796097706285374,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.02682668812312297,
      "loss": 0.7926,
      "num_input_tokens_seen": 22583440,
      "step": 38915
    },
    {
      "epoch": 5.796842418826333,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02682548878367364,
      "loss": 0.8001,
      "num_input_tokens_seen": 22586064,
      "step": 38920
    },
    {
      "epoch": 5.797587131367292,
      "grad_norm": 0.01171875,
      "learning_rate": 0.0268242892444447,
      "loss": 0.7946,
      "num_input_tokens_seen": 22588816,
      "step": 38925
    },
    {
      "epoch": 5.798331843908251,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.026823089505456404,
      "loss": 0.7801,
      "num_input_tokens_seen": 22591472,
      "step": 38930
    },
    {
      "epoch": 5.799076556449211,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.02682188956672903,
      "loss": 0.8126,
      "num_input_tokens_seen": 22594352,
      "step": 38935
    },
    {
      "epoch": 5.79982126899017,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.026820689428282845,
      "loss": 0.8087,
      "num_input_tokens_seen": 22597136,
      "step": 38940
    },
    {
      "epoch": 5.800565981531129,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.026819489090138125,
      "loss": 0.8115,
      "num_input_tokens_seen": 22600144,
      "step": 38945
    },
    {
      "epoch": 5.801310694072088,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02681828855231515,
      "loss": 0.781,
      "num_input_tokens_seen": 22603056,
      "step": 38950
    },
    {
      "epoch": 5.802055406613047,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.026817087814834207,
      "loss": 0.7926,
      "num_input_tokens_seen": 22605936,
      "step": 38955
    },
    {
      "epoch": 5.802800119154006,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.02681588687771557,
      "loss": 0.8073,
      "num_input_tokens_seen": 22608848,
      "step": 38960
    },
    {
      "epoch": 5.803544831694966,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.026814685740979536,
      "loss": 0.8072,
      "num_input_tokens_seen": 22611888,
      "step": 38965
    },
    {
      "epoch": 5.804289544235925,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.026813484404646395,
      "loss": 0.7966,
      "num_input_tokens_seen": 22615056,
      "step": 38970
    },
    {
      "epoch": 5.805034256776884,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02681228286873644,
      "loss": 0.8059,
      "num_input_tokens_seen": 22618064,
      "step": 38975
    },
    {
      "epoch": 5.805778969317843,
      "grad_norm": 0.041015625,
      "learning_rate": 0.026811081133269972,
      "loss": 0.7952,
      "num_input_tokens_seen": 22621072,
      "step": 38980
    },
    {
      "epoch": 5.806523681858803,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.026809879198267293,
      "loss": 0.8,
      "num_input_tokens_seen": 22624080,
      "step": 38985
    },
    {
      "epoch": 5.807268394399761,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.026808677063748707,
      "loss": 0.8594,
      "num_input_tokens_seen": 22628304,
      "step": 38990
    },
    {
      "epoch": 5.808013106940721,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.026807474729734524,
      "loss": 0.8072,
      "num_input_tokens_seen": 22631024,
      "step": 38995
    },
    {
      "epoch": 5.80875781948168,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.026806272196245057,
      "loss": 0.8004,
      "num_input_tokens_seen": 22634288,
      "step": 39000
    },
    {
      "epoch": 5.809502532022639,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02680506946330062,
      "loss": 0.8033,
      "num_input_tokens_seen": 22637264,
      "step": 39005
    },
    {
      "epoch": 5.810247244563598,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.02680386653092153,
      "loss": 0.8062,
      "num_input_tokens_seen": 22640048,
      "step": 39010
    },
    {
      "epoch": 5.810991957104558,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.026802663399128117,
      "loss": 0.8258,
      "num_input_tokens_seen": 22642800,
      "step": 39015
    },
    {
      "epoch": 5.811736669645517,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.026801460067940697,
      "loss": 0.7951,
      "num_input_tokens_seen": 22645648,
      "step": 39020
    },
    {
      "epoch": 5.812481382186476,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026800256537379604,
      "loss": 0.81,
      "num_input_tokens_seen": 22648752,
      "step": 39025
    },
    {
      "epoch": 5.813226094727435,
      "grad_norm": 0.0126953125,
      "learning_rate": 0.02679905280746517,
      "loss": 0.8051,
      "num_input_tokens_seen": 22651664,
      "step": 39030
    },
    {
      "epoch": 5.813970807268395,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.026797848878217737,
      "loss": 0.8078,
      "num_input_tokens_seen": 22654672,
      "step": 39035
    },
    {
      "epoch": 5.814715519809353,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026796644749657633,
      "loss": 0.7979,
      "num_input_tokens_seen": 22657648,
      "step": 39040
    },
    {
      "epoch": 5.815460232350313,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.026795440421805206,
      "loss": 0.8,
      "num_input_tokens_seen": 22660592,
      "step": 39045
    },
    {
      "epoch": 5.816204944891272,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.0267942358946808,
      "loss": 0.8125,
      "num_input_tokens_seen": 22663280,
      "step": 39050
    },
    {
      "epoch": 5.8169496574322315,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02679303116830477,
      "loss": 0.7983,
      "num_input_tokens_seen": 22666416,
      "step": 39055
    },
    {
      "epoch": 5.81769436997319,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02679182624269746,
      "loss": 0.8067,
      "num_input_tokens_seen": 22669296,
      "step": 39060
    },
    {
      "epoch": 5.81843908251415,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.026790621117879228,
      "loss": 0.7933,
      "num_input_tokens_seen": 22672208,
      "step": 39065
    },
    {
      "epoch": 5.819183795055109,
      "grad_norm": 0.01953125,
      "learning_rate": 0.026789415793870442,
      "loss": 0.7911,
      "num_input_tokens_seen": 22675120,
      "step": 39070
    },
    {
      "epoch": 5.819928507596068,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.026788210270691455,
      "loss": 0.8206,
      "num_input_tokens_seen": 22678064,
      "step": 39075
    },
    {
      "epoch": 5.820673220137027,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02678700454836264,
      "loss": 0.8041,
      "num_input_tokens_seen": 22680656,
      "step": 39080
    },
    {
      "epoch": 5.821417932677987,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.026785798626904358,
      "loss": 0.791,
      "num_input_tokens_seen": 22683632,
      "step": 39085
    },
    {
      "epoch": 5.822162645218945,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.026784592506336986,
      "loss": 0.7954,
      "num_input_tokens_seen": 22686576,
      "step": 39090
    },
    {
      "epoch": 5.822907357759904,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.026783386186680906,
      "loss": 0.8052,
      "num_input_tokens_seen": 22689168,
      "step": 39095
    },
    {
      "epoch": 5.823652070300864,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02678217966795649,
      "loss": 0.7875,
      "num_input_tokens_seen": 22691856,
      "step": 39100
    },
    {
      "epoch": 5.8243967828418235,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.026780972950184122,
      "loss": 0.8047,
      "num_input_tokens_seen": 22694800,
      "step": 39105
    },
    {
      "epoch": 5.825141495382782,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02677976603338419,
      "loss": 0.8012,
      "num_input_tokens_seen": 22697808,
      "step": 39110
    },
    {
      "epoch": 5.825886207923741,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.026778558917577088,
      "loss": 0.8289,
      "num_input_tokens_seen": 22700848,
      "step": 39115
    },
    {
      "epoch": 5.826630920464701,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.0267773516027832,
      "loss": 0.8022,
      "num_input_tokens_seen": 22703760,
      "step": 39120
    },
    {
      "epoch": 5.82737563300566,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.026776144089022926,
      "loss": 0.7875,
      "num_input_tokens_seen": 22706544,
      "step": 39125
    },
    {
      "epoch": 5.828120345546619,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.026774936376316668,
      "loss": 0.7921,
      "num_input_tokens_seen": 22709712,
      "step": 39130
    },
    {
      "epoch": 5.828865058087578,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02677372846468483,
      "loss": 0.7916,
      "num_input_tokens_seen": 22712464,
      "step": 39135
    },
    {
      "epoch": 5.8296097706285375,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026772520354147813,
      "loss": 0.7955,
      "num_input_tokens_seen": 22715472,
      "step": 39140
    },
    {
      "epoch": 5.830354483169496,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.026771312044726027,
      "loss": 0.7819,
      "num_input_tokens_seen": 22718448,
      "step": 39145
    },
    {
      "epoch": 5.831099195710456,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02677010353643989,
      "loss": 0.8089,
      "num_input_tokens_seen": 22721104,
      "step": 39150
    },
    {
      "epoch": 5.831843908251415,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.02676889482930982,
      "loss": 0.8073,
      "num_input_tokens_seen": 22724304,
      "step": 39155
    },
    {
      "epoch": 5.832588620792374,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.026767685923356228,
      "loss": 0.8145,
      "num_input_tokens_seen": 22727088,
      "step": 39160
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 0.025390625,
      "learning_rate": 0.026766476818599547,
      "loss": 0.7866,
      "num_input_tokens_seen": 22730000,
      "step": 39165
    },
    {
      "epoch": 5.834078045874293,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02676526751506019,
      "loss": 0.7909,
      "num_input_tokens_seen": 22732912,
      "step": 39170
    },
    {
      "epoch": 5.834822758415251,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.026764058012758607,
      "loss": 0.8319,
      "num_input_tokens_seen": 22735760,
      "step": 39175
    },
    {
      "epoch": 5.835567470956211,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.026762848311715215,
      "loss": 0.7853,
      "num_input_tokens_seen": 22738960,
      "step": 39180
    },
    {
      "epoch": 5.83631218349717,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02676163841195046,
      "loss": 0.801,
      "num_input_tokens_seen": 22741616,
      "step": 39185
    },
    {
      "epoch": 5.8370568960381295,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.026760428313484776,
      "loss": 0.8165,
      "num_input_tokens_seen": 22744560,
      "step": 39190
    },
    {
      "epoch": 5.837801608579088,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.026759218016338605,
      "loss": 0.8048,
      "num_input_tokens_seen": 22747536,
      "step": 39195
    },
    {
      "epoch": 5.838546321120048,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.0267580075205324,
      "loss": 0.7904,
      "num_input_tokens_seen": 22750480,
      "step": 39200
    },
    {
      "epoch": 5.839291033661007,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.026756796826086608,
      "loss": 0.8107,
      "num_input_tokens_seen": 22753264,
      "step": 39205
    },
    {
      "epoch": 5.840035746201966,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.026755585933021683,
      "loss": 0.7924,
      "num_input_tokens_seen": 22756144,
      "step": 39210
    },
    {
      "epoch": 5.840780458742925,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.02675437484135808,
      "loss": 0.8096,
      "num_input_tokens_seen": 22758896,
      "step": 39215
    },
    {
      "epoch": 5.841525171283885,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02675316355111626,
      "loss": 0.8034,
      "num_input_tokens_seen": 22761680,
      "step": 39220
    },
    {
      "epoch": 5.8422698838248435,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.026751952062316692,
      "loss": 0.8014,
      "num_input_tokens_seen": 22764496,
      "step": 39225
    },
    {
      "epoch": 5.843014596365803,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.026750740374979834,
      "loss": 0.7825,
      "num_input_tokens_seen": 22767184,
      "step": 39230
    },
    {
      "epoch": 5.843759308906762,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.026749528489126158,
      "loss": 0.8025,
      "num_input_tokens_seen": 22769968,
      "step": 39235
    },
    {
      "epoch": 5.8445040214477215,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.026748316404776148,
      "loss": 0.8145,
      "num_input_tokens_seen": 22772592,
      "step": 39240
    },
    {
      "epoch": 5.84524873398868,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026747104121950267,
      "loss": 0.8055,
      "num_input_tokens_seen": 22775504,
      "step": 39245
    },
    {
      "epoch": 5.84599344652964,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.026745891640669,
      "loss": 0.809,
      "num_input_tokens_seen": 22778544,
      "step": 39250
    },
    {
      "epoch": 5.846738159070599,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.026744678960952836,
      "loss": 0.8195,
      "num_input_tokens_seen": 22781456,
      "step": 39255
    },
    {
      "epoch": 5.847482871611557,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.026743466082822256,
      "loss": 0.7969,
      "num_input_tokens_seen": 22784304,
      "step": 39260
    },
    {
      "epoch": 5.848227584152517,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02674225300629775,
      "loss": 0.7929,
      "num_input_tokens_seen": 22787056,
      "step": 39265
    },
    {
      "epoch": 5.848972296693477,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.026741039731399813,
      "loss": 0.7795,
      "num_input_tokens_seen": 22789744,
      "step": 39270
    },
    {
      "epoch": 5.8497170092344355,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.026739826258148944,
      "loss": 0.7917,
      "num_input_tokens_seen": 22792496,
      "step": 39275
    },
    {
      "epoch": 5.850461721775394,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.026738612586565645,
      "loss": 0.8157,
      "num_input_tokens_seen": 22795312,
      "step": 39280
    },
    {
      "epoch": 5.851206434316354,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02673739871667042,
      "loss": 0.8263,
      "num_input_tokens_seen": 22798544,
      "step": 39285
    },
    {
      "epoch": 5.8519511468573135,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.026736184648483766,
      "loss": 0.791,
      "num_input_tokens_seen": 22801104,
      "step": 39290
    },
    {
      "epoch": 5.852695859398272,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026734970382026203,
      "loss": 0.7933,
      "num_input_tokens_seen": 22804144,
      "step": 39295
    },
    {
      "epoch": 5.853440571939231,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02673375591731824,
      "loss": 0.8163,
      "num_input_tokens_seen": 22807152,
      "step": 39300
    },
    {
      "epoch": 5.854185284480191,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.0267325412543804,
      "loss": 0.8057,
      "num_input_tokens_seen": 22810032,
      "step": 39305
    },
    {
      "epoch": 5.8549299970211495,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.026731326393233196,
      "loss": 0.787,
      "num_input_tokens_seen": 22812752,
      "step": 39310
    },
    {
      "epoch": 5.855674709562109,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.02673011133389716,
      "loss": 0.7821,
      "num_input_tokens_seen": 22815408,
      "step": 39315
    },
    {
      "epoch": 5.856419422103068,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.026728896076392814,
      "loss": 0.8081,
      "num_input_tokens_seen": 22818704,
      "step": 39320
    },
    {
      "epoch": 5.8571641346440275,
      "grad_norm": 0.021484375,
      "learning_rate": 0.026727680620740688,
      "loss": 0.7817,
      "num_input_tokens_seen": 22821680,
      "step": 39325
    },
    {
      "epoch": 5.857908847184986,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02672646496696132,
      "loss": 0.8138,
      "num_input_tokens_seen": 22824560,
      "step": 39330
    },
    {
      "epoch": 5.858653559725946,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02672524911507524,
      "loss": 0.81,
      "num_input_tokens_seen": 22827344,
      "step": 39335
    },
    {
      "epoch": 5.859398272266905,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.026724033065102997,
      "loss": 0.8271,
      "num_input_tokens_seen": 22830032,
      "step": 39340
    },
    {
      "epoch": 5.860142984807864,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02672281681706513,
      "loss": 0.8066,
      "num_input_tokens_seen": 22833264,
      "step": 39345
    },
    {
      "epoch": 5.860887697348823,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.026721600370982188,
      "loss": 0.8297,
      "num_input_tokens_seen": 22836592,
      "step": 39350
    },
    {
      "epoch": 5.861632409889783,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.026720383726874717,
      "loss": 0.8155,
      "num_input_tokens_seen": 22839664,
      "step": 39355
    },
    {
      "epoch": 5.8623771224307415,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02671916688476328,
      "loss": 0.8062,
      "num_input_tokens_seen": 22842480,
      "step": 39360
    },
    {
      "epoch": 5.863121834971701,
      "grad_norm": 0.0234375,
      "learning_rate": 0.026717949844668427,
      "loss": 0.7938,
      "num_input_tokens_seen": 22845520,
      "step": 39365
    },
    {
      "epoch": 5.86386654751266,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.026716732606610723,
      "loss": 0.8083,
      "num_input_tokens_seen": 22848592,
      "step": 39370
    },
    {
      "epoch": 5.8646112600536195,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.026715515170610726,
      "loss": 0.8085,
      "num_input_tokens_seen": 22851088,
      "step": 39375
    },
    {
      "epoch": 5.865355972594578,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.026714297536689014,
      "loss": 0.8047,
      "num_input_tokens_seen": 22853840,
      "step": 39380
    },
    {
      "epoch": 5.866100685135538,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.026713079704866145,
      "loss": 0.7945,
      "num_input_tokens_seen": 22856816,
      "step": 39385
    },
    {
      "epoch": 5.866845397676497,
      "grad_norm": 0.01422119140625,
      "learning_rate": 0.0267118616751627,
      "loss": 0.7945,
      "num_input_tokens_seen": 22859632,
      "step": 39390
    },
    {
      "epoch": 5.867590110217456,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02671064344759926,
      "loss": 0.8012,
      "num_input_tokens_seen": 22862320,
      "step": 39395
    },
    {
      "epoch": 5.868334822758415,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.026709425022196396,
      "loss": 0.8058,
      "num_input_tokens_seen": 22865008,
      "step": 39400
    },
    {
      "epoch": 5.869079535299375,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.0267082063989747,
      "loss": 0.7924,
      "num_input_tokens_seen": 22867856,
      "step": 39405
    },
    {
      "epoch": 5.8698242478403335,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.02670698757795475,
      "loss": 0.8023,
      "num_input_tokens_seen": 22871888,
      "step": 39410
    },
    {
      "epoch": 5.870568960381293,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.026705768559157152,
      "loss": 0.8003,
      "num_input_tokens_seen": 22874608,
      "step": 39415
    },
    {
      "epoch": 5.871313672922252,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02670454934260249,
      "loss": 0.8046,
      "num_input_tokens_seen": 22877520,
      "step": 39420
    },
    {
      "epoch": 5.872058385463212,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02670332992831136,
      "loss": 0.7973,
      "num_input_tokens_seen": 22880496,
      "step": 39425
    },
    {
      "epoch": 5.87280309800417,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.026702110316304366,
      "loss": 0.8005,
      "num_input_tokens_seen": 22883056,
      "step": 39430
    },
    {
      "epoch": 5.87354781054513,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.026700890506602112,
      "loss": 0.8117,
      "num_input_tokens_seen": 22886160,
      "step": 39435
    },
    {
      "epoch": 5.874292523086089,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.026699670499225205,
      "loss": 0.7922,
      "num_input_tokens_seen": 22889264,
      "step": 39440
    },
    {
      "epoch": 5.8750372356270475,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.026698450294194254,
      "loss": 0.8034,
      "num_input_tokens_seen": 22892400,
      "step": 39445
    },
    {
      "epoch": 5.875781948168007,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02669722989152988,
      "loss": 0.7932,
      "num_input_tokens_seen": 22895280,
      "step": 39450
    },
    {
      "epoch": 5.876526660708967,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.026696009291252692,
      "loss": 0.8116,
      "num_input_tokens_seen": 22898384,
      "step": 39455
    },
    {
      "epoch": 5.8772713732499255,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.026694788493383317,
      "loss": 0.8018,
      "num_input_tokens_seen": 22901232,
      "step": 39460
    },
    {
      "epoch": 5.878016085790884,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.026693567497942373,
      "loss": 0.8129,
      "num_input_tokens_seen": 22904304,
      "step": 39465
    },
    {
      "epoch": 5.878760798331844,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.026692346304950494,
      "loss": 0.7973,
      "num_input_tokens_seen": 22907184,
      "step": 39470
    },
    {
      "epoch": 5.879505510872804,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.026691124914428304,
      "loss": 0.8169,
      "num_input_tokens_seen": 22909808,
      "step": 39475
    },
    {
      "epoch": 5.880250223413762,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.026689903326396442,
      "loss": 0.7885,
      "num_input_tokens_seen": 22912464,
      "step": 39480
    },
    {
      "epoch": 5.880994935954721,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.026688681540875548,
      "loss": 0.8192,
      "num_input_tokens_seen": 22915280,
      "step": 39485
    },
    {
      "epoch": 5.881739648495681,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.026687459557886255,
      "loss": 0.8026,
      "num_input_tokens_seen": 22918160,
      "step": 39490
    },
    {
      "epoch": 5.8824843610366395,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.026686237377449216,
      "loss": 0.797,
      "num_input_tokens_seen": 22921008,
      "step": 39495
    },
    {
      "epoch": 5.883229073577599,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.026685014999585067,
      "loss": 0.7899,
      "num_input_tokens_seen": 22923504,
      "step": 39500
    },
    {
      "epoch": 5.883973786118558,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02668379242431447,
      "loss": 0.8175,
      "num_input_tokens_seen": 22926416,
      "step": 39505
    },
    {
      "epoch": 5.884718498659518,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.026682569651658074,
      "loss": 0.7919,
      "num_input_tokens_seen": 22929168,
      "step": 39510
    },
    {
      "epoch": 5.885463211200476,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02668134668163654,
      "loss": 0.7864,
      "num_input_tokens_seen": 22931952,
      "step": 39515
    },
    {
      "epoch": 5.886207923741436,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.026680123514270522,
      "loss": 0.7986,
      "num_input_tokens_seen": 22935088,
      "step": 39520
    },
    {
      "epoch": 5.886952636282395,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.026678900149580694,
      "loss": 0.8155,
      "num_input_tokens_seen": 22937872,
      "step": 39525
    },
    {
      "epoch": 5.887697348823354,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.026677676587587715,
      "loss": 0.7932,
      "num_input_tokens_seen": 22940784,
      "step": 39530
    },
    {
      "epoch": 5.888442061364313,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026676452828312255,
      "loss": 0.7656,
      "num_input_tokens_seen": 22943664,
      "step": 39535
    },
    {
      "epoch": 5.889186773905273,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.026675228871774998,
      "loss": 0.7899,
      "num_input_tokens_seen": 22946320,
      "step": 39540
    },
    {
      "epoch": 5.8899314864462315,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02667400471799661,
      "loss": 0.8167,
      "num_input_tokens_seen": 22949648,
      "step": 39545
    },
    {
      "epoch": 5.890676198987191,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02667278036699778,
      "loss": 0.8113,
      "num_input_tokens_seen": 22952560,
      "step": 39550
    },
    {
      "epoch": 5.89142091152815,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.026671555818799188,
      "loss": 0.8057,
      "num_input_tokens_seen": 22955408,
      "step": 39555
    },
    {
      "epoch": 5.89216562406911,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.026670331073421524,
      "loss": 0.8105,
      "num_input_tokens_seen": 22958576,
      "step": 39560
    },
    {
      "epoch": 5.892910336610068,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02666910613088548,
      "loss": 0.8101,
      "num_input_tokens_seen": 22961616,
      "step": 39565
    },
    {
      "epoch": 5.893655049151028,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.026667880991211743,
      "loss": 0.8301,
      "num_input_tokens_seen": 22964528,
      "step": 39570
    },
    {
      "epoch": 5.894399761691987,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02666665565442102,
      "loss": 0.8108,
      "num_input_tokens_seen": 22967344,
      "step": 39575
    },
    {
      "epoch": 5.895144474232946,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.026665430120534003,
      "loss": 0.8186,
      "num_input_tokens_seen": 22970416,
      "step": 39580
    },
    {
      "epoch": 5.895889186773905,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.0266642043895714,
      "loss": 0.7934,
      "num_input_tokens_seen": 22973296,
      "step": 39585
    },
    {
      "epoch": 5.896633899314865,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.026662978461553924,
      "loss": 0.7904,
      "num_input_tokens_seen": 22976272,
      "step": 39590
    },
    {
      "epoch": 5.897378611855824,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.026661752336502278,
      "loss": 0.8076,
      "num_input_tokens_seen": 22978992,
      "step": 39595
    },
    {
      "epoch": 5.898123324396783,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.026660526014437178,
      "loss": 0.805,
      "num_input_tokens_seen": 22981840,
      "step": 39600
    },
    {
      "epoch": 5.898868036937742,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.026659299495379343,
      "loss": 0.7869,
      "num_input_tokens_seen": 22984752,
      "step": 39605
    },
    {
      "epoch": 5.899612749478701,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.026658072779349496,
      "loss": 0.8127,
      "num_input_tokens_seen": 22988048,
      "step": 39610
    },
    {
      "epoch": 5.90035746201966,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02665684586636835,
      "loss": 0.8102,
      "num_input_tokens_seen": 22990864,
      "step": 39615
    },
    {
      "epoch": 5.90110217456062,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.026655618756456646,
      "loss": 0.7819,
      "num_input_tokens_seen": 22993584,
      "step": 39620
    },
    {
      "epoch": 5.901846887101579,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.026654391449635106,
      "loss": 0.7904,
      "num_input_tokens_seen": 22996592,
      "step": 39625
    },
    {
      "epoch": 5.9025915996425375,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02665316394592447,
      "loss": 0.801,
      "num_input_tokens_seen": 22999888,
      "step": 39630
    },
    {
      "epoch": 5.903336312183497,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.026651936245345474,
      "loss": 0.8087,
      "num_input_tokens_seen": 23002736,
      "step": 39635
    },
    {
      "epoch": 5.904081024724457,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02665070834791886,
      "loss": 0.8195,
      "num_input_tokens_seen": 23005392,
      "step": 39640
    },
    {
      "epoch": 5.904825737265416,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.026649480253665363,
      "loss": 0.8016,
      "num_input_tokens_seen": 23008304,
      "step": 39645
    },
    {
      "epoch": 5.905570449806374,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.026648251962605743,
      "loss": 0.7654,
      "num_input_tokens_seen": 23011056,
      "step": 39650
    },
    {
      "epoch": 5.906315162347334,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.02664702347476074,
      "loss": 0.7909,
      "num_input_tokens_seen": 23013904,
      "step": 39655
    },
    {
      "epoch": 5.907059874888293,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02664579479015112,
      "loss": 0.8039,
      "num_input_tokens_seen": 23016944,
      "step": 39660
    },
    {
      "epoch": 5.907804587429252,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.026644565908797626,
      "loss": 0.7778,
      "num_input_tokens_seen": 23019824,
      "step": 39665
    },
    {
      "epoch": 5.908549299970211,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.026643336830721027,
      "loss": 0.7866,
      "num_input_tokens_seen": 23022704,
      "step": 39670
    },
    {
      "epoch": 5.909294012511171,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02664210755594209,
      "loss": 0.7674,
      "num_input_tokens_seen": 23025424,
      "step": 39675
    },
    {
      "epoch": 5.91003872505213,
      "grad_norm": 0.048828125,
      "learning_rate": 0.026640878084481574,
      "loss": 0.7885,
      "num_input_tokens_seen": 23028176,
      "step": 39680
    },
    {
      "epoch": 5.910783437593089,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.026639648416360254,
      "loss": 0.8318,
      "num_input_tokens_seen": 23030992,
      "step": 39685
    },
    {
      "epoch": 5.911528150134048,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02663841855159891,
      "loss": 0.8073,
      "num_input_tokens_seen": 23033552,
      "step": 39690
    },
    {
      "epoch": 5.912272862675008,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02663718849021831,
      "loss": 0.8156,
      "num_input_tokens_seen": 23036336,
      "step": 39695
    },
    {
      "epoch": 5.913017575215966,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.026635958232239233,
      "loss": 0.792,
      "num_input_tokens_seen": 23039152,
      "step": 39700
    },
    {
      "epoch": 5.913762287756926,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026634727777682474,
      "loss": 0.8269,
      "num_input_tokens_seen": 23042128,
      "step": 39705
    },
    {
      "epoch": 5.914507000297885,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.026633497126568813,
      "loss": 0.8115,
      "num_input_tokens_seen": 23044976,
      "step": 39710
    },
    {
      "epoch": 5.915251712838844,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.02663226627891904,
      "loss": 0.788,
      "num_input_tokens_seen": 23047888,
      "step": 39715
    },
    {
      "epoch": 5.915996425379803,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.026631035234753953,
      "loss": 0.8026,
      "num_input_tokens_seen": 23050672,
      "step": 39720
    },
    {
      "epoch": 5.916741137920763,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.026629803994094346,
      "loss": 0.8093,
      "num_input_tokens_seen": 23053328,
      "step": 39725
    },
    {
      "epoch": 5.917485850461722,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02662857255696102,
      "loss": 0.8147,
      "num_input_tokens_seen": 23056656,
      "step": 39730
    },
    {
      "epoch": 5.918230563002681,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02662734092337478,
      "loss": 0.7891,
      "num_input_tokens_seen": 23059440,
      "step": 39735
    },
    {
      "epoch": 5.91897527554364,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.026626109093356432,
      "loss": 0.805,
      "num_input_tokens_seen": 23062416,
      "step": 39740
    },
    {
      "epoch": 5.9197199880846,
      "grad_norm": 0.03125,
      "learning_rate": 0.02662487706692679,
      "loss": 0.8246,
      "num_input_tokens_seen": 23065040,
      "step": 39745
    },
    {
      "epoch": 5.920464700625558,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.02662364484410666,
      "loss": 0.8181,
      "num_input_tokens_seen": 23068112,
      "step": 39750
    },
    {
      "epoch": 5.921209413166518,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.026622412424916867,
      "loss": 0.7969,
      "num_input_tokens_seen": 23071056,
      "step": 39755
    },
    {
      "epoch": 5.921954125707477,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02662117980937823,
      "loss": 0.7892,
      "num_input_tokens_seen": 23073904,
      "step": 39760
    },
    {
      "epoch": 5.9226988382484365,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.02661994699751157,
      "loss": 0.8092,
      "num_input_tokens_seen": 23076880,
      "step": 39765
    },
    {
      "epoch": 5.923443550789395,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02661871398933771,
      "loss": 0.8188,
      "num_input_tokens_seen": 23079600,
      "step": 39770
    },
    {
      "epoch": 5.924188263330355,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.026617480784877492,
      "loss": 0.8185,
      "num_input_tokens_seen": 23082640,
      "step": 39775
    },
    {
      "epoch": 5.924932975871314,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02661624738415174,
      "loss": 0.813,
      "num_input_tokens_seen": 23085456,
      "step": 39780
    },
    {
      "epoch": 5.925677688412273,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.026615013787181303,
      "loss": 0.8134,
      "num_input_tokens_seen": 23087984,
      "step": 39785
    },
    {
      "epoch": 5.926422400953232,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.026613779993987004,
      "loss": 0.802,
      "num_input_tokens_seen": 23090640,
      "step": 39790
    },
    {
      "epoch": 5.927167113494191,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.0266125460045897,
      "loss": 0.8163,
      "num_input_tokens_seen": 23093808,
      "step": 39795
    },
    {
      "epoch": 5.92791182603515,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.026611311819010233,
      "loss": 0.8058,
      "num_input_tokens_seen": 23096688,
      "step": 39800
    },
    {
      "epoch": 5.92865653857611,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.026610077437269453,
      "loss": 0.7926,
      "num_input_tokens_seen": 23099376,
      "step": 39805
    },
    {
      "epoch": 5.929401251117069,
      "grad_norm": 0.01080322265625,
      "learning_rate": 0.026608842859388217,
      "loss": 0.8104,
      "num_input_tokens_seen": 23102256,
      "step": 39810
    },
    {
      "epoch": 5.930145963658028,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.026607608085387376,
      "loss": 0.826,
      "num_input_tokens_seen": 23105008,
      "step": 39815
    },
    {
      "epoch": 5.930890676198987,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.026606373115287796,
      "loss": 0.811,
      "num_input_tokens_seen": 23107760,
      "step": 39820
    },
    {
      "epoch": 5.931635388739946,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.026605137949110336,
      "loss": 0.8008,
      "num_input_tokens_seen": 23110544,
      "step": 39825
    },
    {
      "epoch": 5.932380101280906,
      "grad_norm": 0.03125,
      "learning_rate": 0.02660390258687587,
      "loss": 0.8085,
      "num_input_tokens_seen": 23113200,
      "step": 39830
    },
    {
      "epoch": 5.933124813821864,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.026602667028605257,
      "loss": 0.8064,
      "num_input_tokens_seen": 23116112,
      "step": 39835
    },
    {
      "epoch": 5.933869526362824,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.026601431274319385,
      "loss": 0.8119,
      "num_input_tokens_seen": 23119024,
      "step": 39840
    },
    {
      "epoch": 5.934614238903783,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.026600195324039115,
      "loss": 0.8089,
      "num_input_tokens_seen": 23122320,
      "step": 39845
    },
    {
      "epoch": 5.9353589514447425,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.026598959177785337,
      "loss": 0.8145,
      "num_input_tokens_seen": 23125136,
      "step": 39850
    },
    {
      "epoch": 5.936103663985701,
      "grad_norm": 0.010986328125,
      "learning_rate": 0.026597722835578937,
      "loss": 0.8167,
      "num_input_tokens_seen": 23127952,
      "step": 39855
    },
    {
      "epoch": 5.936848376526661,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.026596486297440797,
      "loss": 0.8105,
      "num_input_tokens_seen": 23130704,
      "step": 39860
    },
    {
      "epoch": 5.93759308906762,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.0265952495633918,
      "loss": 0.8055,
      "num_input_tokens_seen": 23133296,
      "step": 39865
    },
    {
      "epoch": 5.938337801608579,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.026594012633452853,
      "loss": 0.8042,
      "num_input_tokens_seen": 23136176,
      "step": 39870
    },
    {
      "epoch": 5.939082514149538,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.026592775507644842,
      "loss": 0.8079,
      "num_input_tokens_seen": 23138928,
      "step": 39875
    },
    {
      "epoch": 5.939827226690498,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.026591538185988677,
      "loss": 0.8031,
      "num_input_tokens_seen": 23141904,
      "step": 39880
    },
    {
      "epoch": 5.940571939231456,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02659030066850525,
      "loss": 0.7991,
      "num_input_tokens_seen": 23144688,
      "step": 39885
    },
    {
      "epoch": 5.941316651772416,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.026589062955215473,
      "loss": 0.7974,
      "num_input_tokens_seen": 23147792,
      "step": 39890
    },
    {
      "epoch": 5.942061364313375,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.026587825046140257,
      "loss": 0.7866,
      "num_input_tokens_seen": 23150576,
      "step": 39895
    },
    {
      "epoch": 5.9428060768543345,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.026586586941300516,
      "loss": 0.8279,
      "num_input_tokens_seen": 23153680,
      "step": 39900
    },
    {
      "epoch": 5.943550789395293,
      "grad_norm": 0.01080322265625,
      "learning_rate": 0.026585348640717165,
      "loss": 0.7971,
      "num_input_tokens_seen": 23156976,
      "step": 39905
    },
    {
      "epoch": 5.944295501936253,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02658411014441112,
      "loss": 0.8037,
      "num_input_tokens_seen": 23160304,
      "step": 39910
    },
    {
      "epoch": 5.945040214477212,
      "grad_norm": 0.0234375,
      "learning_rate": 0.026582871452403307,
      "loss": 0.8003,
      "num_input_tokens_seen": 23163568,
      "step": 39915
    },
    {
      "epoch": 5.945784927018171,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.026581632564714658,
      "loss": 0.8023,
      "num_input_tokens_seen": 23166288,
      "step": 39920
    },
    {
      "epoch": 5.94652963955913,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.026580393481366097,
      "loss": 0.8107,
      "num_input_tokens_seen": 23169072,
      "step": 39925
    },
    {
      "epoch": 5.94727435210009,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.026579154202378556,
      "loss": 0.8044,
      "num_input_tokens_seen": 23171920,
      "step": 39930
    },
    {
      "epoch": 5.9480190646410485,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02657791472777297,
      "loss": 0.7993,
      "num_input_tokens_seen": 23174672,
      "step": 39935
    },
    {
      "epoch": 5.948763777182008,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02657667505757029,
      "loss": 0.8109,
      "num_input_tokens_seen": 23177456,
      "step": 39940
    },
    {
      "epoch": 5.949508489722967,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.026575435191791445,
      "loss": 0.7898,
      "num_input_tokens_seen": 23180400,
      "step": 39945
    },
    {
      "epoch": 5.9502532022639265,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02657419513045739,
      "loss": 0.7839,
      "num_input_tokens_seen": 23183056,
      "step": 39950
    },
    {
      "epoch": 5.950997914804885,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.02657295487358907,
      "loss": 0.8004,
      "num_input_tokens_seen": 23185840,
      "step": 39955
    },
    {
      "epoch": 5.951742627345844,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.02657171442120744,
      "loss": 0.8252,
      "num_input_tokens_seen": 23188816,
      "step": 39960
    },
    {
      "epoch": 5.952487339886804,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.026570473773333454,
      "loss": 0.7979,
      "num_input_tokens_seen": 23191376,
      "step": 39965
    },
    {
      "epoch": 5.953232052427763,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.026569232929988074,
      "loss": 0.7989,
      "num_input_tokens_seen": 23194224,
      "step": 39970
    },
    {
      "epoch": 5.953976764968722,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.026567991891192265,
      "loss": 0.8191,
      "num_input_tokens_seen": 23197008,
      "step": 39975
    },
    {
      "epoch": 5.954721477509681,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02656675065696699,
      "loss": 0.7862,
      "num_input_tokens_seen": 23200080,
      "step": 39980
    },
    {
      "epoch": 5.9554661900506405,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.02656550922733322,
      "loss": 0.7918,
      "num_input_tokens_seen": 23203024,
      "step": 39985
    },
    {
      "epoch": 5.9562109025916,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02656426760231192,
      "loss": 0.8049,
      "num_input_tokens_seen": 23206288,
      "step": 39990
    },
    {
      "epoch": 5.956955615132559,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.026563025781924078,
      "loss": 0.8186,
      "num_input_tokens_seen": 23209136,
      "step": 39995
    },
    {
      "epoch": 5.957700327673518,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02656178376619066,
      "loss": 0.7986,
      "num_input_tokens_seen": 23212272,
      "step": 40000
    },
    {
      "epoch": 5.958445040214477,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.026560541555132662,
      "loss": 0.8036,
      "num_input_tokens_seen": 23215088,
      "step": 40005
    },
    {
      "epoch": 5.959189752755436,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.026559299148771066,
      "loss": 0.7955,
      "num_input_tokens_seen": 23218000,
      "step": 40010
    },
    {
      "epoch": 5.959934465296396,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.026558056547126858,
      "loss": 0.8193,
      "num_input_tokens_seen": 23220880,
      "step": 40015
    },
    {
      "epoch": 5.9606791778373545,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.02655681375022103,
      "loss": 0.8041,
      "num_input_tokens_seen": 23223728,
      "step": 40020
    },
    {
      "epoch": 5.961423890378314,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.026555570758074583,
      "loss": 0.8216,
      "num_input_tokens_seen": 23226608,
      "step": 40025
    },
    {
      "epoch": 5.962168602919273,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.026554327570708504,
      "loss": 0.8206,
      "num_input_tokens_seen": 23229264,
      "step": 40030
    },
    {
      "epoch": 5.9629133154602325,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.026553084188143807,
      "loss": 0.794,
      "num_input_tokens_seen": 23232368,
      "step": 40035
    },
    {
      "epoch": 5.963658028001191,
      "grad_norm": 0.01025390625,
      "learning_rate": 0.0265518406104015,
      "loss": 0.7869,
      "num_input_tokens_seen": 23235120,
      "step": 40040
    },
    {
      "epoch": 5.964402740542151,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.026550596837502587,
      "loss": 0.8063,
      "num_input_tokens_seen": 23238192,
      "step": 40045
    },
    {
      "epoch": 5.96514745308311,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.026549352869468076,
      "loss": 0.8037,
      "num_input_tokens_seen": 23241328,
      "step": 40050
    },
    {
      "epoch": 5.965892165624069,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.026548108706318987,
      "loss": 0.796,
      "num_input_tokens_seen": 23244304,
      "step": 40055
    },
    {
      "epoch": 5.966636878165028,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02654686434807634,
      "loss": 0.7927,
      "num_input_tokens_seen": 23247056,
      "step": 40060
    },
    {
      "epoch": 5.967381590705988,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.026545619794761156,
      "loss": 0.8145,
      "num_input_tokens_seen": 23250128,
      "step": 40065
    },
    {
      "epoch": 5.9681263032469465,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.026544375046394453,
      "loss": 0.8101,
      "num_input_tokens_seen": 23253008,
      "step": 40070
    },
    {
      "epoch": 5.968871015787906,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.026543130102997272,
      "loss": 0.7879,
      "num_input_tokens_seen": 23255504,
      "step": 40075
    },
    {
      "epoch": 5.969615728328865,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02654188496459064,
      "loss": 0.7991,
      "num_input_tokens_seen": 23258288,
      "step": 40080
    },
    {
      "epoch": 5.9703604408698245,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026540639631195594,
      "loss": 0.7988,
      "num_input_tokens_seen": 23261168,
      "step": 40085
    },
    {
      "epoch": 5.971105153410783,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02653939410283317,
      "loss": 0.8119,
      "num_input_tokens_seen": 23263888,
      "step": 40090
    },
    {
      "epoch": 5.971849865951743,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02653814837952441,
      "loss": 0.7986,
      "num_input_tokens_seen": 23266992,
      "step": 40095
    },
    {
      "epoch": 5.972594578492702,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.026536902461290354,
      "loss": 0.8091,
      "num_input_tokens_seen": 23269904,
      "step": 40100
    },
    {
      "epoch": 5.973339291033661,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.026535656348152064,
      "loss": 0.7886,
      "num_input_tokens_seen": 23272752,
      "step": 40105
    },
    {
      "epoch": 5.97408400357462,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.026534410040130584,
      "loss": 0.8137,
      "num_input_tokens_seen": 23275600,
      "step": 40110
    },
    {
      "epoch": 5.97482871611558,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.026533163537246964,
      "loss": 0.8063,
      "num_input_tokens_seen": 23278416,
      "step": 40115
    },
    {
      "epoch": 5.9755734286565385,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02653191683952227,
      "loss": 0.8015,
      "num_input_tokens_seen": 23281424,
      "step": 40120
    },
    {
      "epoch": 5.976318141197497,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02653066994697757,
      "loss": 0.8029,
      "num_input_tokens_seen": 23284240,
      "step": 40125
    },
    {
      "epoch": 5.977062853738457,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02652942285963391,
      "loss": 0.818,
      "num_input_tokens_seen": 23287056,
      "step": 40130
    },
    {
      "epoch": 5.977807566279417,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.026528175577512372,
      "loss": 0.8122,
      "num_input_tokens_seen": 23289904,
      "step": 40135
    },
    {
      "epoch": 5.978552278820375,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026526928100634027,
      "loss": 0.7966,
      "num_input_tokens_seen": 23292912,
      "step": 40140
    },
    {
      "epoch": 5.979296991361334,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.026525680429019944,
      "loss": 0.8063,
      "num_input_tokens_seen": 23295952,
      "step": 40145
    },
    {
      "epoch": 5.980041703902294,
      "grad_norm": 0.033203125,
      "learning_rate": 0.026524432562691203,
      "loss": 0.8201,
      "num_input_tokens_seen": 23298896,
      "step": 40150
    },
    {
      "epoch": 5.980786416443253,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.026523184501668892,
      "loss": 0.8016,
      "num_input_tokens_seen": 23301840,
      "step": 40155
    },
    {
      "epoch": 5.981531128984212,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.026521936245974084,
      "loss": 0.8002,
      "num_input_tokens_seen": 23304752,
      "step": 40160
    },
    {
      "epoch": 5.982275841525171,
      "grad_norm": 0.017578125,
      "learning_rate": 0.02652068779562788,
      "loss": 0.7952,
      "num_input_tokens_seen": 23307760,
      "step": 40165
    },
    {
      "epoch": 5.9830205540661305,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.026519439150651358,
      "loss": 0.808,
      "num_input_tokens_seen": 23310608,
      "step": 40170
    },
    {
      "epoch": 5.983765266607089,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.026518190311065625,
      "loss": 0.7981,
      "num_input_tokens_seen": 23313456,
      "step": 40175
    },
    {
      "epoch": 5.984509979148049,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.026516941276891772,
      "loss": 0.8119,
      "num_input_tokens_seen": 23316752,
      "step": 40180
    },
    {
      "epoch": 5.985254691689008,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.0265156920481509,
      "loss": 0.789,
      "num_input_tokens_seen": 23319568,
      "step": 40185
    },
    {
      "epoch": 5.985999404229967,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.026514442624864118,
      "loss": 0.8002,
      "num_input_tokens_seen": 23322416,
      "step": 40190
    },
    {
      "epoch": 5.986744116770926,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026513193007052526,
      "loss": 0.7795,
      "num_input_tokens_seen": 23325232,
      "step": 40195
    },
    {
      "epoch": 5.987488829311886,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.026511943194737243,
      "loss": 0.7847,
      "num_input_tokens_seen": 23328272,
      "step": 40200
    },
    {
      "epoch": 5.9882335418528445,
      "grad_norm": 0.017578125,
      "learning_rate": 0.026510693187939378,
      "loss": 0.7838,
      "num_input_tokens_seen": 23331376,
      "step": 40205
    },
    {
      "epoch": 5.988978254393804,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02650944298668005,
      "loss": 0.8135,
      "num_input_tokens_seen": 23334352,
      "step": 40210
    },
    {
      "epoch": 5.989722966934763,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02650819259098038,
      "loss": 0.7883,
      "num_input_tokens_seen": 23337296,
      "step": 40215
    },
    {
      "epoch": 5.990467679475723,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.0265069420008615,
      "loss": 0.8037,
      "num_input_tokens_seen": 23340432,
      "step": 40220
    },
    {
      "epoch": 5.991212392016681,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.026505691216344524,
      "loss": 0.8193,
      "num_input_tokens_seen": 23343120,
      "step": 40225
    },
    {
      "epoch": 5.991957104557641,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.026504440237450584,
      "loss": 0.8121,
      "num_input_tokens_seen": 23345936,
      "step": 40230
    },
    {
      "epoch": 5.9927018170986,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.026503189064200825,
      "loss": 0.8388,
      "num_input_tokens_seen": 23348912,
      "step": 40235
    },
    {
      "epoch": 5.993446529639559,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02650193769661638,
      "loss": 0.7864,
      "num_input_tokens_seen": 23352208,
      "step": 40240
    },
    {
      "epoch": 5.994191242180518,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.026500686134718385,
      "loss": 0.8144,
      "num_input_tokens_seen": 23355376,
      "step": 40245
    },
    {
      "epoch": 5.994935954721478,
      "grad_norm": 0.025390625,
      "learning_rate": 0.026499434378527986,
      "loss": 0.7952,
      "num_input_tokens_seen": 23358160,
      "step": 40250
    },
    {
      "epoch": 5.9956806672624365,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02649818242806633,
      "loss": 0.815,
      "num_input_tokens_seen": 23361040,
      "step": 40255
    },
    {
      "epoch": 5.996425379803396,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.026496930283354567,
      "loss": 0.8053,
      "num_input_tokens_seen": 23364016,
      "step": 40260
    },
    {
      "epoch": 5.997170092344355,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.026495677944413855,
      "loss": 0.8148,
      "num_input_tokens_seen": 23366800,
      "step": 40265
    },
    {
      "epoch": 5.997914804885315,
      "grad_norm": 0.0111083984375,
      "learning_rate": 0.026494425411265347,
      "loss": 0.8127,
      "num_input_tokens_seen": 23369296,
      "step": 40270
    },
    {
      "epoch": 5.998659517426273,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.026493172683930203,
      "loss": 0.8071,
      "num_input_tokens_seen": 23372112,
      "step": 40275
    },
    {
      "epoch": 5.999404229967233,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.026491919762429587,
      "loss": 0.815,
      "num_input_tokens_seen": 23374992,
      "step": 40280
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.8028715252876282,
      "eval_runtime": 70.6804,
      "eval_samples_per_second": 42.218,
      "eval_steps_per_second": 10.555,
      "num_input_tokens_seen": 23376776,
      "step": 40284
    },
    {
      "epoch": 6.000148942508192,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.026490666646784667,
      "loss": 0.7954,
      "num_input_tokens_seen": 23377352,
      "step": 40285
    },
    {
      "epoch": 6.000893655049151,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.026489413337016616,
      "loss": 0.8146,
      "num_input_tokens_seen": 23379944,
      "step": 40290
    },
    {
      "epoch": 6.00163836759011,
      "grad_norm": 0.0234375,
      "learning_rate": 0.026488159833146604,
      "loss": 0.8028,
      "num_input_tokens_seen": 23382792,
      "step": 40295
    },
    {
      "epoch": 6.00238308013107,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.026486906135195804,
      "loss": 0.8099,
      "num_input_tokens_seen": 23385960,
      "step": 40300
    },
    {
      "epoch": 6.003127792672029,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.0264856522431854,
      "loss": 0.8159,
      "num_input_tokens_seen": 23389160,
      "step": 40305
    },
    {
      "epoch": 6.003872505212988,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.026484398157136574,
      "loss": 0.7987,
      "num_input_tokens_seen": 23392040,
      "step": 40310
    },
    {
      "epoch": 6.004617217753947,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.026483143877070518,
      "loss": 0.8204,
      "num_input_tokens_seen": 23394984,
      "step": 40315
    },
    {
      "epoch": 6.005361930294906,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.026481889403008413,
      "loss": 0.8126,
      "num_input_tokens_seen": 23397768,
      "step": 40320
    },
    {
      "epoch": 6.006106642835865,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.026480634734971458,
      "loss": 0.8167,
      "num_input_tokens_seen": 23400648,
      "step": 40325
    },
    {
      "epoch": 6.006851355376824,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.026479379872980845,
      "loss": 0.8028,
      "num_input_tokens_seen": 23403720,
      "step": 40330
    },
    {
      "epoch": 6.007596067917784,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.026478124817057778,
      "loss": 0.8111,
      "num_input_tokens_seen": 23406664,
      "step": 40335
    },
    {
      "epoch": 6.0083407804587425,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.02647686956722346,
      "loss": 0.7979,
      "num_input_tokens_seen": 23409352,
      "step": 40340
    },
    {
      "epoch": 6.009085492999702,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02647561412349909,
      "loss": 0.7953,
      "num_input_tokens_seen": 23412104,
      "step": 40345
    },
    {
      "epoch": 6.009830205540661,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02647435848590589,
      "loss": 0.8078,
      "num_input_tokens_seen": 23414888,
      "step": 40350
    },
    {
      "epoch": 6.010574918081621,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02647310265446506,
      "loss": 0.8053,
      "num_input_tokens_seen": 23417608,
      "step": 40355
    },
    {
      "epoch": 6.011319630622579,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.026471846629197823,
      "loss": 0.7956,
      "num_input_tokens_seen": 23420584,
      "step": 40360
    },
    {
      "epoch": 6.012064343163539,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026470590410125394,
      "loss": 0.7988,
      "num_input_tokens_seen": 23423528,
      "step": 40365
    },
    {
      "epoch": 6.012809055704498,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.026469333997269,
      "loss": 0.806,
      "num_input_tokens_seen": 23426280,
      "step": 40370
    },
    {
      "epoch": 6.013553768245457,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.026468077390649866,
      "loss": 0.794,
      "num_input_tokens_seen": 23429416,
      "step": 40375
    },
    {
      "epoch": 6.014298480786416,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.026466820590289222,
      "loss": 0.8207,
      "num_input_tokens_seen": 23432360,
      "step": 40380
    },
    {
      "epoch": 6.015043193327376,
      "grad_norm": 0.025390625,
      "learning_rate": 0.026465563596208296,
      "loss": 0.7922,
      "num_input_tokens_seen": 23435336,
      "step": 40385
    },
    {
      "epoch": 6.015787905868335,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.026464306408428326,
      "loss": 0.7879,
      "num_input_tokens_seen": 23438408,
      "step": 40390
    },
    {
      "epoch": 6.016532618409294,
      "grad_norm": 0.017578125,
      "learning_rate": 0.026463049026970553,
      "loss": 0.8054,
      "num_input_tokens_seen": 23441416,
      "step": 40395
    },
    {
      "epoch": 6.017277330950253,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02646179145185621,
      "loss": 0.8142,
      "num_input_tokens_seen": 23444392,
      "step": 40400
    },
    {
      "epoch": 6.018022043491213,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.026460533683106555,
      "loss": 0.7965,
      "num_input_tokens_seen": 23447272,
      "step": 40405
    },
    {
      "epoch": 6.018766756032171,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026459275720742834,
      "loss": 0.8062,
      "num_input_tokens_seen": 23450344,
      "step": 40410
    },
    {
      "epoch": 6.019511468573131,
      "grad_norm": 0.0111083984375,
      "learning_rate": 0.026458017564786294,
      "loss": 0.8125,
      "num_input_tokens_seen": 23453192,
      "step": 40415
    },
    {
      "epoch": 6.02025618111409,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.026456759215258193,
      "loss": 0.8032,
      "num_input_tokens_seen": 23456040,
      "step": 40420
    },
    {
      "epoch": 6.021000893655049,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.026455500672179787,
      "loss": 0.7908,
      "num_input_tokens_seen": 23458664,
      "step": 40425
    },
    {
      "epoch": 6.021745606196008,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.026454241935572345,
      "loss": 0.8181,
      "num_input_tokens_seen": 23461512,
      "step": 40430
    },
    {
      "epoch": 6.022490318736968,
      "grad_norm": 0.01953125,
      "learning_rate": 0.026452983005457124,
      "loss": 0.7959,
      "num_input_tokens_seen": 23464136,
      "step": 40435
    },
    {
      "epoch": 6.023235031277927,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.026451723881855394,
      "loss": 0.8031,
      "num_input_tokens_seen": 23466920,
      "step": 40440
    },
    {
      "epoch": 6.023979743818886,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.026450464564788428,
      "loss": 0.8144,
      "num_input_tokens_seen": 23469640,
      "step": 40445
    },
    {
      "epoch": 6.024724456359845,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.026449205054277503,
      "loss": 0.8042,
      "num_input_tokens_seen": 23472552,
      "step": 40450
    },
    {
      "epoch": 6.025469168900805,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.026447945350343897,
      "loss": 0.8077,
      "num_input_tokens_seen": 23475688,
      "step": 40455
    },
    {
      "epoch": 6.026213881441763,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.026446685453008885,
      "loss": 0.8154,
      "num_input_tokens_seen": 23478664,
      "step": 40460
    },
    {
      "epoch": 6.026958593982723,
      "grad_norm": 0.01123046875,
      "learning_rate": 0.02644542536229376,
      "loss": 0.7999,
      "num_input_tokens_seen": 23481352,
      "step": 40465
    },
    {
      "epoch": 6.027703306523682,
      "grad_norm": 0.0101318359375,
      "learning_rate": 0.0264441650782198,
      "loss": 0.7985,
      "num_input_tokens_seen": 23484040,
      "step": 40470
    },
    {
      "epoch": 6.0284480190646414,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.026442904600808313,
      "loss": 0.816,
      "num_input_tokens_seen": 23486952,
      "step": 40475
    },
    {
      "epoch": 6.0291927316056,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.026441643930080572,
      "loss": 0.8003,
      "num_input_tokens_seen": 23489768,
      "step": 40480
    },
    {
      "epoch": 6.02993744414656,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.026440383066057893,
      "loss": 0.8164,
      "num_input_tokens_seen": 23492904,
      "step": 40485
    },
    {
      "epoch": 6.030682156687519,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.026439122008761567,
      "loss": 0.8122,
      "num_input_tokens_seen": 23495624,
      "step": 40490
    },
    {
      "epoch": 6.031426869228477,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.0264378607582129,
      "loss": 0.7888,
      "num_input_tokens_seen": 23498664,
      "step": 40495
    },
    {
      "epoch": 6.032171581769437,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.026436599314433196,
      "loss": 0.8028,
      "num_input_tokens_seen": 23501704,
      "step": 40500
    },
    {
      "epoch": 6.032916294310396,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.026435337677443774,
      "loss": 0.807,
      "num_input_tokens_seen": 23504936,
      "step": 40505
    },
    {
      "epoch": 6.033661006851355,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.026434075847265946,
      "loss": 0.8056,
      "num_input_tokens_seen": 23507528,
      "step": 40510
    },
    {
      "epoch": 6.034405719392314,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.026432813823921025,
      "loss": 0.8005,
      "num_input_tokens_seen": 23510312,
      "step": 40515
    },
    {
      "epoch": 6.035150431933274,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02643155160743034,
      "loss": 0.8055,
      "num_input_tokens_seen": 23513416,
      "step": 40520
    },
    {
      "epoch": 6.035895144474233,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.026430289197815204,
      "loss": 0.8154,
      "num_input_tokens_seen": 23516552,
      "step": 40525
    },
    {
      "epoch": 6.036639857015192,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.026429026595096952,
      "loss": 0.8237,
      "num_input_tokens_seen": 23519336,
      "step": 40530
    },
    {
      "epoch": 6.037384569556151,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02642776379929691,
      "loss": 0.7911,
      "num_input_tokens_seen": 23522504,
      "step": 40535
    },
    {
      "epoch": 6.038129282097111,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.026426500810436414,
      "loss": 0.806,
      "num_input_tokens_seen": 23525032,
      "step": 40540
    },
    {
      "epoch": 6.038873994638069,
      "grad_norm": 0.025390625,
      "learning_rate": 0.026425237628536796,
      "loss": 0.7945,
      "num_input_tokens_seen": 23528072,
      "step": 40545
    },
    {
      "epoch": 6.039618707179029,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.026423974253619406,
      "loss": 0.7861,
      "num_input_tokens_seen": 23530888,
      "step": 40550
    },
    {
      "epoch": 6.040363419719988,
      "grad_norm": 0.02734375,
      "learning_rate": 0.026422710685705578,
      "loss": 0.8113,
      "num_input_tokens_seen": 23534024,
      "step": 40555
    },
    {
      "epoch": 6.0411081322609474,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.026421446924816665,
      "loss": 0.7755,
      "num_input_tokens_seen": 23536744,
      "step": 40560
    },
    {
      "epoch": 6.041852844801906,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.026420182970974013,
      "loss": 0.7917,
      "num_input_tokens_seen": 23539720,
      "step": 40565
    },
    {
      "epoch": 6.042597557342866,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02641891882419898,
      "loss": 0.7819,
      "num_input_tokens_seen": 23542664,
      "step": 40570
    },
    {
      "epoch": 6.043342269883825,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.02641765448451291,
      "loss": 0.8231,
      "num_input_tokens_seen": 23545512,
      "step": 40575
    },
    {
      "epoch": 6.044086982424784,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02641638995193718,
      "loss": 0.7948,
      "num_input_tokens_seen": 23548424,
      "step": 40580
    },
    {
      "epoch": 6.044831694965743,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.026415125226493145,
      "loss": 0.7866,
      "num_input_tokens_seen": 23551400,
      "step": 40585
    },
    {
      "epoch": 6.045576407506703,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026413860308202173,
      "loss": 0.812,
      "num_input_tokens_seen": 23554568,
      "step": 40590
    },
    {
      "epoch": 6.046321120047661,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.026412595197085626,
      "loss": 0.7883,
      "num_input_tokens_seen": 23557256,
      "step": 40595
    },
    {
      "epoch": 6.047065832588621,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02641132989316488,
      "loss": 0.7871,
      "num_input_tokens_seen": 23560040,
      "step": 40600
    },
    {
      "epoch": 6.04781054512958,
      "grad_norm": 0.017578125,
      "learning_rate": 0.02641006439646132,
      "loss": 0.8106,
      "num_input_tokens_seen": 23562760,
      "step": 40605
    },
    {
      "epoch": 6.0485552576705395,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.026408798706996316,
      "loss": 0.8016,
      "num_input_tokens_seen": 23565768,
      "step": 40610
    },
    {
      "epoch": 6.049299970211498,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.026407532824791254,
      "loss": 0.8247,
      "num_input_tokens_seen": 23568712,
      "step": 40615
    },
    {
      "epoch": 6.050044682752458,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02640626674986752,
      "loss": 0.8307,
      "num_input_tokens_seen": 23571592,
      "step": 40620
    },
    {
      "epoch": 6.050789395293417,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.026405000482246498,
      "loss": 0.8055,
      "num_input_tokens_seen": 23574600,
      "step": 40625
    },
    {
      "epoch": 6.051534107834376,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02640373402194959,
      "loss": 0.8037,
      "num_input_tokens_seen": 23577448,
      "step": 40630
    },
    {
      "epoch": 6.052278820375335,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.02640246736899818,
      "loss": 0.7992,
      "num_input_tokens_seen": 23580040,
      "step": 40635
    },
    {
      "epoch": 6.053023532916295,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.026401200523413677,
      "loss": 0.8002,
      "num_input_tokens_seen": 23583048,
      "step": 40640
    },
    {
      "epoch": 6.0537682454572534,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.026399933485217478,
      "loss": 0.7959,
      "num_input_tokens_seen": 23585832,
      "step": 40645
    },
    {
      "epoch": 6.054512957998213,
      "grad_norm": 0.01226806640625,
      "learning_rate": 0.026398666254430988,
      "loss": 0.7908,
      "num_input_tokens_seen": 23588808,
      "step": 40650
    },
    {
      "epoch": 6.055257670539172,
      "grad_norm": 0.017578125,
      "learning_rate": 0.026397398831075617,
      "loss": 0.8123,
      "num_input_tokens_seen": 23591656,
      "step": 40655
    },
    {
      "epoch": 6.0560023830801315,
      "grad_norm": 0.025390625,
      "learning_rate": 0.026396131215172775,
      "loss": 0.809,
      "num_input_tokens_seen": 23594184,
      "step": 40660
    },
    {
      "epoch": 6.05674709562109,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.026394863406743883,
      "loss": 0.7866,
      "num_input_tokens_seen": 23597032,
      "step": 40665
    },
    {
      "epoch": 6.057491808162049,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.026393595405810353,
      "loss": 0.8011,
      "num_input_tokens_seen": 23599880,
      "step": 40670
    },
    {
      "epoch": 6.058236520703009,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02639232721239361,
      "loss": 0.7865,
      "num_input_tokens_seen": 23602728,
      "step": 40675
    },
    {
      "epoch": 6.058981233243967,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.026391058826515074,
      "loss": 0.7972,
      "num_input_tokens_seen": 23605640,
      "step": 40680
    },
    {
      "epoch": 6.059725945784927,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.02638979024819618,
      "loss": 0.8177,
      "num_input_tokens_seen": 23608264,
      "step": 40685
    },
    {
      "epoch": 6.060470658325886,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.026388521477458352,
      "loss": 0.8158,
      "num_input_tokens_seen": 23611176,
      "step": 40690
    },
    {
      "epoch": 6.0612153708668455,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02638725251432303,
      "loss": 0.8091,
      "num_input_tokens_seen": 23614344,
      "step": 40695
    },
    {
      "epoch": 6.061960083407804,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.02638598335881165,
      "loss": 0.8151,
      "num_input_tokens_seen": 23617192,
      "step": 40700
    },
    {
      "epoch": 6.062704795948764,
      "grad_norm": 0.011474609375,
      "learning_rate": 0.026384714010945654,
      "loss": 0.7886,
      "num_input_tokens_seen": 23620456,
      "step": 40705
    },
    {
      "epoch": 6.063449508489723,
      "grad_norm": 0.013671875,
      "learning_rate": 0.026383444470746486,
      "loss": 0.8105,
      "num_input_tokens_seen": 23623464,
      "step": 40710
    },
    {
      "epoch": 6.064194221030682,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.026382174738235592,
      "loss": 0.7948,
      "num_input_tokens_seen": 23626696,
      "step": 40715
    },
    {
      "epoch": 6.064938933571641,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.026380904813434424,
      "loss": 0.812,
      "num_input_tokens_seen": 23629608,
      "step": 40720
    },
    {
      "epoch": 6.065683646112601,
      "grad_norm": 0.0118408203125,
      "learning_rate": 0.026379634696364437,
      "loss": 0.7823,
      "num_input_tokens_seen": 23632520,
      "step": 40725
    },
    {
      "epoch": 6.0664283586535594,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.026378364387047087,
      "loss": 0.7999,
      "num_input_tokens_seen": 23635528,
      "step": 40730
    },
    {
      "epoch": 6.067173071194519,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.026377093885503838,
      "loss": 0.7989,
      "num_input_tokens_seen": 23638376,
      "step": 40735
    },
    {
      "epoch": 6.067917783735478,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.026375823191756145,
      "loss": 0.7811,
      "num_input_tokens_seen": 23641192,
      "step": 40740
    },
    {
      "epoch": 6.0686624962764375,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.026374552305825485,
      "loss": 0.8358,
      "num_input_tokens_seen": 23644040,
      "step": 40745
    },
    {
      "epoch": 6.069407208817396,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.026373281227733327,
      "loss": 0.7959,
      "num_input_tokens_seen": 23647048,
      "step": 40750
    },
    {
      "epoch": 6.070151921358356,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02637200995750114,
      "loss": 0.804,
      "num_input_tokens_seen": 23649768,
      "step": 40755
    },
    {
      "epoch": 6.070896633899315,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.0263707384951504,
      "loss": 0.8259,
      "num_input_tokens_seen": 23652648,
      "step": 40760
    },
    {
      "epoch": 6.071641346440274,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.026369466840702595,
      "loss": 0.7867,
      "num_input_tokens_seen": 23655336,
      "step": 40765
    },
    {
      "epoch": 6.072386058981233,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.0263681949941792,
      "loss": 0.8031,
      "num_input_tokens_seen": 23658504,
      "step": 40770
    },
    {
      "epoch": 6.073130771522193,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.026366922955601706,
      "loss": 0.8248,
      "num_input_tokens_seen": 23661288,
      "step": 40775
    },
    {
      "epoch": 6.0738754840631515,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.0263656507249916,
      "loss": 0.794,
      "num_input_tokens_seen": 23664328,
      "step": 40780
    },
    {
      "epoch": 6.074620196604111,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.026364378302370373,
      "loss": 0.7996,
      "num_input_tokens_seen": 23667240,
      "step": 40785
    },
    {
      "epoch": 6.07536490914507,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02636310568775953,
      "loss": 0.7946,
      "num_input_tokens_seen": 23669960,
      "step": 40790
    },
    {
      "epoch": 6.0761096216860295,
      "grad_norm": 0.01092529296875,
      "learning_rate": 0.026361832881180564,
      "loss": 0.8111,
      "num_input_tokens_seen": 23672904,
      "step": 40795
    },
    {
      "epoch": 6.076854334226988,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02636055988265498,
      "loss": 0.7885,
      "num_input_tokens_seen": 23676040,
      "step": 40800
    },
    {
      "epoch": 6.077599046767948,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.026359286692204276,
      "loss": 0.7916,
      "num_input_tokens_seen": 23678984,
      "step": 40805
    },
    {
      "epoch": 6.078343759308907,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.026358013309849975,
      "loss": 0.8009,
      "num_input_tokens_seen": 23681800,
      "step": 40810
    },
    {
      "epoch": 6.079088471849866,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.026356739735613584,
      "loss": 0.8261,
      "num_input_tokens_seen": 23684936,
      "step": 40815
    },
    {
      "epoch": 6.079833184390825,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.026355465969516618,
      "loss": 0.8069,
      "num_input_tokens_seen": 23687816,
      "step": 40820
    },
    {
      "epoch": 6.080577896931785,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02635419201158059,
      "loss": 0.8111,
      "num_input_tokens_seen": 23690760,
      "step": 40825
    },
    {
      "epoch": 6.0813226094727435,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.026352917861827033,
      "loss": 0.8156,
      "num_input_tokens_seen": 23693608,
      "step": 40830
    },
    {
      "epoch": 6.082067322013703,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.026351643520277468,
      "loss": 0.7926,
      "num_input_tokens_seen": 23696360,
      "step": 40835
    },
    {
      "epoch": 6.082812034554662,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02635036898695342,
      "loss": 0.8036,
      "num_input_tokens_seen": 23699080,
      "step": 40840
    },
    {
      "epoch": 6.083556747095621,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.026349094261876423,
      "loss": 0.8129,
      "num_input_tokens_seen": 23702184,
      "step": 40845
    },
    {
      "epoch": 6.08430145963658,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.026347819345068014,
      "loss": 0.8128,
      "num_input_tokens_seen": 23704936,
      "step": 40850
    },
    {
      "epoch": 6.085046172177539,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.026346544236549734,
      "loss": 0.7866,
      "num_input_tokens_seen": 23707496,
      "step": 40855
    },
    {
      "epoch": 6.085790884718499,
      "grad_norm": 0.01129150390625,
      "learning_rate": 0.026345268936343116,
      "loss": 0.8024,
      "num_input_tokens_seen": 23710216,
      "step": 40860
    },
    {
      "epoch": 6.0865355972594575,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.026343993444469714,
      "loss": 0.8242,
      "num_input_tokens_seen": 23713160,
      "step": 40865
    },
    {
      "epoch": 6.087280309800417,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02634271776095107,
      "loss": 0.8146,
      "num_input_tokens_seen": 23716200,
      "step": 40870
    },
    {
      "epoch": 6.088025022341376,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02634144188580874,
      "loss": 0.8121,
      "num_input_tokens_seen": 23719464,
      "step": 40875
    },
    {
      "epoch": 6.0887697348823355,
      "grad_norm": 0.01953125,
      "learning_rate": 0.026340165819064277,
      "loss": 0.7726,
      "num_input_tokens_seen": 23722312,
      "step": 40880
    },
    {
      "epoch": 6.089514447423294,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.026338889560739234,
      "loss": 0.7953,
      "num_input_tokens_seen": 23724968,
      "step": 40885
    },
    {
      "epoch": 6.090259159964254,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.026337613110855182,
      "loss": 0.7877,
      "num_input_tokens_seen": 23727752,
      "step": 40890
    },
    {
      "epoch": 6.091003872505213,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02633633646943367,
      "loss": 0.8073,
      "num_input_tokens_seen": 23730600,
      "step": 40895
    },
    {
      "epoch": 6.091748585046172,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.026335059636496284,
      "loss": 0.8004,
      "num_input_tokens_seen": 23733544,
      "step": 40900
    },
    {
      "epoch": 6.092493297587131,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.026333782612064583,
      "loss": 0.8155,
      "num_input_tokens_seen": 23736584,
      "step": 40905
    },
    {
      "epoch": 6.093238010128091,
      "grad_norm": 0.013671875,
      "learning_rate": 0.026332505396160145,
      "loss": 0.8014,
      "num_input_tokens_seen": 23739592,
      "step": 40910
    },
    {
      "epoch": 6.0939827226690495,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02633122798880454,
      "loss": 0.7884,
      "num_input_tokens_seen": 23742408,
      "step": 40915
    },
    {
      "epoch": 6.094727435210009,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.026329950390019362,
      "loss": 0.8076,
      "num_input_tokens_seen": 23745256,
      "step": 40920
    },
    {
      "epoch": 6.095472147750968,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.026328672599826186,
      "loss": 0.7979,
      "num_input_tokens_seen": 23748168,
      "step": 40925
    },
    {
      "epoch": 6.0962168602919276,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.026327394618246597,
      "loss": 0.788,
      "num_input_tokens_seen": 23751112,
      "step": 40930
    },
    {
      "epoch": 6.096961572832886,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02632611644530219,
      "loss": 0.7847,
      "num_input_tokens_seen": 23753960,
      "step": 40935
    },
    {
      "epoch": 6.097706285373846,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.026324838081014563,
      "loss": 0.8126,
      "num_input_tokens_seen": 23756808,
      "step": 40940
    },
    {
      "epoch": 6.098450997914805,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.0263235595254053,
      "loss": 0.8036,
      "num_input_tokens_seen": 23759816,
      "step": 40945
    },
    {
      "epoch": 6.099195710455764,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.026322280778496007,
      "loss": 0.8048,
      "num_input_tokens_seen": 23762568,
      "step": 40950
    },
    {
      "epoch": 6.099940422996723,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.026321001840308292,
      "loss": 0.7913,
      "num_input_tokens_seen": 23765512,
      "step": 40955
    },
    {
      "epoch": 6.100685135537683,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.026319722710863754,
      "loss": 0.786,
      "num_input_tokens_seen": 23768296,
      "step": 40960
    },
    {
      "epoch": 6.1014298480786415,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.026318443390184007,
      "loss": 0.8154,
      "num_input_tokens_seen": 23771144,
      "step": 40965
    },
    {
      "epoch": 6.102174560619601,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.026317163878290663,
      "loss": 0.8091,
      "num_input_tokens_seen": 23774248,
      "step": 40970
    },
    {
      "epoch": 6.10291927316056,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.026315884175205338,
      "loss": 0.8,
      "num_input_tokens_seen": 23777384,
      "step": 40975
    },
    {
      "epoch": 6.10366398570152,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.026314604280949647,
      "loss": 0.7829,
      "num_input_tokens_seen": 23780456,
      "step": 40980
    },
    {
      "epoch": 6.104408698242478,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02631332419554522,
      "loss": 0.7886,
      "num_input_tokens_seen": 23783464,
      "step": 40985
    },
    {
      "epoch": 6.105153410783438,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.026312043919013678,
      "loss": 0.787,
      "num_input_tokens_seen": 23786344,
      "step": 40990
    },
    {
      "epoch": 6.105898123324397,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02631076345137665,
      "loss": 0.7942,
      "num_input_tokens_seen": 23789480,
      "step": 40995
    },
    {
      "epoch": 6.106642835865356,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.02630948279265577,
      "loss": 0.81,
      "num_input_tokens_seen": 23792840,
      "step": 41000
    },
    {
      "epoch": 6.107387548406315,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02630820194287267,
      "loss": 0.8163,
      "num_input_tokens_seen": 23795432,
      "step": 41005
    },
    {
      "epoch": 6.108132260947274,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.026306920902048996,
      "loss": 0.7814,
      "num_input_tokens_seen": 23798216,
      "step": 41010
    },
    {
      "epoch": 6.1088769734882336,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.026305639670206382,
      "loss": 0.7812,
      "num_input_tokens_seen": 23801064,
      "step": 41015
    },
    {
      "epoch": 6.109621686029192,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.026304358247366478,
      "loss": 0.814,
      "num_input_tokens_seen": 23803752,
      "step": 41020
    },
    {
      "epoch": 6.110366398570152,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.026303076633550932,
      "loss": 0.8029,
      "num_input_tokens_seen": 23806696,
      "step": 41025
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02630179482878139,
      "loss": 0.8431,
      "num_input_tokens_seen": 23809480,
      "step": 41030
    },
    {
      "epoch": 6.11185582365207,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.026300512833079517,
      "loss": 0.8024,
      "num_input_tokens_seen": 23811944,
      "step": 41035
    },
    {
      "epoch": 6.112600536193029,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02629923064646696,
      "loss": 0.7979,
      "num_input_tokens_seen": 23814856,
      "step": 41040
    },
    {
      "epoch": 6.113345248733989,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.026297948268965383,
      "loss": 0.7969,
      "num_input_tokens_seen": 23817672,
      "step": 41045
    },
    {
      "epoch": 6.1140899612749475,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02629666570059646,
      "loss": 0.7893,
      "num_input_tokens_seen": 23820584,
      "step": 41050
    },
    {
      "epoch": 6.114834673815907,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.02629538294138185,
      "loss": 0.8061,
      "num_input_tokens_seen": 23823496,
      "step": 41055
    },
    {
      "epoch": 6.115579386356866,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02629409999134322,
      "loss": 0.8023,
      "num_input_tokens_seen": 23826472,
      "step": 41060
    },
    {
      "epoch": 6.116324098897826,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.026292816850502252,
      "loss": 0.7999,
      "num_input_tokens_seen": 23829480,
      "step": 41065
    },
    {
      "epoch": 6.117068811438784,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02629153351888062,
      "loss": 0.791,
      "num_input_tokens_seen": 23831976,
      "step": 41070
    },
    {
      "epoch": 6.117813523979744,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02629024999650001,
      "loss": 0.8366,
      "num_input_tokens_seen": 23834824,
      "step": 41075
    },
    {
      "epoch": 6.118558236520703,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.026288966283382096,
      "loss": 0.7868,
      "num_input_tokens_seen": 23837960,
      "step": 41080
    },
    {
      "epoch": 6.119302949061662,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.026287682379548576,
      "loss": 0.7982,
      "num_input_tokens_seen": 23841000,
      "step": 41085
    },
    {
      "epoch": 6.120047661602621,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.026286398285021127,
      "loss": 0.7857,
      "num_input_tokens_seen": 23843720,
      "step": 41090
    },
    {
      "epoch": 6.120792374143581,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.026285113999821454,
      "loss": 0.8144,
      "num_input_tokens_seen": 23846888,
      "step": 41095
    },
    {
      "epoch": 6.1215370866845396,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026283829523971253,
      "loss": 0.7944,
      "num_input_tokens_seen": 23849640,
      "step": 41100
    },
    {
      "epoch": 6.122281799225499,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.026282544857492214,
      "loss": 0.7914,
      "num_input_tokens_seen": 23852808,
      "step": 41105
    },
    {
      "epoch": 6.123026511766458,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.026281260000406047,
      "loss": 0.7787,
      "num_input_tokens_seen": 23855784,
      "step": 41110
    },
    {
      "epoch": 6.123771224307418,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.026279974952734465,
      "loss": 0.8246,
      "num_input_tokens_seen": 23858568,
      "step": 41115
    },
    {
      "epoch": 6.124515936848376,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.026278689714499166,
      "loss": 0.7989,
      "num_input_tokens_seen": 23861224,
      "step": 41120
    },
    {
      "epoch": 6.125260649389336,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.026277404285721867,
      "loss": 0.8099,
      "num_input_tokens_seen": 23864264,
      "step": 41125
    },
    {
      "epoch": 6.126005361930295,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02627611866642428,
      "loss": 0.8114,
      "num_input_tokens_seen": 23866952,
      "step": 41130
    },
    {
      "epoch": 6.126750074471254,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02627483285662814,
      "loss": 0.7979,
      "num_input_tokens_seen": 23870056,
      "step": 41135
    },
    {
      "epoch": 6.127494787012213,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02627354685635515,
      "loss": 0.8098,
      "num_input_tokens_seen": 23873032,
      "step": 41140
    },
    {
      "epoch": 6.128239499553173,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.026272260665627045,
      "loss": 0.7934,
      "num_input_tokens_seen": 23875880,
      "step": 41145
    },
    {
      "epoch": 6.128984212094132,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.026270974284465554,
      "loss": 0.8088,
      "num_input_tokens_seen": 23878600,
      "step": 41150
    },
    {
      "epoch": 6.129728924635091,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.02626968771289241,
      "loss": 0.81,
      "num_input_tokens_seen": 23881544,
      "step": 41155
    },
    {
      "epoch": 6.13047363717605,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.026268400950929343,
      "loss": 0.8108,
      "num_input_tokens_seen": 23884456,
      "step": 41160
    },
    {
      "epoch": 6.13121834971701,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.02626711399859809,
      "loss": 0.7999,
      "num_input_tokens_seen": 23887464,
      "step": 41165
    },
    {
      "epoch": 6.131963062257968,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.026265826855920403,
      "loss": 0.797,
      "num_input_tokens_seen": 23890280,
      "step": 41170
    },
    {
      "epoch": 6.132707774798928,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02626453952291802,
      "loss": 0.793,
      "num_input_tokens_seen": 23893768,
      "step": 41175
    },
    {
      "epoch": 6.133452487339887,
      "grad_norm": 0.02734375,
      "learning_rate": 0.026263251999612696,
      "loss": 0.7975,
      "num_input_tokens_seen": 23896616,
      "step": 41180
    },
    {
      "epoch": 6.134197199880846,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02626196428602617,
      "loss": 0.7967,
      "num_input_tokens_seen": 23899720,
      "step": 41185
    },
    {
      "epoch": 6.134941912421805,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.026260676382180204,
      "loss": 0.7757,
      "num_input_tokens_seen": 23902856,
      "step": 41190
    },
    {
      "epoch": 6.135686624962764,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.02625938828809656,
      "loss": 0.7943,
      "num_input_tokens_seen": 23905736,
      "step": 41195
    },
    {
      "epoch": 6.136431337503724,
      "grad_norm": 0.025390625,
      "learning_rate": 0.026258100003796987,
      "loss": 0.8121,
      "num_input_tokens_seen": 23908680,
      "step": 41200
    },
    {
      "epoch": 6.137176050044682,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.026256811529303262,
      "loss": 0.7824,
      "num_input_tokens_seen": 23911368,
      "step": 41205
    },
    {
      "epoch": 6.137920762585642,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.026255522864637146,
      "loss": 0.784,
      "num_input_tokens_seen": 23914312,
      "step": 41210
    },
    {
      "epoch": 6.138665475126601,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.026254234009820413,
      "loss": 0.7605,
      "num_input_tokens_seen": 23917128,
      "step": 41215
    },
    {
      "epoch": 6.13941018766756,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.026252944964874837,
      "loss": 0.7658,
      "num_input_tokens_seen": 23919624,
      "step": 41220
    },
    {
      "epoch": 6.140154900208519,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.026251655729822184,
      "loss": 0.802,
      "num_input_tokens_seen": 23922504,
      "step": 41225
    },
    {
      "epoch": 6.140899612749479,
      "grad_norm": 0.025390625,
      "learning_rate": 0.026250366304684244,
      "loss": 0.8235,
      "num_input_tokens_seen": 23925256,
      "step": 41230
    },
    {
      "epoch": 6.141644325290438,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.026249076689482807,
      "loss": 0.802,
      "num_input_tokens_seen": 23928072,
      "step": 41235
    },
    {
      "epoch": 6.142389037831397,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.026247786884239645,
      "loss": 0.79,
      "num_input_tokens_seen": 23931144,
      "step": 41240
    },
    {
      "epoch": 6.143133750372356,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.026246496888976564,
      "loss": 0.8375,
      "num_input_tokens_seen": 23933736,
      "step": 41245
    },
    {
      "epoch": 6.143878462913316,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.026245206703715343,
      "loss": 0.8046,
      "num_input_tokens_seen": 23936712,
      "step": 41250
    },
    {
      "epoch": 6.144623175454274,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.026243916328477782,
      "loss": 0.7956,
      "num_input_tokens_seen": 23939784,
      "step": 41255
    },
    {
      "epoch": 6.145367887995234,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.026242625763285687,
      "loss": 0.8074,
      "num_input_tokens_seen": 23942664,
      "step": 41260
    },
    {
      "epoch": 6.146112600536193,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02624133500816085,
      "loss": 0.7898,
      "num_input_tokens_seen": 23945608,
      "step": 41265
    },
    {
      "epoch": 6.146857313077152,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02624004406312509,
      "loss": 0.8363,
      "num_input_tokens_seen": 23948520,
      "step": 41270
    },
    {
      "epoch": 6.147602025618111,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02623875292820021,
      "loss": 0.8222,
      "num_input_tokens_seen": 23951912,
      "step": 41275
    },
    {
      "epoch": 6.148346738159071,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.026237461603408015,
      "loss": 0.7835,
      "num_input_tokens_seen": 23954696,
      "step": 41280
    },
    {
      "epoch": 6.14909145070003,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.026236170088770328,
      "loss": 0.8035,
      "num_input_tokens_seen": 23957448,
      "step": 41285
    },
    {
      "epoch": 6.149836163240989,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02623487838430897,
      "loss": 0.7963,
      "num_input_tokens_seen": 23960360,
      "step": 41290
    },
    {
      "epoch": 6.150580875781948,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.026233586490045764,
      "loss": 0.8198,
      "num_input_tokens_seen": 23963176,
      "step": 41295
    },
    {
      "epoch": 6.151325588322908,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.026232294406002526,
      "loss": 0.8223,
      "num_input_tokens_seen": 23965896,
      "step": 41300
    },
    {
      "epoch": 6.152070300863866,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02623100213220109,
      "loss": 0.7915,
      "num_input_tokens_seen": 23968840,
      "step": 41305
    },
    {
      "epoch": 6.152815013404826,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02622970966866329,
      "loss": 0.8227,
      "num_input_tokens_seen": 23971784,
      "step": 41310
    },
    {
      "epoch": 6.153559725945785,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02622841701541096,
      "loss": 0.7912,
      "num_input_tokens_seen": 23974792,
      "step": 41315
    },
    {
      "epoch": 6.1543044384867445,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02622712417246594,
      "loss": 0.8178,
      "num_input_tokens_seen": 23977608,
      "step": 41320
    },
    {
      "epoch": 6.155049151027703,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.02622583113985006,
      "loss": 0.8061,
      "num_input_tokens_seen": 23980328,
      "step": 41325
    },
    {
      "epoch": 6.155793863568663,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.026224537917585174,
      "loss": 0.7892,
      "num_input_tokens_seen": 23982984,
      "step": 41330
    },
    {
      "epoch": 6.156538576109622,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.026223244505693126,
      "loss": 0.7902,
      "num_input_tokens_seen": 23986024,
      "step": 41335
    },
    {
      "epoch": 6.157283288650581,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.026221950904195775,
      "loss": 0.8208,
      "num_input_tokens_seen": 23988808,
      "step": 41340
    },
    {
      "epoch": 6.15802800119154,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.026220657113114963,
      "loss": 0.7858,
      "num_input_tokens_seen": 23991528,
      "step": 41345
    },
    {
      "epoch": 6.1587727137325,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.026219363132472556,
      "loss": 0.8018,
      "num_input_tokens_seen": 23994440,
      "step": 41350
    },
    {
      "epoch": 6.159517426273458,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.026218068962290415,
      "loss": 0.7982,
      "num_input_tokens_seen": 23997320,
      "step": 41355
    },
    {
      "epoch": 6.160262138814417,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.026216774602590392,
      "loss": 0.8069,
      "num_input_tokens_seen": 24000296,
      "step": 41360
    },
    {
      "epoch": 6.161006851355377,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02621548005339437,
      "loss": 0.8281,
      "num_input_tokens_seen": 24003144,
      "step": 41365
    },
    {
      "epoch": 6.161751563896336,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02621418531472421,
      "loss": 0.8107,
      "num_input_tokens_seen": 24006376,
      "step": 41370
    },
    {
      "epoch": 6.162496276437295,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.026212890386601784,
      "loss": 0.7809,
      "num_input_tokens_seen": 24009032,
      "step": 41375
    },
    {
      "epoch": 6.163240988978254,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02621159526904897,
      "loss": 0.7928,
      "num_input_tokens_seen": 24011944,
      "step": 41380
    },
    {
      "epoch": 6.163985701519214,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.026210299962087655,
      "loss": 0.7973,
      "num_input_tokens_seen": 24014760,
      "step": 41385
    },
    {
      "epoch": 6.164730414060172,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.02620900446573971,
      "loss": 0.8016,
      "num_input_tokens_seen": 24017864,
      "step": 41390
    },
    {
      "epoch": 6.165475126601132,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.026207708780027032,
      "loss": 0.8193,
      "num_input_tokens_seen": 24020744,
      "step": 41395
    },
    {
      "epoch": 6.166219839142091,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.026206412904971506,
      "loss": 0.8006,
      "num_input_tokens_seen": 24023400,
      "step": 41400
    },
    {
      "epoch": 6.1669645516830505,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.026205116840595018,
      "loss": 0.8221,
      "num_input_tokens_seen": 24026312,
      "step": 41405
    },
    {
      "epoch": 6.167709264224009,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.026203820586919467,
      "loss": 0.8132,
      "num_input_tokens_seen": 24029160,
      "step": 41410
    },
    {
      "epoch": 6.168453976764969,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02620252414396676,
      "loss": 0.8061,
      "num_input_tokens_seen": 24032392,
      "step": 41415
    },
    {
      "epoch": 6.169198689305928,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.026201227511758792,
      "loss": 0.8137,
      "num_input_tokens_seen": 24035784,
      "step": 41420
    },
    {
      "epoch": 6.169943401846887,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02619993069031747,
      "loss": 0.8083,
      "num_input_tokens_seen": 24038696,
      "step": 41425
    },
    {
      "epoch": 6.170688114387846,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.026198633679664702,
      "loss": 0.8107,
      "num_input_tokens_seen": 24041672,
      "step": 41430
    },
    {
      "epoch": 6.171432826928806,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.026197336479822402,
      "loss": 0.8211,
      "num_input_tokens_seen": 24044456,
      "step": 41435
    },
    {
      "epoch": 6.172177539469764,
      "grad_norm": 0.01116943359375,
      "learning_rate": 0.02619603909081248,
      "loss": 0.8009,
      "num_input_tokens_seen": 24047240,
      "step": 41440
    },
    {
      "epoch": 6.172922252010724,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.026194741512656855,
      "loss": 0.7924,
      "num_input_tokens_seen": 24050248,
      "step": 41445
    },
    {
      "epoch": 6.173666964551683,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02619344374537745,
      "loss": 0.7927,
      "num_input_tokens_seen": 24053128,
      "step": 41450
    },
    {
      "epoch": 6.1744116770926425,
      "grad_norm": 0.05078125,
      "learning_rate": 0.026192145788996195,
      "loss": 0.8469,
      "num_input_tokens_seen": 24055976,
      "step": 41455
    },
    {
      "epoch": 6.175156389633601,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.02619084764353501,
      "loss": 0.8059,
      "num_input_tokens_seen": 24059176,
      "step": 41460
    },
    {
      "epoch": 6.175901102174561,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026189549309015823,
      "loss": 0.8202,
      "num_input_tokens_seen": 24062056,
      "step": 41465
    },
    {
      "epoch": 6.17664581471552,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.026188250785460578,
      "loss": 0.8253,
      "num_input_tokens_seen": 24064616,
      "step": 41470
    },
    {
      "epoch": 6.177390527256479,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.026186952072891203,
      "loss": 0.8115,
      "num_input_tokens_seen": 24067848,
      "step": 41475
    },
    {
      "epoch": 6.178135239797438,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02618565317132964,
      "loss": 0.8021,
      "num_input_tokens_seen": 24070536,
      "step": 41480
    },
    {
      "epoch": 6.178879952338398,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.026184354080797844,
      "loss": 0.8076,
      "num_input_tokens_seen": 24073320,
      "step": 41485
    },
    {
      "epoch": 6.1796246648793565,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02618305480131775,
      "loss": 0.792,
      "num_input_tokens_seen": 24076232,
      "step": 41490
    },
    {
      "epoch": 6.180369377420316,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02618175533291131,
      "loss": 0.7946,
      "num_input_tokens_seen": 24079112,
      "step": 41495
    },
    {
      "epoch": 6.181114089961275,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02618045567560048,
      "loss": 0.8002,
      "num_input_tokens_seen": 24082120,
      "step": 41500
    },
    {
      "epoch": 6.1818588025022345,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.026179155829407212,
      "loss": 0.8028,
      "num_input_tokens_seen": 24084936,
      "step": 41505
    },
    {
      "epoch": 6.182603515043193,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.02617785579435347,
      "loss": 0.8036,
      "num_input_tokens_seen": 24087944,
      "step": 41510
    },
    {
      "epoch": 6.183348227584153,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.026176555570461217,
      "loss": 0.8086,
      "num_input_tokens_seen": 24090632,
      "step": 41515
    },
    {
      "epoch": 6.184092940125112,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.026175255157752413,
      "loss": 0.8122,
      "num_input_tokens_seen": 24093736,
      "step": 41520
    },
    {
      "epoch": 6.18483765266607,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.026173954556249032,
      "loss": 0.7961,
      "num_input_tokens_seen": 24096744,
      "step": 41525
    },
    {
      "epoch": 6.18558236520703,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02617265376597305,
      "loss": 0.8081,
      "num_input_tokens_seen": 24099624,
      "step": 41530
    },
    {
      "epoch": 6.18632707774799,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02617135278694643,
      "loss": 0.803,
      "num_input_tokens_seen": 24102600,
      "step": 41535
    },
    {
      "epoch": 6.1870717902889485,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.026170051619191163,
      "loss": 0.7954,
      "num_input_tokens_seen": 24105640,
      "step": 41540
    },
    {
      "epoch": 6.187816502829907,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02616875026272923,
      "loss": 0.8039,
      "num_input_tokens_seen": 24108392,
      "step": 41545
    },
    {
      "epoch": 6.188561215370867,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.026167448717582604,
      "loss": 0.804,
      "num_input_tokens_seen": 24111304,
      "step": 41550
    },
    {
      "epoch": 6.189305927911826,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.02616614698377329,
      "loss": 0.81,
      "num_input_tokens_seen": 24114280,
      "step": 41555
    },
    {
      "epoch": 6.190050640452785,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.02616484506132327,
      "loss": 0.8119,
      "num_input_tokens_seen": 24117448,
      "step": 41560
    },
    {
      "epoch": 6.190795352993744,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02616354295025454,
      "loss": 0.8288,
      "num_input_tokens_seen": 24120040,
      "step": 41565
    },
    {
      "epoch": 6.191540065534704,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.026162240650589096,
      "loss": 0.7909,
      "num_input_tokens_seen": 24123336,
      "step": 41570
    },
    {
      "epoch": 6.1922847780756625,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02616093816234894,
      "loss": 0.8119,
      "num_input_tokens_seen": 24125896,
      "step": 41575
    },
    {
      "epoch": 6.193029490616622,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02615963548555608,
      "loss": 0.7957,
      "num_input_tokens_seen": 24129032,
      "step": 41580
    },
    {
      "epoch": 6.193774203157581,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.02615833262023252,
      "loss": 0.794,
      "num_input_tokens_seen": 24132072,
      "step": 41585
    },
    {
      "epoch": 6.1945189156985405,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.026157029566400272,
      "loss": 0.8035,
      "num_input_tokens_seen": 24134792,
      "step": 41590
    },
    {
      "epoch": 6.195263628239499,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02615572632408135,
      "loss": 0.8035,
      "num_input_tokens_seen": 24137416,
      "step": 41595
    },
    {
      "epoch": 6.196008340780459,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.026154422893297766,
      "loss": 0.7977,
      "num_input_tokens_seen": 24140264,
      "step": 41600
    },
    {
      "epoch": 6.196753053321418,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.026153119274071546,
      "loss": 0.8064,
      "num_input_tokens_seen": 24143624,
      "step": 41605
    },
    {
      "epoch": 6.197497765862377,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02615181546642471,
      "loss": 0.8113,
      "num_input_tokens_seen": 24146696,
      "step": 41610
    },
    {
      "epoch": 6.198242478403336,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.026150511470379285,
      "loss": 0.8094,
      "num_input_tokens_seen": 24149448,
      "step": 41615
    },
    {
      "epoch": 6.198987190944296,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02614920728595731,
      "loss": 0.8063,
      "num_input_tokens_seen": 24152232,
      "step": 41620
    },
    {
      "epoch": 6.1997319034852545,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.0261479029131808,
      "loss": 0.8035,
      "num_input_tokens_seen": 24154984,
      "step": 41625
    },
    {
      "epoch": 6.200476616026214,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026146598352071802,
      "loss": 0.7954,
      "num_input_tokens_seen": 24157800,
      "step": 41630
    },
    {
      "epoch": 6.201221328567173,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.026145293602652355,
      "loss": 0.803,
      "num_input_tokens_seen": 24160552,
      "step": 41635
    },
    {
      "epoch": 6.2019660411081325,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.026143988664944502,
      "loss": 0.8042,
      "num_input_tokens_seen": 24163304,
      "step": 41640
    },
    {
      "epoch": 6.202710753649091,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.026142683538970285,
      "loss": 0.7987,
      "num_input_tokens_seen": 24166600,
      "step": 41645
    },
    {
      "epoch": 6.203455466190051,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.026141378224751755,
      "loss": 0.7954,
      "num_input_tokens_seen": 24169704,
      "step": 41650
    },
    {
      "epoch": 6.20420017873101,
      "grad_norm": 0.017578125,
      "learning_rate": 0.026140072722310965,
      "loss": 0.8183,
      "num_input_tokens_seen": 24172392,
      "step": 41655
    },
    {
      "epoch": 6.204944891271969,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.026138767031669966,
      "loss": 0.8042,
      "num_input_tokens_seen": 24175656,
      "step": 41660
    },
    {
      "epoch": 6.205689603812928,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02613746115285082,
      "loss": 0.8056,
      "num_input_tokens_seen": 24178760,
      "step": 41665
    },
    {
      "epoch": 6.206434316353888,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.026136155085875586,
      "loss": 0.7913,
      "num_input_tokens_seen": 24181448,
      "step": 41670
    },
    {
      "epoch": 6.2071790288948465,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.026134848830766333,
      "loss": 0.8012,
      "num_input_tokens_seen": 24184456,
      "step": 41675
    },
    {
      "epoch": 6.207923741435806,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.026133542387545124,
      "loss": 0.8171,
      "num_input_tokens_seen": 24187400,
      "step": 41680
    },
    {
      "epoch": 6.208668453976765,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.026132235756234034,
      "loss": 0.8071,
      "num_input_tokens_seen": 24190440,
      "step": 41685
    },
    {
      "epoch": 6.209413166517725,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02613092893685513,
      "loss": 0.8099,
      "num_input_tokens_seen": 24193160,
      "step": 41690
    },
    {
      "epoch": 6.210157879058683,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.0261296219294305,
      "loss": 0.8112,
      "num_input_tokens_seen": 24196008,
      "step": 41695
    },
    {
      "epoch": 6.210902591599643,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02612831473398222,
      "loss": 0.805,
      "num_input_tokens_seen": 24198792,
      "step": 41700
    },
    {
      "epoch": 6.211647304140602,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.026127007350532368,
      "loss": 0.8068,
      "num_input_tokens_seen": 24201832,
      "step": 41705
    },
    {
      "epoch": 6.2123920166815605,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.026125699779103038,
      "loss": 0.8049,
      "num_input_tokens_seen": 24204648,
      "step": 41710
    },
    {
      "epoch": 6.21313672922252,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02612439201971632,
      "loss": 0.8031,
      "num_input_tokens_seen": 24207656,
      "step": 41715
    },
    {
      "epoch": 6.213881441763479,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.026123084072394304,
      "loss": 0.803,
      "num_input_tokens_seen": 24210440,
      "step": 41720
    },
    {
      "epoch": 6.2146261543044385,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.026121775937159086,
      "loss": 0.805,
      "num_input_tokens_seen": 24213512,
      "step": 41725
    },
    {
      "epoch": 6.215370866845397,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.026120467614032767,
      "loss": 0.8019,
      "num_input_tokens_seen": 24216584,
      "step": 41730
    },
    {
      "epoch": 6.216115579386357,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.026119159103037453,
      "loss": 0.8106,
      "num_input_tokens_seen": 24219560,
      "step": 41735
    },
    {
      "epoch": 6.216860291927316,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.026117850404195245,
      "loss": 0.814,
      "num_input_tokens_seen": 24222216,
      "step": 41740
    },
    {
      "epoch": 6.217605004468275,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.026116541517528253,
      "loss": 0.802,
      "num_input_tokens_seen": 24225064,
      "step": 41745
    },
    {
      "epoch": 6.218349717009234,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02611523244305859,
      "loss": 0.7917,
      "num_input_tokens_seen": 24228232,
      "step": 41750
    },
    {
      "epoch": 6.219094429550194,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.026113923180808372,
      "loss": 0.8183,
      "num_input_tokens_seen": 24231208,
      "step": 41755
    },
    {
      "epoch": 6.2198391420911525,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02611261373079972,
      "loss": 0.8075,
      "num_input_tokens_seen": 24233800,
      "step": 41760
    },
    {
      "epoch": 6.220583854632112,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02611130409305475,
      "loss": 0.8031,
      "num_input_tokens_seen": 24236840,
      "step": 41765
    },
    {
      "epoch": 6.221328567173071,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.026109994267595596,
      "loss": 0.7975,
      "num_input_tokens_seen": 24239688,
      "step": 41770
    },
    {
      "epoch": 6.222073279714031,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.026108684254444372,
      "loss": 0.8069,
      "num_input_tokens_seen": 24242696,
      "step": 41775
    },
    {
      "epoch": 6.222817992254989,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.026107374053623227,
      "loss": 0.7963,
      "num_input_tokens_seen": 24245576,
      "step": 41780
    },
    {
      "epoch": 6.223562704795949,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.026106063665154285,
      "loss": 0.7923,
      "num_input_tokens_seen": 24248712,
      "step": 41785
    },
    {
      "epoch": 6.224307417336908,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.026104753089059678,
      "loss": 0.816,
      "num_input_tokens_seen": 24251624,
      "step": 41790
    },
    {
      "epoch": 6.225052129877867,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02610344232536156,
      "loss": 0.7859,
      "num_input_tokens_seen": 24254600,
      "step": 41795
    },
    {
      "epoch": 6.225796842418826,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.026102131374082068,
      "loss": 0.8135,
      "num_input_tokens_seen": 24257512,
      "step": 41800
    },
    {
      "epoch": 6.226541554959786,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02610082023524335,
      "loss": 0.8173,
      "num_input_tokens_seen": 24260296,
      "step": 41805
    },
    {
      "epoch": 6.2272862675007445,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.026099508908867555,
      "loss": 0.8007,
      "num_input_tokens_seen": 24263080,
      "step": 41810
    },
    {
      "epoch": 6.228030980041704,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.02609819739497684,
      "loss": 0.8069,
      "num_input_tokens_seen": 24265928,
      "step": 41815
    },
    {
      "epoch": 6.228775692582663,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.026096885693593357,
      "loss": 0.8044,
      "num_input_tokens_seen": 24269128,
      "step": 41820
    },
    {
      "epoch": 6.229520405123623,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.02609557380473927,
      "loss": 0.8118,
      "num_input_tokens_seen": 24271912,
      "step": 41825
    },
    {
      "epoch": 6.230265117664581,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.02609426172843674,
      "loss": 0.8011,
      "num_input_tokens_seen": 24275656,
      "step": 41830
    },
    {
      "epoch": 6.231009830205541,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.026092949464707936,
      "loss": 0.807,
      "num_input_tokens_seen": 24278248,
      "step": 41835
    },
    {
      "epoch": 6.2317545427465,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.026091637013575024,
      "loss": 0.8,
      "num_input_tokens_seen": 24281288,
      "step": 41840
    },
    {
      "epoch": 6.232499255287459,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.026090324375060175,
      "loss": 0.7979,
      "num_input_tokens_seen": 24284040,
      "step": 41845
    },
    {
      "epoch": 6.233243967828418,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02608901154918557,
      "loss": 0.8075,
      "num_input_tokens_seen": 24287144,
      "step": 41850
    },
    {
      "epoch": 6.233988680369378,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.026087698535973387,
      "loss": 0.8086,
      "num_input_tokens_seen": 24290120,
      "step": 41855
    },
    {
      "epoch": 6.234733392910337,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.026086385335445807,
      "loss": 0.8069,
      "num_input_tokens_seen": 24293096,
      "step": 41860
    },
    {
      "epoch": 6.235478105451296,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.026085071947625008,
      "loss": 0.8083,
      "num_input_tokens_seen": 24295944,
      "step": 41865
    },
    {
      "epoch": 6.236222817992255,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02608375837253319,
      "loss": 0.7958,
      "num_input_tokens_seen": 24298952,
      "step": 41870
    },
    {
      "epoch": 6.236967530533214,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02608244461019254,
      "loss": 0.8112,
      "num_input_tokens_seen": 24301768,
      "step": 41875
    },
    {
      "epoch": 6.237712243074173,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02608113066062525,
      "loss": 0.8102,
      "num_input_tokens_seen": 24304776,
      "step": 41880
    },
    {
      "epoch": 6.238456955615132,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02607981652385352,
      "loss": 0.7957,
      "num_input_tokens_seen": 24307784,
      "step": 41885
    },
    {
      "epoch": 6.239201668156092,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02607850219989955,
      "loss": 0.7996,
      "num_input_tokens_seen": 24310856,
      "step": 41890
    },
    {
      "epoch": 6.2399463806970505,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.026077187688785544,
      "loss": 0.7979,
      "num_input_tokens_seen": 24314088,
      "step": 41895
    },
    {
      "epoch": 6.24069109323801,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.026075872990533714,
      "loss": 0.8006,
      "num_input_tokens_seen": 24317192,
      "step": 41900
    },
    {
      "epoch": 6.241435805778969,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.026074558105166262,
      "loss": 0.8026,
      "num_input_tokens_seen": 24319848,
      "step": 41905
    },
    {
      "epoch": 6.242180518319929,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02607324303270541,
      "loss": 0.8011,
      "num_input_tokens_seen": 24322824,
      "step": 41910
    },
    {
      "epoch": 6.242925230860887,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02607192777317337,
      "loss": 0.8037,
      "num_input_tokens_seen": 24325736,
      "step": 41915
    },
    {
      "epoch": 6.243669943401847,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.026070612326592368,
      "loss": 0.7955,
      "num_input_tokens_seen": 24328360,
      "step": 41920
    },
    {
      "epoch": 6.244414655942806,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.026069296692984616,
      "loss": 0.819,
      "num_input_tokens_seen": 24331848,
      "step": 41925
    },
    {
      "epoch": 6.245159368483765,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02606798087237235,
      "loss": 0.7794,
      "num_input_tokens_seen": 24334664,
      "step": 41930
    },
    {
      "epoch": 6.245904081024724,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02606666486477779,
      "loss": 0.7984,
      "num_input_tokens_seen": 24337640,
      "step": 41935
    },
    {
      "epoch": 6.246648793565684,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.026065348670223188,
      "loss": 0.8048,
      "num_input_tokens_seen": 24340712,
      "step": 41940
    },
    {
      "epoch": 6.247393506106643,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.026064032288730755,
      "loss": 0.8164,
      "num_input_tokens_seen": 24343560,
      "step": 41945
    },
    {
      "epoch": 6.248138218647602,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02606271572032275,
      "loss": 0.8009,
      "num_input_tokens_seen": 24346568,
      "step": 41950
    },
    {
      "epoch": 6.248882931188561,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.026061398965021396,
      "loss": 0.8083,
      "num_input_tokens_seen": 24349256,
      "step": 41955
    },
    {
      "epoch": 6.249627643729521,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.026060082022848957,
      "loss": 0.8313,
      "num_input_tokens_seen": 24352040,
      "step": 41960
    },
    {
      "epoch": 6.250372356270479,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02605876489382767,
      "loss": 0.8106,
      "num_input_tokens_seen": 24355176,
      "step": 41965
    },
    {
      "epoch": 6.251117068811439,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.02605744757797979,
      "loss": 0.8198,
      "num_input_tokens_seen": 24358056,
      "step": 41970
    },
    {
      "epoch": 6.251861781352398,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02605613007532757,
      "loss": 0.8085,
      "num_input_tokens_seen": 24361192,
      "step": 41975
    },
    {
      "epoch": 6.252606493893357,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.026054812385893272,
      "loss": 0.8168,
      "num_input_tokens_seen": 24364168,
      "step": 41980
    },
    {
      "epoch": 6.253351206434316,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.026053494509699154,
      "loss": 0.8042,
      "num_input_tokens_seen": 24367144,
      "step": 41985
    },
    {
      "epoch": 6.254095918975276,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.026052176446767477,
      "loss": 0.7915,
      "num_input_tokens_seen": 24369992,
      "step": 41990
    },
    {
      "epoch": 6.254840631516235,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.026050858197120514,
      "loss": 0.7916,
      "num_input_tokens_seen": 24373480,
      "step": 41995
    },
    {
      "epoch": 6.255585344057194,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02604953976078054,
      "loss": 0.7967,
      "num_input_tokens_seen": 24376520,
      "step": 42000
    },
    {
      "epoch": 6.256330056598153,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.026048221137769813,
      "loss": 0.8051,
      "num_input_tokens_seen": 24379752,
      "step": 42005
    },
    {
      "epoch": 6.257074769139113,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.026046902328110622,
      "loss": 0.7994,
      "num_input_tokens_seen": 24382728,
      "step": 42010
    },
    {
      "epoch": 6.257819481680071,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.026045583331825245,
      "loss": 0.801,
      "num_input_tokens_seen": 24385704,
      "step": 42015
    },
    {
      "epoch": 6.258564194221031,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.026044264148935962,
      "loss": 0.7875,
      "num_input_tokens_seen": 24388232,
      "step": 42020
    },
    {
      "epoch": 6.25930890676199,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.026042944779465064,
      "loss": 0.7961,
      "num_input_tokens_seen": 24391112,
      "step": 42025
    },
    {
      "epoch": 6.2600536193029495,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.026041625223434832,
      "loss": 0.7959,
      "num_input_tokens_seen": 24393928,
      "step": 42030
    },
    {
      "epoch": 6.260798331843908,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.026040305480867568,
      "loss": 0.7959,
      "num_input_tokens_seen": 24397000,
      "step": 42035
    },
    {
      "epoch": 6.261543044384867,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.026038985551785566,
      "loss": 0.788,
      "num_input_tokens_seen": 24399880,
      "step": 42040
    },
    {
      "epoch": 6.262287756925827,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.026037665436211117,
      "loss": 0.8134,
      "num_input_tokens_seen": 24402664,
      "step": 42045
    },
    {
      "epoch": 6.263032469466786,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.026036345134166528,
      "loss": 0.8099,
      "num_input_tokens_seen": 24405640,
      "step": 42050
    },
    {
      "epoch": 6.263777182007745,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.026035024645674112,
      "loss": 0.8047,
      "num_input_tokens_seen": 24409128,
      "step": 42055
    },
    {
      "epoch": 6.264521894548704,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.026033703970756165,
      "loss": 0.8049,
      "num_input_tokens_seen": 24411816,
      "step": 42060
    },
    {
      "epoch": 6.265266607089663,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.026032383109435005,
      "loss": 0.7939,
      "num_input_tokens_seen": 24414536,
      "step": 42065
    },
    {
      "epoch": 6.266011319630622,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.026031062061732942,
      "loss": 0.8128,
      "num_input_tokens_seen": 24417192,
      "step": 42070
    },
    {
      "epoch": 6.266756032171582,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.0260297408276723,
      "loss": 0.808,
      "num_input_tokens_seen": 24419816,
      "step": 42075
    },
    {
      "epoch": 6.267500744712541,
      "grad_norm": 0.017578125,
      "learning_rate": 0.026028419407275398,
      "loss": 0.7962,
      "num_input_tokens_seen": 24422728,
      "step": 42080
    },
    {
      "epoch": 6.2682454572535,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.026027097800564558,
      "loss": 0.7897,
      "num_input_tokens_seen": 24425448,
      "step": 42085
    },
    {
      "epoch": 6.268990169794459,
      "grad_norm": 0.033203125,
      "learning_rate": 0.026025776007562108,
      "loss": 0.8239,
      "num_input_tokens_seen": 24428488,
      "step": 42090
    },
    {
      "epoch": 6.269734882335419,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02602445402829038,
      "loss": 0.7886,
      "num_input_tokens_seen": 24431720,
      "step": 42095
    },
    {
      "epoch": 6.270479594876377,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.026023131862771703,
      "loss": 0.8096,
      "num_input_tokens_seen": 24434440,
      "step": 42100
    },
    {
      "epoch": 6.271224307417337,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.026021809511028417,
      "loss": 0.7938,
      "num_input_tokens_seen": 24437352,
      "step": 42105
    },
    {
      "epoch": 6.271969019958296,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.026020486973082863,
      "loss": 0.8102,
      "num_input_tokens_seen": 24440136,
      "step": 42110
    },
    {
      "epoch": 6.2727137324992555,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02601916424895738,
      "loss": 0.806,
      "num_input_tokens_seen": 24443144,
      "step": 42115
    },
    {
      "epoch": 6.273458445040214,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02601784133867432,
      "loss": 0.7971,
      "num_input_tokens_seen": 24445896,
      "step": 42120
    },
    {
      "epoch": 6.274203157581174,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.026016518242256023,
      "loss": 0.7957,
      "num_input_tokens_seen": 24449160,
      "step": 42125
    },
    {
      "epoch": 6.274947870122133,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.026015194959724854,
      "loss": 0.8054,
      "num_input_tokens_seen": 24452168,
      "step": 42130
    },
    {
      "epoch": 6.275692582663092,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.026013871491103156,
      "loss": 0.7977,
      "num_input_tokens_seen": 24455048,
      "step": 42135
    },
    {
      "epoch": 6.276437295204051,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.026012547836413297,
      "loss": 0.8195,
      "num_input_tokens_seen": 24457800,
      "step": 42140
    },
    {
      "epoch": 6.277182007745011,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.02601122399567763,
      "loss": 0.8062,
      "num_input_tokens_seen": 24460520,
      "step": 42145
    },
    {
      "epoch": 6.277926720285969,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.026009899968918528,
      "loss": 0.7955,
      "num_input_tokens_seen": 24463208,
      "step": 42150
    },
    {
      "epoch": 6.278671432826929,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.026008575756158354,
      "loss": 0.8198,
      "num_input_tokens_seen": 24466184,
      "step": 42155
    },
    {
      "epoch": 6.279416145367888,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.026007251357419485,
      "loss": 0.8194,
      "num_input_tokens_seen": 24469256,
      "step": 42160
    },
    {
      "epoch": 6.2801608579088475,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02600592677272429,
      "loss": 0.8208,
      "num_input_tokens_seen": 24472104,
      "step": 42165
    },
    {
      "epoch": 6.280905570449806,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.026004602002095145,
      "loss": 0.7768,
      "num_input_tokens_seen": 24474728,
      "step": 42170
    },
    {
      "epoch": 6.281650282990766,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.026003277045554437,
      "loss": 0.7992,
      "num_input_tokens_seen": 24477480,
      "step": 42175
    },
    {
      "epoch": 6.282394995531725,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02600195190312455,
      "loss": 0.7891,
      "num_input_tokens_seen": 24480424,
      "step": 42180
    },
    {
      "epoch": 6.283139708072684,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02600062657482786,
      "loss": 0.8211,
      "num_input_tokens_seen": 24483560,
      "step": 42185
    },
    {
      "epoch": 6.283884420613643,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.025999301060686767,
      "loss": 0.8263,
      "num_input_tokens_seen": 24486504,
      "step": 42190
    },
    {
      "epoch": 6.284629133154603,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.02599797536072367,
      "loss": 0.8033,
      "num_input_tokens_seen": 24489544,
      "step": 42195
    },
    {
      "epoch": 6.2853738456955615,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.025996649474960946,
      "loss": 0.806,
      "num_input_tokens_seen": 24492456,
      "step": 42200
    },
    {
      "epoch": 6.286118558236521,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.02599532340342101,
      "loss": 0.8144,
      "num_input_tokens_seen": 24495368,
      "step": 42205
    },
    {
      "epoch": 6.28686327077748,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.025993997146126258,
      "loss": 0.7992,
      "num_input_tokens_seen": 24498344,
      "step": 42210
    },
    {
      "epoch": 6.2876079833184395,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.0259926707030991,
      "loss": 0.8134,
      "num_input_tokens_seen": 24501480,
      "step": 42215
    },
    {
      "epoch": 6.288352695859398,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.025991344074361943,
      "loss": 0.8129,
      "num_input_tokens_seen": 24504456,
      "step": 42220
    },
    {
      "epoch": 6.289097408400357,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.0259900172599372,
      "loss": 0.7952,
      "num_input_tokens_seen": 24507272,
      "step": 42225
    },
    {
      "epoch": 6.289842120941317,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.02598869025984728,
      "loss": 0.8111,
      "num_input_tokens_seen": 24510216,
      "step": 42230
    },
    {
      "epoch": 6.290586833482275,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.025987363074114613,
      "loss": 0.8186,
      "num_input_tokens_seen": 24512872,
      "step": 42235
    },
    {
      "epoch": 6.291331546023235,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.025986035702761607,
      "loss": 0.7761,
      "num_input_tokens_seen": 24515592,
      "step": 42240
    },
    {
      "epoch": 6.292076258564194,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.025984708145810706,
      "loss": 0.7937,
      "num_input_tokens_seen": 24518504,
      "step": 42245
    },
    {
      "epoch": 6.2928209711051535,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.025983380403284313,
      "loss": 0.8015,
      "num_input_tokens_seen": 24521480,
      "step": 42250
    },
    {
      "epoch": 6.293565683646112,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.025982052475204877,
      "loss": 0.8083,
      "num_input_tokens_seen": 24524328,
      "step": 42255
    },
    {
      "epoch": 6.294310396187072,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02598072436159482,
      "loss": 0.8174,
      "num_input_tokens_seen": 24527144,
      "step": 42260
    },
    {
      "epoch": 6.295055108728031,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02597939606247659,
      "loss": 0.8053,
      "num_input_tokens_seen": 24530344,
      "step": 42265
    },
    {
      "epoch": 6.29579982126899,
      "grad_norm": 0.03125,
      "learning_rate": 0.025978067577872622,
      "loss": 0.8132,
      "num_input_tokens_seen": 24533160,
      "step": 42270
    },
    {
      "epoch": 6.296544533809949,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.025976738907805365,
      "loss": 0.794,
      "num_input_tokens_seen": 24536328,
      "step": 42275
    },
    {
      "epoch": 6.297289246350909,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.025975410052297254,
      "loss": 0.8004,
      "num_input_tokens_seen": 24539144,
      "step": 42280
    },
    {
      "epoch": 6.2980339588918675,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.025974081011370744,
      "loss": 0.8079,
      "num_input_tokens_seen": 24541928,
      "step": 42285
    },
    {
      "epoch": 6.298778671432827,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02597275178504829,
      "loss": 0.7966,
      "num_input_tokens_seen": 24544744,
      "step": 42290
    },
    {
      "epoch": 6.299523383973786,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02597142237335235,
      "loss": 0.8024,
      "num_input_tokens_seen": 24547592,
      "step": 42295
    },
    {
      "epoch": 6.3002680965147455,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.025970092776305374,
      "loss": 0.809,
      "num_input_tokens_seen": 24550440,
      "step": 42300
    },
    {
      "epoch": 6.301012809055704,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.025968762993929837,
      "loss": 0.7988,
      "num_input_tokens_seen": 24553544,
      "step": 42305
    },
    {
      "epoch": 6.301757521596664,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.025967433026248196,
      "loss": 0.7984,
      "num_input_tokens_seen": 24556360,
      "step": 42310
    },
    {
      "epoch": 6.302502234137623,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.025966102873282916,
      "loss": 0.8029,
      "num_input_tokens_seen": 24559272,
      "step": 42315
    },
    {
      "epoch": 6.303246946678582,
      "grad_norm": 0.01953125,
      "learning_rate": 0.025964772535056475,
      "loss": 0.8008,
      "num_input_tokens_seen": 24562216,
      "step": 42320
    },
    {
      "epoch": 6.303991659219541,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.025963442011591353,
      "loss": 0.7865,
      "num_input_tokens_seen": 24565288,
      "step": 42325
    },
    {
      "epoch": 6.304736371760501,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.025962111302910013,
      "loss": 0.797,
      "num_input_tokens_seen": 24568296,
      "step": 42330
    },
    {
      "epoch": 6.3054810843014595,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.025960780409034948,
      "loss": 0.8151,
      "num_input_tokens_seen": 24571176,
      "step": 42335
    },
    {
      "epoch": 6.306225796842419,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.025959449329988633,
      "loss": 0.8172,
      "num_input_tokens_seen": 24574376,
      "step": 42340
    },
    {
      "epoch": 6.306970509383378,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.025958118065793567,
      "loss": 0.8093,
      "num_input_tokens_seen": 24577256,
      "step": 42345
    },
    {
      "epoch": 6.3077152219243375,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.02595678661647223,
      "loss": 0.778,
      "num_input_tokens_seen": 24580136,
      "step": 42350
    },
    {
      "epoch": 6.308459934465296,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.025955454982047117,
      "loss": 0.8168,
      "num_input_tokens_seen": 24583272,
      "step": 42355
    },
    {
      "epoch": 6.309204647006256,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02595412316254073,
      "loss": 0.8125,
      "num_input_tokens_seen": 24586376,
      "step": 42360
    },
    {
      "epoch": 6.309949359547215,
      "grad_norm": 0.021484375,
      "learning_rate": 0.025952791157975566,
      "loss": 0.788,
      "num_input_tokens_seen": 24589320,
      "step": 42365
    },
    {
      "epoch": 6.310694072088174,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.025951458968374124,
      "loss": 0.8012,
      "num_input_tokens_seen": 24592072,
      "step": 42370
    },
    {
      "epoch": 6.311438784629133,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.025950126593758914,
      "loss": 0.8092,
      "num_input_tokens_seen": 24594792,
      "step": 42375
    },
    {
      "epoch": 6.312183497170093,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02594879403415245,
      "loss": 0.807,
      "num_input_tokens_seen": 24597576,
      "step": 42380
    },
    {
      "epoch": 6.3129282097110515,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.025947461289577232,
      "loss": 0.8058,
      "num_input_tokens_seen": 24600840,
      "step": 42385
    },
    {
      "epoch": 6.31367292225201,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.025946128360055783,
      "loss": 0.8014,
      "num_input_tokens_seen": 24603784,
      "step": 42390
    },
    {
      "epoch": 6.31441763479297,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.025944795245610622,
      "loss": 0.8072,
      "num_input_tokens_seen": 24606952,
      "step": 42395
    },
    {
      "epoch": 6.31516234733393,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02594346194626427,
      "loss": 0.8071,
      "num_input_tokens_seen": 24609736,
      "step": 42400
    },
    {
      "epoch": 6.315907059874888,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.025942128462039252,
      "loss": 0.8016,
      "num_input_tokens_seen": 24612648,
      "step": 42405
    },
    {
      "epoch": 6.316651772415847,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02594079479295809,
      "loss": 0.8102,
      "num_input_tokens_seen": 24615304,
      "step": 42410
    },
    {
      "epoch": 6.317396484956807,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02593946093904332,
      "loss": 0.8047,
      "num_input_tokens_seen": 24618408,
      "step": 42415
    },
    {
      "epoch": 6.3181411974977655,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.025938126900317483,
      "loss": 0.7959,
      "num_input_tokens_seen": 24621672,
      "step": 42420
    },
    {
      "epoch": 6.318885910038725,
      "grad_norm": 0.021484375,
      "learning_rate": 0.025936792676803103,
      "loss": 0.7971,
      "num_input_tokens_seen": 24624488,
      "step": 42425
    },
    {
      "epoch": 6.319630622579684,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.025935458268522726,
      "loss": 0.8113,
      "num_input_tokens_seen": 24627528,
      "step": 42430
    },
    {
      "epoch": 6.3203753351206435,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.025934123675498896,
      "loss": 0.8021,
      "num_input_tokens_seen": 24630216,
      "step": 42435
    },
    {
      "epoch": 6.321120047661602,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.025932788897754164,
      "loss": 0.7986,
      "num_input_tokens_seen": 24633320,
      "step": 42440
    },
    {
      "epoch": 6.321864760202562,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.025931453935311072,
      "loss": 0.8032,
      "num_input_tokens_seen": 24636040,
      "step": 42445
    },
    {
      "epoch": 6.322609472743521,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.025930118788192173,
      "loss": 0.8154,
      "num_input_tokens_seen": 24639272,
      "step": 42450
    },
    {
      "epoch": 6.32335418528448,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.025928783456420027,
      "loss": 0.7936,
      "num_input_tokens_seen": 24642312,
      "step": 42455
    },
    {
      "epoch": 6.324098897825439,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.025927447940017195,
      "loss": 0.8031,
      "num_input_tokens_seen": 24645448,
      "step": 42460
    },
    {
      "epoch": 6.324843610366399,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.025926112239006236,
      "loss": 0.8009,
      "num_input_tokens_seen": 24648264,
      "step": 42465
    },
    {
      "epoch": 6.3255883229073575,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.025924776353409715,
      "loss": 0.8114,
      "num_input_tokens_seen": 24651144,
      "step": 42470
    },
    {
      "epoch": 6.326333035448317,
      "grad_norm": 0.01953125,
      "learning_rate": 0.0259234402832502,
      "loss": 0.7782,
      "num_input_tokens_seen": 24653928,
      "step": 42475
    },
    {
      "epoch": 6.327077747989276,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.025922104028550264,
      "loss": 0.7974,
      "num_input_tokens_seen": 24656840,
      "step": 42480
    },
    {
      "epoch": 6.327822460530236,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.025920767589332483,
      "loss": 0.8176,
      "num_input_tokens_seen": 24659656,
      "step": 42485
    },
    {
      "epoch": 6.328567173071194,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02591943096561943,
      "loss": 0.79,
      "num_input_tokens_seen": 24662440,
      "step": 42490
    },
    {
      "epoch": 6.329311885612154,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.025918094157433694,
      "loss": 0.792,
      "num_input_tokens_seen": 24665224,
      "step": 42495
    },
    {
      "epoch": 6.330056598153113,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.025916757164797844,
      "loss": 0.8031,
      "num_input_tokens_seen": 24668168,
      "step": 42500
    },
    {
      "epoch": 6.330801310694072,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.025915419987734487,
      "loss": 0.8032,
      "num_input_tokens_seen": 24670760,
      "step": 42505
    },
    {
      "epoch": 6.331546023235031,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.0259140826262662,
      "loss": 0.8123,
      "num_input_tokens_seen": 24673832,
      "step": 42510
    },
    {
      "epoch": 6.332290735775991,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.025912745080415574,
      "loss": 0.7929,
      "num_input_tokens_seen": 24676776,
      "step": 42515
    },
    {
      "epoch": 6.3330354483169495,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.025911407350205216,
      "loss": 0.785,
      "num_input_tokens_seen": 24679688,
      "step": 42520
    },
    {
      "epoch": 6.333780160857909,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02591006943565772,
      "loss": 0.7896,
      "num_input_tokens_seen": 24682408,
      "step": 42525
    },
    {
      "epoch": 6.334524873398868,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.025908731336795693,
      "loss": 0.7938,
      "num_input_tokens_seen": 24685256,
      "step": 42530
    },
    {
      "epoch": 6.335269585939828,
      "grad_norm": 0.013671875,
      "learning_rate": 0.02590739305364173,
      "loss": 0.791,
      "num_input_tokens_seen": 24688008,
      "step": 42535
    },
    {
      "epoch": 6.336014298480786,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.025906054586218457,
      "loss": 0.7865,
      "num_input_tokens_seen": 24690824,
      "step": 42540
    },
    {
      "epoch": 6.336759011021746,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.025904715934548468,
      "loss": 0.7824,
      "num_input_tokens_seen": 24693832,
      "step": 42545
    },
    {
      "epoch": 6.337503723562705,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.025903377098654385,
      "loss": 0.7751,
      "num_input_tokens_seen": 24696872,
      "step": 42550
    },
    {
      "epoch": 6.338248436103664,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.025902038078558832,
      "loss": 0.7723,
      "num_input_tokens_seen": 24699784,
      "step": 42555
    },
    {
      "epoch": 6.338993148644623,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02590069887428443,
      "loss": 0.7799,
      "num_input_tokens_seen": 24702920,
      "step": 42560
    },
    {
      "epoch": 6.339737861185583,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02589935948585379,
      "loss": 0.7915,
      "num_input_tokens_seen": 24705928,
      "step": 42565
    },
    {
      "epoch": 6.340482573726542,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.025898019913289556,
      "loss": 0.7623,
      "num_input_tokens_seen": 24708808,
      "step": 42570
    },
    {
      "epoch": 6.3412272862675,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02589668015661435,
      "loss": 0.7848,
      "num_input_tokens_seen": 24711624,
      "step": 42575
    },
    {
      "epoch": 6.34197199880846,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.02589534021585081,
      "loss": 0.8208,
      "num_input_tokens_seen": 24714728,
      "step": 42580
    },
    {
      "epoch": 6.342716711349419,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.025894000091021566,
      "loss": 0.7905,
      "num_input_tokens_seen": 24717384,
      "step": 42585
    },
    {
      "epoch": 6.343461423890378,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.025892659782149264,
      "loss": 0.8197,
      "num_input_tokens_seen": 24720296,
      "step": 42590
    },
    {
      "epoch": 6.344206136431337,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.025891319289256545,
      "loss": 0.8021,
      "num_input_tokens_seen": 24723272,
      "step": 42595
    },
    {
      "epoch": 6.344950848972297,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.025889978612366058,
      "loss": 0.7931,
      "num_input_tokens_seen": 24726152,
      "step": 42600
    },
    {
      "epoch": 6.3456955615132555,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.025888637751500448,
      "loss": 0.8213,
      "num_input_tokens_seen": 24729000,
      "step": 42605
    },
    {
      "epoch": 6.346440274054215,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.025887296706682376,
      "loss": 0.8143,
      "num_input_tokens_seen": 24731816,
      "step": 42610
    },
    {
      "epoch": 6.347184986595174,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025885955477934486,
      "loss": 0.8078,
      "num_input_tokens_seen": 24734568,
      "step": 42615
    },
    {
      "epoch": 6.347929699136134,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.025884614065279445,
      "loss": 0.7744,
      "num_input_tokens_seen": 24737672,
      "step": 42620
    },
    {
      "epoch": 6.348674411677092,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02588327246873991,
      "loss": 0.779,
      "num_input_tokens_seen": 24740424,
      "step": 42625
    },
    {
      "epoch": 6.349419124218052,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02588193068833855,
      "loss": 0.7782,
      "num_input_tokens_seen": 24743240,
      "step": 42630
    },
    {
      "epoch": 6.350163836759011,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.025880588724098024,
      "loss": 0.7779,
      "num_input_tokens_seen": 24746120,
      "step": 42635
    },
    {
      "epoch": 6.35090854929997,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02587924657604102,
      "loss": 0.7881,
      "num_input_tokens_seen": 24749160,
      "step": 42640
    },
    {
      "epoch": 6.351653261840929,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.025877904244190196,
      "loss": 0.7917,
      "num_input_tokens_seen": 24752008,
      "step": 42645
    },
    {
      "epoch": 6.352397974381889,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.025876561728568236,
      "loss": 0.7805,
      "num_input_tokens_seen": 24754824,
      "step": 42650
    },
    {
      "epoch": 6.353142686922848,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02587521902919782,
      "loss": 0.7812,
      "num_input_tokens_seen": 24757576,
      "step": 42655
    },
    {
      "epoch": 6.353887399463807,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.025873876146101638,
      "loss": 0.8076,
      "num_input_tokens_seen": 24760168,
      "step": 42660
    },
    {
      "epoch": 6.354632112004766,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.025872533079302363,
      "loss": 0.7782,
      "num_input_tokens_seen": 24763240,
      "step": 42665
    },
    {
      "epoch": 6.355376824545726,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02587118982882269,
      "loss": 0.8324,
      "num_input_tokens_seen": 24765960,
      "step": 42670
    },
    {
      "epoch": 6.356121537086684,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02586984639468532,
      "loss": 0.8179,
      "num_input_tokens_seen": 24768808,
      "step": 42675
    },
    {
      "epoch": 6.356866249627644,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.025868502776912942,
      "loss": 0.7999,
      "num_input_tokens_seen": 24771912,
      "step": 42680
    },
    {
      "epoch": 6.357610962168603,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.025867158975528257,
      "loss": 0.8019,
      "num_input_tokens_seen": 24774952,
      "step": 42685
    },
    {
      "epoch": 6.358355674709562,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02586581499055396,
      "loss": 0.794,
      "num_input_tokens_seen": 24777832,
      "step": 42690
    },
    {
      "epoch": 6.359100387250521,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.025864470822012764,
      "loss": 0.7984,
      "num_input_tokens_seen": 24781160,
      "step": 42695
    },
    {
      "epoch": 6.359845099791481,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02586312646992738,
      "loss": 0.8114,
      "num_input_tokens_seen": 24784104,
      "step": 42700
    },
    {
      "epoch": 6.36058981233244,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.02586178193432051,
      "loss": 0.7814,
      "num_input_tokens_seen": 24786632,
      "step": 42705
    },
    {
      "epoch": 6.361334524873399,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.025860437215214877,
      "loss": 0.8152,
      "num_input_tokens_seen": 24789544,
      "step": 42710
    },
    {
      "epoch": 6.362079237414358,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.025859092312633194,
      "loss": 0.8113,
      "num_input_tokens_seen": 24792488,
      "step": 42715
    },
    {
      "epoch": 6.362823949955318,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.025857747226598185,
      "loss": 0.7997,
      "num_input_tokens_seen": 24795272,
      "step": 42720
    },
    {
      "epoch": 6.363568662496276,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.025856401957132572,
      "loss": 0.8266,
      "num_input_tokens_seen": 24798152,
      "step": 42725
    },
    {
      "epoch": 6.364313375037236,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.025855056504259077,
      "loss": 0.795,
      "num_input_tokens_seen": 24801000,
      "step": 42730
    },
    {
      "epoch": 6.365058087578195,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.02585371086800044,
      "loss": 0.8349,
      "num_input_tokens_seen": 24803816,
      "step": 42735
    },
    {
      "epoch": 6.365802800119154,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02585236504837939,
      "loss": 0.7767,
      "num_input_tokens_seen": 24806856,
      "step": 42740
    },
    {
      "epoch": 6.366547512660113,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.025851019045418658,
      "loss": 0.8055,
      "num_input_tokens_seen": 24809608,
      "step": 42745
    },
    {
      "epoch": 6.367292225201073,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.025849672859140988,
      "loss": 0.8231,
      "num_input_tokens_seen": 24812584,
      "step": 42750
    },
    {
      "epoch": 6.368036937742032,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.025848326489569123,
      "loss": 0.811,
      "num_input_tokens_seen": 24815272,
      "step": 42755
    },
    {
      "epoch": 6.36878165028299,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.025846979936725807,
      "loss": 0.7981,
      "num_input_tokens_seen": 24818376,
      "step": 42760
    },
    {
      "epoch": 6.36952636282395,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02584563320063379,
      "loss": 0.7766,
      "num_input_tokens_seen": 24821160,
      "step": 42765
    },
    {
      "epoch": 6.370271075364909,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.025844286281315822,
      "loss": 0.8255,
      "num_input_tokens_seen": 24824008,
      "step": 42770
    },
    {
      "epoch": 6.371015787905868,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.025842939178794665,
      "loss": 0.8257,
      "num_input_tokens_seen": 24826696,
      "step": 42775
    },
    {
      "epoch": 6.371760500446827,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.025841591893093065,
      "loss": 0.804,
      "num_input_tokens_seen": 24829384,
      "step": 42780
    },
    {
      "epoch": 6.372505212987787,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02584024442423379,
      "loss": 0.7952,
      "num_input_tokens_seen": 24832168,
      "step": 42785
    },
    {
      "epoch": 6.373249925528746,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.025838896772239606,
      "loss": 0.8025,
      "num_input_tokens_seen": 24835208,
      "step": 42790
    },
    {
      "epoch": 6.373994638069705,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.025837548937133272,
      "loss": 0.7978,
      "num_input_tokens_seen": 24837896,
      "step": 42795
    },
    {
      "epoch": 6.374739350610664,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.025836200918937565,
      "loss": 0.8249,
      "num_input_tokens_seen": 24840936,
      "step": 42800
    },
    {
      "epoch": 6.375484063151624,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.025834852717675258,
      "loss": 0.8152,
      "num_input_tokens_seen": 24843816,
      "step": 42805
    },
    {
      "epoch": 6.376228775692582,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.025833504333369128,
      "loss": 0.8065,
      "num_input_tokens_seen": 24846440,
      "step": 42810
    },
    {
      "epoch": 6.376973488233542,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02583215576604195,
      "loss": 0.7968,
      "num_input_tokens_seen": 24849480,
      "step": 42815
    },
    {
      "epoch": 6.377718200774501,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.025830807015716512,
      "loss": 0.8107,
      "num_input_tokens_seen": 24852360,
      "step": 42820
    },
    {
      "epoch": 6.3784629133154604,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.025829458082415595,
      "loss": 0.7891,
      "num_input_tokens_seen": 24855464,
      "step": 42825
    },
    {
      "epoch": 6.379207625856419,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.025828108966161993,
      "loss": 0.7997,
      "num_input_tokens_seen": 24858344,
      "step": 42830
    },
    {
      "epoch": 6.379952338397379,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.025826759666978494,
      "loss": 0.8336,
      "num_input_tokens_seen": 24861352,
      "step": 42835
    },
    {
      "epoch": 6.380697050938338,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.025825410184887893,
      "loss": 0.8049,
      "num_input_tokens_seen": 24864264,
      "step": 42840
    },
    {
      "epoch": 6.381441763479297,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02582406051991299,
      "loss": 0.7916,
      "num_input_tokens_seen": 24867656,
      "step": 42845
    },
    {
      "epoch": 6.382186476020256,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.025822710672076588,
      "loss": 0.777,
      "num_input_tokens_seen": 24870344,
      "step": 42850
    },
    {
      "epoch": 6.382931188561216,
      "grad_norm": 0.02734375,
      "learning_rate": 0.025821360641401486,
      "loss": 0.8088,
      "num_input_tokens_seen": 24873224,
      "step": 42855
    },
    {
      "epoch": 6.383675901102174,
      "grad_norm": 0.021484375,
      "learning_rate": 0.025820010427910496,
      "loss": 0.7873,
      "num_input_tokens_seen": 24875976,
      "step": 42860
    },
    {
      "epoch": 6.384420613643134,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.025818660031626424,
      "loss": 0.7951,
      "num_input_tokens_seen": 24879144,
      "step": 42865
    },
    {
      "epoch": 6.385165326184093,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.025817309452572088,
      "loss": 0.8028,
      "num_input_tokens_seen": 24881960,
      "step": 42870
    },
    {
      "epoch": 6.3859100387250525,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.025815958690770302,
      "loss": 0.7812,
      "num_input_tokens_seen": 24884840,
      "step": 42875
    },
    {
      "epoch": 6.386654751266011,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.02581460774624389,
      "loss": 0.8144,
      "num_input_tokens_seen": 24887464,
      "step": 42880
    },
    {
      "epoch": 6.387399463806971,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.025813256619015672,
      "loss": 0.8237,
      "num_input_tokens_seen": 24890408,
      "step": 42885
    },
    {
      "epoch": 6.38814417634793,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.02581190530910847,
      "loss": 0.811,
      "num_input_tokens_seen": 24893608,
      "step": 42890
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.025810553816545116,
      "loss": 0.8092,
      "num_input_tokens_seen": 24896136,
      "step": 42895
    },
    {
      "epoch": 6.389633601429848,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.025809202141348447,
      "loss": 0.8109,
      "num_input_tokens_seen": 24899208,
      "step": 42900
    },
    {
      "epoch": 6.390378313970807,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.02580785028354129,
      "loss": 0.7857,
      "num_input_tokens_seen": 24902120,
      "step": 42905
    },
    {
      "epoch": 6.3911230265117664,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02580649824314649,
      "loss": 0.788,
      "num_input_tokens_seen": 24904776,
      "step": 42910
    },
    {
      "epoch": 6.391867739052726,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.025805146020186884,
      "loss": 0.7942,
      "num_input_tokens_seen": 24907688,
      "step": 42915
    },
    {
      "epoch": 6.392612451593685,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.025803793614685315,
      "loss": 0.8055,
      "num_input_tokens_seen": 24910792,
      "step": 42920
    },
    {
      "epoch": 6.393357164134644,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.025802441026664638,
      "loss": 0.794,
      "num_input_tokens_seen": 24913736,
      "step": 42925
    },
    {
      "epoch": 6.394101876675603,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.025801088256147697,
      "loss": 0.8318,
      "num_input_tokens_seen": 24916712,
      "step": 42930
    },
    {
      "epoch": 6.394846589216562,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.025799735303157344,
      "loss": 0.8229,
      "num_input_tokens_seen": 24919432,
      "step": 42935
    },
    {
      "epoch": 6.395591301757522,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.025798382167716447,
      "loss": 0.7874,
      "num_input_tokens_seen": 24922184,
      "step": 42940
    },
    {
      "epoch": 6.39633601429848,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02579702884984785,
      "loss": 0.8091,
      "num_input_tokens_seen": 24925000,
      "step": 42945
    },
    {
      "epoch": 6.39708072683944,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02579567534957443,
      "loss": 0.7906,
      "num_input_tokens_seen": 24928232,
      "step": 42950
    },
    {
      "epoch": 6.397825439380399,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.025794321666919046,
      "loss": 0.8083,
      "num_input_tokens_seen": 24930952,
      "step": 42955
    },
    {
      "epoch": 6.3985701519213585,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.025792967801904567,
      "loss": 0.7935,
      "num_input_tokens_seen": 24933864,
      "step": 42960
    },
    {
      "epoch": 6.399314864462317,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02579161375455387,
      "loss": 0.8229,
      "num_input_tokens_seen": 24936648,
      "step": 42965
    },
    {
      "epoch": 6.400059577003277,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.025790259524889822,
      "loss": 0.8018,
      "num_input_tokens_seen": 24939688,
      "step": 42970
    },
    {
      "epoch": 6.400804289544236,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02578890511293531,
      "loss": 0.7941,
      "num_input_tokens_seen": 24942376,
      "step": 42975
    },
    {
      "epoch": 6.401549002085195,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.025787550518713207,
      "loss": 0.7962,
      "num_input_tokens_seen": 24945000,
      "step": 42980
    },
    {
      "epoch": 6.402293714626154,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.025786195742246405,
      "loss": 0.8222,
      "num_input_tokens_seen": 24947848,
      "step": 42985
    },
    {
      "epoch": 6.403038427167114,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02578484078355779,
      "loss": 0.7878,
      "num_input_tokens_seen": 24950696,
      "step": 42990
    },
    {
      "epoch": 6.4037831397080724,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.02578348564267025,
      "loss": 0.7837,
      "num_input_tokens_seen": 24953736,
      "step": 42995
    },
    {
      "epoch": 6.404527852249032,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.025782130319606678,
      "loss": 0.7848,
      "num_input_tokens_seen": 24956616,
      "step": 43000
    },
    {
      "epoch": 6.405272564789991,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02578077481438997,
      "loss": 0.7971,
      "num_input_tokens_seen": 24959656,
      "step": 43005
    },
    {
      "epoch": 6.4060172773309505,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.025779419127043037,
      "loss": 0.7945,
      "num_input_tokens_seen": 24962504,
      "step": 43010
    },
    {
      "epoch": 6.406761989871909,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.025778063257588767,
      "loss": 0.8008,
      "num_input_tokens_seen": 24965288,
      "step": 43015
    },
    {
      "epoch": 6.407506702412869,
      "grad_norm": 0.010498046875,
      "learning_rate": 0.025776707206050072,
      "loss": 0.7808,
      "num_input_tokens_seen": 24968168,
      "step": 43020
    },
    {
      "epoch": 6.408251414953828,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.025775350972449866,
      "loss": 0.817,
      "num_input_tokens_seen": 24970856,
      "step": 43025
    },
    {
      "epoch": 6.408996127494787,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.025773994556811057,
      "loss": 0.7884,
      "num_input_tokens_seen": 24973736,
      "step": 43030
    },
    {
      "epoch": 6.409740840035746,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02577263795915656,
      "loss": 0.8161,
      "num_input_tokens_seen": 24976648,
      "step": 43035
    },
    {
      "epoch": 6.410485552576706,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.025771281179509293,
      "loss": 0.8438,
      "num_input_tokens_seen": 24979496,
      "step": 43040
    },
    {
      "epoch": 6.4112302651176645,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.025769924217892176,
      "loss": 0.8094,
      "num_input_tokens_seen": 24982248,
      "step": 43045
    },
    {
      "epoch": 6.411974977658624,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.025768567074328133,
      "loss": 0.7957,
      "num_input_tokens_seen": 24984936,
      "step": 43050
    },
    {
      "epoch": 6.412719690199583,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.025767209748840104,
      "loss": 0.7855,
      "num_input_tokens_seen": 24988040,
      "step": 43055
    },
    {
      "epoch": 6.4134644027405425,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.025765852241450997,
      "loss": 0.8217,
      "num_input_tokens_seen": 24990824,
      "step": 43060
    },
    {
      "epoch": 6.414209115281501,
      "grad_norm": 0.013671875,
      "learning_rate": 0.025764494552183763,
      "loss": 0.8096,
      "num_input_tokens_seen": 24993736,
      "step": 43065
    },
    {
      "epoch": 6.414953827822461,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.025763136681061335,
      "loss": 0.8141,
      "num_input_tokens_seen": 24996520,
      "step": 43070
    },
    {
      "epoch": 6.41569854036342,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.025761778628106656,
      "loss": 0.7796,
      "num_input_tokens_seen": 24999240,
      "step": 43075
    },
    {
      "epoch": 6.416443252904379,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02576042039334266,
      "loss": 0.8117,
      "num_input_tokens_seen": 25002216,
      "step": 43080
    },
    {
      "epoch": 6.417187965445338,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.0257590619767923,
      "loss": 0.8071,
      "num_input_tokens_seen": 25004872,
      "step": 43085
    },
    {
      "epoch": 6.417932677986297,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.025757703378478523,
      "loss": 0.8,
      "num_input_tokens_seen": 25008136,
      "step": 43090
    },
    {
      "epoch": 6.4186773905272565,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.025756344598424276,
      "loss": 0.7914,
      "num_input_tokens_seen": 25011048,
      "step": 43095
    },
    {
      "epoch": 6.419422103068215,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.025754985636652523,
      "loss": 0.7896,
      "num_input_tokens_seen": 25013896,
      "step": 43100
    },
    {
      "epoch": 6.420166815609175,
      "grad_norm": 0.01422119140625,
      "learning_rate": 0.025753626493186223,
      "loss": 0.7801,
      "num_input_tokens_seen": 25016648,
      "step": 43105
    },
    {
      "epoch": 6.420911528150134,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.025752267168048324,
      "loss": 0.8215,
      "num_input_tokens_seen": 25019880,
      "step": 43110
    },
    {
      "epoch": 6.421656240691093,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.025750907661261806,
      "loss": 0.8243,
      "num_input_tokens_seen": 25022760,
      "step": 43115
    },
    {
      "epoch": 6.422400953232052,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.02574954797284963,
      "loss": 0.8095,
      "num_input_tokens_seen": 25025608,
      "step": 43120
    },
    {
      "epoch": 6.423145665773012,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.025748188102834763,
      "loss": 0.7996,
      "num_input_tokens_seen": 25028648,
      "step": 43125
    },
    {
      "epoch": 6.4238903783139705,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.025746828051240185,
      "loss": 0.7983,
      "num_input_tokens_seen": 25031784,
      "step": 43130
    },
    {
      "epoch": 6.42463509085493,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02574546781808886,
      "loss": 0.7977,
      "num_input_tokens_seen": 25034664,
      "step": 43135
    },
    {
      "epoch": 6.425379803395889,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.025744107403403788,
      "loss": 0.8037,
      "num_input_tokens_seen": 25037576,
      "step": 43140
    },
    {
      "epoch": 6.4261245159368485,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02574274680720794,
      "loss": 0.7838,
      "num_input_tokens_seen": 25040296,
      "step": 43145
    },
    {
      "epoch": 6.426869228477807,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.0257413860295243,
      "loss": 0.7976,
      "num_input_tokens_seen": 25043016,
      "step": 43150
    },
    {
      "epoch": 6.427613941018767,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.025740025070375865,
      "loss": 0.7942,
      "num_input_tokens_seen": 25046184,
      "step": 43155
    },
    {
      "epoch": 6.428358653559726,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02573866392978562,
      "loss": 0.7962,
      "num_input_tokens_seen": 25049064,
      "step": 43160
    },
    {
      "epoch": 6.429103366100685,
      "grad_norm": 0.01953125,
      "learning_rate": 0.025737302607776555,
      "loss": 0.8163,
      "num_input_tokens_seen": 25051848,
      "step": 43165
    },
    {
      "epoch": 6.429848078641644,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02573594110437168,
      "loss": 0.8178,
      "num_input_tokens_seen": 25054696,
      "step": 43170
    },
    {
      "epoch": 6.430592791182604,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.025734579419593995,
      "loss": 0.8164,
      "num_input_tokens_seen": 25057544,
      "step": 43175
    },
    {
      "epoch": 6.4313375037235625,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.025733217553466498,
      "loss": 0.8021,
      "num_input_tokens_seen": 25060648,
      "step": 43180
    },
    {
      "epoch": 6.432082216264522,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.025731855506012202,
      "loss": 0.8263,
      "num_input_tokens_seen": 25063432,
      "step": 43185
    },
    {
      "epoch": 6.432826928805481,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.025730493277254112,
      "loss": 0.7976,
      "num_input_tokens_seen": 25066056,
      "step": 43190
    },
    {
      "epoch": 6.4335716413464406,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.025729130867215243,
      "loss": 0.7892,
      "num_input_tokens_seen": 25069384,
      "step": 43195
    },
    {
      "epoch": 6.434316353887399,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.025727768275918612,
      "loss": 0.8266,
      "num_input_tokens_seen": 25072168,
      "step": 43200
    },
    {
      "epoch": 6.435061066428359,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.025726405503387246,
      "loss": 0.7686,
      "num_input_tokens_seen": 25075016,
      "step": 43205
    },
    {
      "epoch": 6.435805778969318,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.025725042549644157,
      "loss": 0.7965,
      "num_input_tokens_seen": 25077768,
      "step": 43210
    },
    {
      "epoch": 6.436550491510277,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.025723679414712375,
      "loss": 0.814,
      "num_input_tokens_seen": 25080360,
      "step": 43215
    },
    {
      "epoch": 6.437295204051236,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.025722316098614924,
      "loss": 0.8414,
      "num_input_tokens_seen": 25083496,
      "step": 43220
    },
    {
      "epoch": 6.438039916592196,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.025720952601374844,
      "loss": 0.8293,
      "num_input_tokens_seen": 25086536,
      "step": 43225
    },
    {
      "epoch": 6.4387846291331545,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.025719588923015166,
      "loss": 0.8057,
      "num_input_tokens_seen": 25089448,
      "step": 43230
    },
    {
      "epoch": 6.439529341674114,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.025718225063558926,
      "loss": 0.7921,
      "num_input_tokens_seen": 25092296,
      "step": 43235
    },
    {
      "epoch": 6.440274054215073,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.025716861023029166,
      "loss": 0.8044,
      "num_input_tokens_seen": 25095176,
      "step": 43240
    },
    {
      "epoch": 6.441018766756033,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.025715496801448936,
      "loss": 0.8142,
      "num_input_tokens_seen": 25097992,
      "step": 43245
    },
    {
      "epoch": 6.441763479296991,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.025714132398841277,
      "loss": 0.8152,
      "num_input_tokens_seen": 25100776,
      "step": 43250
    },
    {
      "epoch": 6.44250819183795,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02571276781522924,
      "loss": 0.8137,
      "num_input_tokens_seen": 25103880,
      "step": 43255
    },
    {
      "epoch": 6.44325290437891,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.025711403050635878,
      "loss": 0.7932,
      "num_input_tokens_seen": 25106760,
      "step": 43260
    },
    {
      "epoch": 6.443997616919869,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.025710038105084248,
      "loss": 0.7974,
      "num_input_tokens_seen": 25109672,
      "step": 43265
    },
    {
      "epoch": 6.444742329460828,
      "grad_norm": 0.01123046875,
      "learning_rate": 0.025708672978597406,
      "loss": 0.7964,
      "num_input_tokens_seen": 25112424,
      "step": 43270
    },
    {
      "epoch": 6.445487042001787,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.025707307671198425,
      "loss": 0.8092,
      "num_input_tokens_seen": 25115112,
      "step": 43275
    },
    {
      "epoch": 6.4462317545427466,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.025705942182910356,
      "loss": 0.8064,
      "num_input_tokens_seen": 25118056,
      "step": 43280
    },
    {
      "epoch": 6.446976467083705,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02570457651375628,
      "loss": 0.8049,
      "num_input_tokens_seen": 25121064,
      "step": 43285
    },
    {
      "epoch": 6.447721179624665,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.025703210663759263,
      "loss": 0.8093,
      "num_input_tokens_seen": 25124040,
      "step": 43290
    },
    {
      "epoch": 6.448465892165624,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02570184463294238,
      "loss": 0.8055,
      "num_input_tokens_seen": 25126856,
      "step": 43295
    },
    {
      "epoch": 6.449210604706583,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025700478421328707,
      "loss": 0.7999,
      "num_input_tokens_seen": 25129704,
      "step": 43300
    },
    {
      "epoch": 6.449955317247542,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.025699112028941328,
      "loss": 0.8002,
      "num_input_tokens_seen": 25132520,
      "step": 43305
    },
    {
      "epoch": 6.450700029788502,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.025697745455803327,
      "loss": 0.7882,
      "num_input_tokens_seen": 25135272,
      "step": 43310
    },
    {
      "epoch": 6.4514447423294605,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02569637870193778,
      "loss": 0.7997,
      "num_input_tokens_seen": 25138440,
      "step": 43315
    },
    {
      "epoch": 6.45218945487042,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.025695011767367798,
      "loss": 0.8191,
      "num_input_tokens_seen": 25141576,
      "step": 43320
    },
    {
      "epoch": 6.452934167411379,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.025693644652116456,
      "loss": 0.8081,
      "num_input_tokens_seen": 25144264,
      "step": 43325
    },
    {
      "epoch": 6.453678879952339,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.025692277356206855,
      "loss": 0.799,
      "num_input_tokens_seen": 25147400,
      "step": 43330
    },
    {
      "epoch": 6.454423592493297,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.0256909098796621,
      "loss": 0.792,
      "num_input_tokens_seen": 25150472,
      "step": 43335
    },
    {
      "epoch": 6.455168305034257,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.025689542222505288,
      "loss": 0.7769,
      "num_input_tokens_seen": 25153192,
      "step": 43340
    },
    {
      "epoch": 6.455913017575216,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02568817438475952,
      "loss": 0.8035,
      "num_input_tokens_seen": 25156200,
      "step": 43345
    },
    {
      "epoch": 6.456657730116175,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.025686806366447913,
      "loss": 0.7837,
      "num_input_tokens_seen": 25158984,
      "step": 43350
    },
    {
      "epoch": 6.457402442657134,
      "grad_norm": 0.01031494140625,
      "learning_rate": 0.02568543816759357,
      "loss": 0.7882,
      "num_input_tokens_seen": 25161640,
      "step": 43355
    },
    {
      "epoch": 6.458147155198094,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.02568406978821961,
      "loss": 0.7791,
      "num_input_tokens_seen": 25164360,
      "step": 43360
    },
    {
      "epoch": 6.4588918677390526,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.025682701228349154,
      "loss": 0.8035,
      "num_input_tokens_seen": 25167112,
      "step": 43365
    },
    {
      "epoch": 6.459636580280012,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.025681332488005318,
      "loss": 0.8018,
      "num_input_tokens_seen": 25169928,
      "step": 43370
    },
    {
      "epoch": 6.460381292820971,
      "grad_norm": 0.0126953125,
      "learning_rate": 0.025679963567211223,
      "loss": 0.7796,
      "num_input_tokens_seen": 25172712,
      "step": 43375
    },
    {
      "epoch": 6.461126005361931,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.025678594465989996,
      "loss": 0.7779,
      "num_input_tokens_seen": 25175592,
      "step": 43380
    },
    {
      "epoch": 6.461870717902889,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02567722518436477,
      "loss": 0.8013,
      "num_input_tokens_seen": 25178376,
      "step": 43385
    },
    {
      "epoch": 6.462615430443849,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.025675855722358677,
      "loss": 0.778,
      "num_input_tokens_seen": 25181064,
      "step": 43390
    },
    {
      "epoch": 6.463360142984808,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.025674486079994854,
      "loss": 0.815,
      "num_input_tokens_seen": 25184104,
      "step": 43395
    },
    {
      "epoch": 6.464104855525767,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.025673116257296437,
      "loss": 0.7917,
      "num_input_tokens_seen": 25186920,
      "step": 43400
    },
    {
      "epoch": 6.464849568066726,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.02567174625428657,
      "loss": 0.7978,
      "num_input_tokens_seen": 25189928,
      "step": 43405
    },
    {
      "epoch": 6.465594280607686,
      "grad_norm": 0.0123291015625,
      "learning_rate": 0.02567037607098839,
      "loss": 0.8138,
      "num_input_tokens_seen": 25192872,
      "step": 43410
    },
    {
      "epoch": 6.466338993148645,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.025669005707425054,
      "loss": 0.8274,
      "num_input_tokens_seen": 25195496,
      "step": 43415
    },
    {
      "epoch": 6.467083705689604,
      "grad_norm": 0.0234375,
      "learning_rate": 0.025667635163619713,
      "loss": 0.7846,
      "num_input_tokens_seen": 25198088,
      "step": 43420
    },
    {
      "epoch": 6.467828418230563,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.025666264439595517,
      "loss": 0.7917,
      "num_input_tokens_seen": 25201224,
      "step": 43425
    },
    {
      "epoch": 6.468573130771523,
      "grad_norm": 0.01141357421875,
      "learning_rate": 0.02566489353537562,
      "loss": 0.7979,
      "num_input_tokens_seen": 25204488,
      "step": 43430
    },
    {
      "epoch": 6.469317843312481,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.025663522450983193,
      "loss": 0.7848,
      "num_input_tokens_seen": 25207752,
      "step": 43435
    },
    {
      "epoch": 6.47006255585344,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.025662151186441388,
      "loss": 0.7878,
      "num_input_tokens_seen": 25210664,
      "step": 43440
    },
    {
      "epoch": 6.4708072683944,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.025660779741773374,
      "loss": 0.8404,
      "num_input_tokens_seen": 25213512,
      "step": 43445
    },
    {
      "epoch": 6.4715519809353586,
      "grad_norm": 0.017578125,
      "learning_rate": 0.025659408117002323,
      "loss": 0.8172,
      "num_input_tokens_seen": 25216456,
      "step": 43450
    },
    {
      "epoch": 6.472296693476318,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.025658036312151408,
      "loss": 0.8187,
      "num_input_tokens_seen": 25219144,
      "step": 43455
    },
    {
      "epoch": 6.473041406017277,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.025656664327243797,
      "loss": 0.8206,
      "num_input_tokens_seen": 25222024,
      "step": 43460
    },
    {
      "epoch": 6.473786118558237,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.025655292162302674,
      "loss": 0.806,
      "num_input_tokens_seen": 25224616,
      "step": 43465
    },
    {
      "epoch": 6.474530831099195,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.025653919817351226,
      "loss": 0.7785,
      "num_input_tokens_seen": 25227400,
      "step": 43470
    },
    {
      "epoch": 6.475275543640155,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02565254729241262,
      "loss": 0.8017,
      "num_input_tokens_seen": 25230088,
      "step": 43475
    },
    {
      "epoch": 6.476020256181114,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.02565117458751006,
      "loss": 0.7848,
      "num_input_tokens_seen": 25233032,
      "step": 43480
    },
    {
      "epoch": 6.476764968722073,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.025649801702666734,
      "loss": 0.7892,
      "num_input_tokens_seen": 25235848,
      "step": 43485
    },
    {
      "epoch": 6.477509681263032,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.02564842863790583,
      "loss": 0.7942,
      "num_input_tokens_seen": 25239080,
      "step": 43490
    },
    {
      "epoch": 6.478254393803992,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.025647055393250545,
      "loss": 0.8076,
      "num_input_tokens_seen": 25242120,
      "step": 43495
    },
    {
      "epoch": 6.478999106344951,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02564568196872408,
      "loss": 0.7852,
      "num_input_tokens_seen": 25245032,
      "step": 43500
    },
    {
      "epoch": 6.47974381888591,
      "grad_norm": 0.021484375,
      "learning_rate": 0.025644308364349642,
      "loss": 0.7911,
      "num_input_tokens_seen": 25247688,
      "step": 43505
    },
    {
      "epoch": 6.480488531426869,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.025642934580150427,
      "loss": 0.7923,
      "num_input_tokens_seen": 25250728,
      "step": 43510
    },
    {
      "epoch": 6.481233243967829,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.025641560616149653,
      "loss": 0.7986,
      "num_input_tokens_seen": 25253544,
      "step": 43515
    },
    {
      "epoch": 6.481977956508787,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.025640186472370527,
      "loss": 0.8281,
      "num_input_tokens_seen": 25256424,
      "step": 43520
    },
    {
      "epoch": 6.482722669049747,
      "grad_norm": 0.01092529296875,
      "learning_rate": 0.025638812148836263,
      "loss": 0.8064,
      "num_input_tokens_seen": 25259432,
      "step": 43525
    },
    {
      "epoch": 6.483467381590706,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.025637437645570083,
      "loss": 0.807,
      "num_input_tokens_seen": 25262152,
      "step": 43530
    },
    {
      "epoch": 6.484212094131665,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.025636062962595203,
      "loss": 0.7716,
      "num_input_tokens_seen": 25264936,
      "step": 43535
    },
    {
      "epoch": 6.484956806672624,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.02563468809993485,
      "loss": 0.7905,
      "num_input_tokens_seen": 25267656,
      "step": 43540
    },
    {
      "epoch": 6.485701519213584,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.025633313057612253,
      "loss": 0.7989,
      "num_input_tokens_seen": 25270536,
      "step": 43545
    },
    {
      "epoch": 6.486446231754543,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02563193783565064,
      "loss": 0.789,
      "num_input_tokens_seen": 25273448,
      "step": 43550
    },
    {
      "epoch": 6.487190944295502,
      "grad_norm": 0.02734375,
      "learning_rate": 0.025630562434073238,
      "loss": 0.8368,
      "num_input_tokens_seen": 25276200,
      "step": 43555
    },
    {
      "epoch": 6.487935656836461,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02562918685290329,
      "loss": 0.8144,
      "num_input_tokens_seen": 25279048,
      "step": 43560
    },
    {
      "epoch": 6.488680369377421,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.025627811092164034,
      "loss": 0.8167,
      "num_input_tokens_seen": 25281832,
      "step": 43565
    },
    {
      "epoch": 6.489425081918379,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02562643515187871,
      "loss": 0.8034,
      "num_input_tokens_seen": 25284872,
      "step": 43570
    },
    {
      "epoch": 6.490169794459339,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.025625059032070567,
      "loss": 0.8124,
      "num_input_tokens_seen": 25288072,
      "step": 43575
    },
    {
      "epoch": 6.490914507000298,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.025623682732762845,
      "loss": 0.7895,
      "num_input_tokens_seen": 25290984,
      "step": 43580
    },
    {
      "epoch": 6.4916592195412575,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.025622306253978804,
      "loss": 0.7978,
      "num_input_tokens_seen": 25293704,
      "step": 43585
    },
    {
      "epoch": 6.492403932082216,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.0256209295957417,
      "loss": 0.8107,
      "num_input_tokens_seen": 25296904,
      "step": 43590
    },
    {
      "epoch": 6.493148644623176,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.025619552758074777,
      "loss": 0.8126,
      "num_input_tokens_seen": 25299816,
      "step": 43595
    },
    {
      "epoch": 6.493893357164135,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.025618175741001305,
      "loss": 0.7803,
      "num_input_tokens_seen": 25302504,
      "step": 43600
    },
    {
      "epoch": 6.494638069705093,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02561679854454455,
      "loss": 0.7987,
      "num_input_tokens_seen": 25305384,
      "step": 43605
    },
    {
      "epoch": 6.495382782246053,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02561542116872777,
      "loss": 0.7871,
      "num_input_tokens_seen": 25308488,
      "step": 43610
    },
    {
      "epoch": 6.496127494787013,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02561404361357424,
      "loss": 0.8244,
      "num_input_tokens_seen": 25311112,
      "step": 43615
    },
    {
      "epoch": 6.496872207327971,
      "grad_norm": 0.01263427734375,
      "learning_rate": 0.025612665879107228,
      "loss": 0.7868,
      "num_input_tokens_seen": 25313832,
      "step": 43620
    },
    {
      "epoch": 6.49761691986893,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.025611287965350015,
      "loss": 0.7925,
      "num_input_tokens_seen": 25316808,
      "step": 43625
    },
    {
      "epoch": 6.49836163240989,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02560990987232588,
      "loss": 0.8078,
      "num_input_tokens_seen": 25319752,
      "step": 43630
    },
    {
      "epoch": 6.499106344950849,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.025608531600058097,
      "loss": 0.8144,
      "num_input_tokens_seen": 25322632,
      "step": 43635
    },
    {
      "epoch": 6.499851057491808,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.025607153148569956,
      "loss": 0.8051,
      "num_input_tokens_seen": 25325448,
      "step": 43640
    },
    {
      "epoch": 6.500595770032767,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.025605774517884736,
      "loss": 0.7959,
      "num_input_tokens_seen": 25328104,
      "step": 43645
    },
    {
      "epoch": 6.501340482573727,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.025604395708025744,
      "loss": 0.8029,
      "num_input_tokens_seen": 25331304,
      "step": 43650
    },
    {
      "epoch": 6.502085195114685,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.025603016719016262,
      "loss": 0.7969,
      "num_input_tokens_seen": 25334216,
      "step": 43655
    },
    {
      "epoch": 6.502829907655645,
      "grad_norm": 0.013671875,
      "learning_rate": 0.02560163755087959,
      "loss": 0.8298,
      "num_input_tokens_seen": 25337064,
      "step": 43660
    },
    {
      "epoch": 6.503574620196604,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.025600258203639023,
      "loss": 0.8093,
      "num_input_tokens_seen": 25340296,
      "step": 43665
    },
    {
      "epoch": 6.5043193327375635,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02559887867731787,
      "loss": 0.8088,
      "num_input_tokens_seen": 25343496,
      "step": 43670
    },
    {
      "epoch": 6.505064045278522,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02559749897193943,
      "loss": 0.8054,
      "num_input_tokens_seen": 25346312,
      "step": 43675
    },
    {
      "epoch": 6.505808757819482,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02559611908752702,
      "loss": 0.7835,
      "num_input_tokens_seen": 25349192,
      "step": 43680
    },
    {
      "epoch": 6.506553470360441,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.025594739024103946,
      "loss": 0.8187,
      "num_input_tokens_seen": 25352264,
      "step": 43685
    },
    {
      "epoch": 6.5072981829014,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.025593358781693526,
      "loss": 0.7984,
      "num_input_tokens_seen": 25355656,
      "step": 43690
    },
    {
      "epoch": 6.508042895442359,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02559197836031907,
      "loss": 0.7945,
      "num_input_tokens_seen": 25358440,
      "step": 43695
    },
    {
      "epoch": 6.508787607983319,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.025590597760003914,
      "loss": 0.781,
      "num_input_tokens_seen": 25361160,
      "step": 43700
    },
    {
      "epoch": 6.509532320524277,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02558921698077137,
      "loss": 0.8085,
      "num_input_tokens_seen": 25364200,
      "step": 43705
    },
    {
      "epoch": 6.510277033065237,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.025587836022644764,
      "loss": 0.8075,
      "num_input_tokens_seen": 25367016,
      "step": 43710
    },
    {
      "epoch": 6.511021745606196,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.02558645488564743,
      "loss": 0.8099,
      "num_input_tokens_seen": 25369864,
      "step": 43715
    },
    {
      "epoch": 6.5117664581471555,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.0255850735698027,
      "loss": 0.7813,
      "num_input_tokens_seen": 25372904,
      "step": 43720
    },
    {
      "epoch": 6.512511170688114,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.025583692075133915,
      "loss": 0.8051,
      "num_input_tokens_seen": 25375784,
      "step": 43725
    },
    {
      "epoch": 6.513255883229074,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.0255823104016644,
      "loss": 0.7904,
      "num_input_tokens_seen": 25378664,
      "step": 43730
    },
    {
      "epoch": 6.514000595770033,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.025580928549417514,
      "loss": 0.8201,
      "num_input_tokens_seen": 25381448,
      "step": 43735
    },
    {
      "epoch": 6.514745308310992,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02557954651841659,
      "loss": 0.8236,
      "num_input_tokens_seen": 25384296,
      "step": 43740
    },
    {
      "epoch": 6.515490020851951,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025578164308684984,
      "loss": 0.8044,
      "num_input_tokens_seen": 25387272,
      "step": 43745
    },
    {
      "epoch": 6.516234733392911,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02557678192024604,
      "loss": 0.8059,
      "num_input_tokens_seen": 25390696,
      "step": 43750
    },
    {
      "epoch": 6.5169794459338695,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.025575399353123116,
      "loss": 0.8,
      "num_input_tokens_seen": 25393640,
      "step": 43755
    },
    {
      "epoch": 6.517724158474829,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02557401660733957,
      "loss": 0.826,
      "num_input_tokens_seen": 25396552,
      "step": 43760
    },
    {
      "epoch": 6.518468871015788,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.025572633682918757,
      "loss": 0.7836,
      "num_input_tokens_seen": 25399560,
      "step": 43765
    },
    {
      "epoch": 6.519213583556747,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.025571250579884047,
      "loss": 0.806,
      "num_input_tokens_seen": 25402568,
      "step": 43770
    },
    {
      "epoch": 6.519958296097706,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.0255698672982588,
      "loss": 0.8103,
      "num_input_tokens_seen": 25405576,
      "step": 43775
    },
    {
      "epoch": 6.520703008638666,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.02556848383806639,
      "loss": 0.8126,
      "num_input_tokens_seen": 25408520,
      "step": 43780
    },
    {
      "epoch": 6.521447721179625,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.025567100199330187,
      "loss": 0.7819,
      "num_input_tokens_seen": 25411176,
      "step": 43785
    },
    {
      "epoch": 6.522192433720583,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.025565716382073563,
      "loss": 0.8064,
      "num_input_tokens_seen": 25413992,
      "step": 43790
    },
    {
      "epoch": 6.522937146261543,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025564332386319902,
      "loss": 0.7956,
      "num_input_tokens_seen": 25417480,
      "step": 43795
    },
    {
      "epoch": 6.523681858802503,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.025562948212092578,
      "loss": 0.7945,
      "num_input_tokens_seen": 25420264,
      "step": 43800
    },
    {
      "epoch": 6.5244265713434615,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.025561563859414983,
      "loss": 0.8041,
      "num_input_tokens_seen": 25423464,
      "step": 43805
    },
    {
      "epoch": 6.52517128388442,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.025560179328310503,
      "loss": 0.7867,
      "num_input_tokens_seen": 25426120,
      "step": 43810
    },
    {
      "epoch": 6.52591599642538,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.025558794618802522,
      "loss": 0.8009,
      "num_input_tokens_seen": 25429192,
      "step": 43815
    },
    {
      "epoch": 6.526660708966339,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.02555740973091444,
      "loss": 0.7901,
      "num_input_tokens_seen": 25431944,
      "step": 43820
    },
    {
      "epoch": 6.527405421507298,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.025556024664669653,
      "loss": 0.8089,
      "num_input_tokens_seen": 25434728,
      "step": 43825
    },
    {
      "epoch": 6.528150134048257,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.025554639420091552,
      "loss": 0.823,
      "num_input_tokens_seen": 25437640,
      "step": 43830
    },
    {
      "epoch": 6.528894846589217,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.025553253997203554,
      "loss": 0.8096,
      "num_input_tokens_seen": 25440520,
      "step": 43835
    },
    {
      "epoch": 6.5296395591301755,
      "grad_norm": 0.0234375,
      "learning_rate": 0.025551868396029045,
      "loss": 0.8067,
      "num_input_tokens_seen": 25443336,
      "step": 43840
    },
    {
      "epoch": 6.530384271671135,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.025550482616591447,
      "loss": 0.8125,
      "num_input_tokens_seen": 25446376,
      "step": 43845
    },
    {
      "epoch": 6.531128984212094,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.025549096658914174,
      "loss": 0.7977,
      "num_input_tokens_seen": 25449224,
      "step": 43850
    },
    {
      "epoch": 6.5318736967530535,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.025547710523020628,
      "loss": 0.7814,
      "num_input_tokens_seen": 25452072,
      "step": 43855
    },
    {
      "epoch": 6.532618409294012,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.025546324208934237,
      "loss": 0.7836,
      "num_input_tokens_seen": 25454792,
      "step": 43860
    },
    {
      "epoch": 6.533363121834972,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.025544937716678418,
      "loss": 0.8103,
      "num_input_tokens_seen": 25457736,
      "step": 43865
    },
    {
      "epoch": 6.534107834375931,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.025543551046276588,
      "loss": 0.7906,
      "num_input_tokens_seen": 25460584,
      "step": 43870
    },
    {
      "epoch": 6.53485254691689,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.025542164197752185,
      "loss": 0.7923,
      "num_input_tokens_seen": 25463272,
      "step": 43875
    },
    {
      "epoch": 6.535597259457849,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.02554077717112863,
      "loss": 0.8019,
      "num_input_tokens_seen": 25465960,
      "step": 43880
    },
    {
      "epoch": 6.536341971998809,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.025539389966429356,
      "loss": 0.7843,
      "num_input_tokens_seen": 25468904,
      "step": 43885
    },
    {
      "epoch": 6.5370866845397675,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.0255380025836778,
      "loss": 0.807,
      "num_input_tokens_seen": 25471848,
      "step": 43890
    },
    {
      "epoch": 6.537831397080727,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.0255366150228974,
      "loss": 0.8287,
      "num_input_tokens_seen": 25474696,
      "step": 43895
    },
    {
      "epoch": 6.538576109621686,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.025535227284111603,
      "loss": 0.79,
      "num_input_tokens_seen": 25477608,
      "step": 43900
    },
    {
      "epoch": 6.5393208221626455,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02553383936734385,
      "loss": 0.7845,
      "num_input_tokens_seen": 25480808,
      "step": 43905
    },
    {
      "epoch": 6.540065534703604,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.025532451272617582,
      "loss": 0.8072,
      "num_input_tokens_seen": 25483784,
      "step": 43910
    },
    {
      "epoch": 6.540810247244564,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.025531062999956257,
      "loss": 0.782,
      "num_input_tokens_seen": 25486856,
      "step": 43915
    },
    {
      "epoch": 6.541554959785523,
      "grad_norm": 0.010498046875,
      "learning_rate": 0.025529674549383322,
      "loss": 0.7967,
      "num_input_tokens_seen": 25489544,
      "step": 43920
    },
    {
      "epoch": 6.542299672326482,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.025528285920922242,
      "loss": 0.7877,
      "num_input_tokens_seen": 25492488,
      "step": 43925
    },
    {
      "epoch": 6.543044384867441,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.025526897114596468,
      "loss": 0.8088,
      "num_input_tokens_seen": 25495400,
      "step": 43930
    },
    {
      "epoch": 6.5437890974084,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02552550813042947,
      "loss": 0.8228,
      "num_input_tokens_seen": 25498344,
      "step": 43935
    },
    {
      "epoch": 6.5445338099493595,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.025524118968444708,
      "loss": 0.79,
      "num_input_tokens_seen": 25501128,
      "step": 43940
    },
    {
      "epoch": 6.545278522490319,
      "grad_norm": 0.01422119140625,
      "learning_rate": 0.02552272962866565,
      "loss": 0.8173,
      "num_input_tokens_seen": 25504168,
      "step": 43945
    },
    {
      "epoch": 6.546023235031278,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.025521340111115776,
      "loss": 0.7863,
      "num_input_tokens_seen": 25506952,
      "step": 43950
    },
    {
      "epoch": 6.546767947572237,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.025519950415818553,
      "loss": 0.8176,
      "num_input_tokens_seen": 25509768,
      "step": 43955
    },
    {
      "epoch": 6.547512660113196,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.025518560542797452,
      "loss": 0.8199,
      "num_input_tokens_seen": 25512936,
      "step": 43960
    },
    {
      "epoch": 6.548257372654156,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02551717049207597,
      "loss": 0.8153,
      "num_input_tokens_seen": 25515944,
      "step": 43965
    },
    {
      "epoch": 6.549002085195115,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.025515780263677575,
      "loss": 0.8142,
      "num_input_tokens_seen": 25518888,
      "step": 43970
    },
    {
      "epoch": 6.5497467977360735,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.025514389857625764,
      "loss": 0.8004,
      "num_input_tokens_seen": 25521544,
      "step": 43975
    },
    {
      "epoch": 6.550491510277033,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02551299927394402,
      "loss": 0.7991,
      "num_input_tokens_seen": 25524424,
      "step": 43980
    },
    {
      "epoch": 6.551236222817992,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.025511608512655843,
      "loss": 0.8083,
      "num_input_tokens_seen": 25527208,
      "step": 43985
    },
    {
      "epoch": 6.5519809353589515,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.025510217573784723,
      "loss": 0.8153,
      "num_input_tokens_seen": 25530088,
      "step": 43990
    },
    {
      "epoch": 6.55272564789991,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.025508826457354153,
      "loss": 0.8148,
      "num_input_tokens_seen": 25532744,
      "step": 43995
    },
    {
      "epoch": 6.55347036044087,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.025507435163387646,
      "loss": 0.8058,
      "num_input_tokens_seen": 25535496,
      "step": 44000
    },
    {
      "epoch": 6.554215072981829,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.0255060436919087,
      "loss": 0.8194,
      "num_input_tokens_seen": 25538600,
      "step": 44005
    },
    {
      "epoch": 6.554959785522788,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.025504652042940824,
      "loss": 0.8256,
      "num_input_tokens_seen": 25541384,
      "step": 44010
    },
    {
      "epoch": 6.555704498063747,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.025503260216507527,
      "loss": 0.8006,
      "num_input_tokens_seen": 25544168,
      "step": 44015
    },
    {
      "epoch": 6.556449210604707,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.025501868212632322,
      "loss": 0.8041,
      "num_input_tokens_seen": 25547048,
      "step": 44020
    },
    {
      "epoch": 6.5571939231456655,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.025500476031338728,
      "loss": 0.8072,
      "num_input_tokens_seen": 25549768,
      "step": 44025
    },
    {
      "epoch": 6.557938635686625,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02549908367265027,
      "loss": 0.8178,
      "num_input_tokens_seen": 25553128,
      "step": 44030
    },
    {
      "epoch": 6.558683348227584,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02549769113659046,
      "loss": 0.797,
      "num_input_tokens_seen": 25555976,
      "step": 44035
    },
    {
      "epoch": 6.559428060768544,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.025496298423182826,
      "loss": 0.7895,
      "num_input_tokens_seen": 25558728,
      "step": 44040
    },
    {
      "epoch": 6.560172773309502,
      "grad_norm": 0.0234375,
      "learning_rate": 0.025494905532450898,
      "loss": 0.8145,
      "num_input_tokens_seen": 25561608,
      "step": 44045
    },
    {
      "epoch": 6.560917485850462,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.02549351246441821,
      "loss": 0.8214,
      "num_input_tokens_seen": 25564616,
      "step": 44050
    },
    {
      "epoch": 6.561662198391421,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.025492119219108295,
      "loss": 0.7984,
      "num_input_tokens_seen": 25567592,
      "step": 44055
    },
    {
      "epoch": 6.56240691093238,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.025490725796544683,
      "loss": 0.7929,
      "num_input_tokens_seen": 25570376,
      "step": 44060
    },
    {
      "epoch": 6.563151623473339,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.025489332196750928,
      "loss": 0.7838,
      "num_input_tokens_seen": 25573352,
      "step": 44065
    },
    {
      "epoch": 6.563896336014299,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.025487938419750562,
      "loss": 0.8107,
      "num_input_tokens_seen": 25576232,
      "step": 44070
    },
    {
      "epoch": 6.5646410485552575,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02548654446556714,
      "loss": 0.7882,
      "num_input_tokens_seen": 25579016,
      "step": 44075
    },
    {
      "epoch": 6.565385761096217,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.025485150334224202,
      "loss": 0.7933,
      "num_input_tokens_seen": 25581896,
      "step": 44080
    },
    {
      "epoch": 6.566130473637176,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.025483756025745312,
      "loss": 0.8115,
      "num_input_tokens_seen": 25584616,
      "step": 44085
    },
    {
      "epoch": 6.566875186178136,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02548236154015401,
      "loss": 0.8241,
      "num_input_tokens_seen": 25587688,
      "step": 44090
    },
    {
      "epoch": 6.567619898719094,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.025480966877473872,
      "loss": 0.8102,
      "num_input_tokens_seen": 25590760,
      "step": 44095
    },
    {
      "epoch": 6.568364611260054,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02547957203772845,
      "loss": 0.7912,
      "num_input_tokens_seen": 25593448,
      "step": 44100
    },
    {
      "epoch": 6.569109323801013,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.025478177020941305,
      "loss": 0.8138,
      "num_input_tokens_seen": 25596264,
      "step": 44105
    },
    {
      "epoch": 6.569854036341972,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.025476781827136016,
      "loss": 0.791,
      "num_input_tokens_seen": 25599080,
      "step": 44110
    },
    {
      "epoch": 6.570598748882931,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02547538645633614,
      "loss": 0.8058,
      "num_input_tokens_seen": 25602088,
      "step": 44115
    },
    {
      "epoch": 6.57134346142389,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02547399090856526,
      "loss": 0.7994,
      "num_input_tokens_seen": 25604968,
      "step": 44120
    },
    {
      "epoch": 6.57208817396485,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.025472595183846945,
      "loss": 0.8125,
      "num_input_tokens_seen": 25607912,
      "step": 44125
    },
    {
      "epoch": 6.572832886505809,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02547119928220478,
      "loss": 0.7816,
      "num_input_tokens_seen": 25611144,
      "step": 44130
    },
    {
      "epoch": 6.573577599046768,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.025469803203662347,
      "loss": 0.8039,
      "num_input_tokens_seen": 25613992,
      "step": 44135
    },
    {
      "epoch": 6.574322311587727,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.02546840694824323,
      "loss": 0.7885,
      "num_input_tokens_seen": 25616776,
      "step": 44140
    },
    {
      "epoch": 6.575067024128686,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.025467010515971014,
      "loss": 0.7945,
      "num_input_tokens_seen": 25619592,
      "step": 44145
    },
    {
      "epoch": 6.575811736669645,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.025465613906869296,
      "loss": 0.7865,
      "num_input_tokens_seen": 25622440,
      "step": 44150
    },
    {
      "epoch": 6.576556449210605,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.02546421712096167,
      "loss": 0.8172,
      "num_input_tokens_seen": 25625672,
      "step": 44155
    },
    {
      "epoch": 6.5773011617515635,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.025462820158271723,
      "loss": 0.7939,
      "num_input_tokens_seen": 25628616,
      "step": 44160
    },
    {
      "epoch": 6.578045874292523,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02546142301882307,
      "loss": 0.8424,
      "num_input_tokens_seen": 25631432,
      "step": 44165
    },
    {
      "epoch": 6.578790586833482,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.025460025702639305,
      "loss": 0.8023,
      "num_input_tokens_seen": 25634728,
      "step": 44170
    },
    {
      "epoch": 6.579535299374442,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.02545862820974404,
      "loss": 0.7983,
      "num_input_tokens_seen": 25638024,
      "step": 44175
    },
    {
      "epoch": 6.5802800119154,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.025457230540160882,
      "loss": 0.7964,
      "num_input_tokens_seen": 25641000,
      "step": 44180
    },
    {
      "epoch": 6.58102472445636,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.025455832693913435,
      "loss": 0.7935,
      "num_input_tokens_seen": 25643848,
      "step": 44185
    },
    {
      "epoch": 6.581769436997319,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.025454434671025327,
      "loss": 0.7816,
      "num_input_tokens_seen": 25646472,
      "step": 44190
    },
    {
      "epoch": 6.582514149538278,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.025453036471520172,
      "loss": 0.7994,
      "num_input_tokens_seen": 25649384,
      "step": 44195
    },
    {
      "epoch": 6.583258862079237,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02545163809542159,
      "loss": 0.8024,
      "num_input_tokens_seen": 25652520,
      "step": 44200
    },
    {
      "epoch": 6.584003574620197,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.0254502395427532,
      "loss": 0.7967,
      "num_input_tokens_seen": 25655560,
      "step": 44205
    },
    {
      "epoch": 6.584748287161156,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.025448840813538636,
      "loss": 0.8056,
      "num_input_tokens_seen": 25658632,
      "step": 44210
    },
    {
      "epoch": 6.585492999702115,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025447441907801526,
      "loss": 0.7882,
      "num_input_tokens_seen": 25661864,
      "step": 44215
    },
    {
      "epoch": 6.586237712243074,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.0254460428255655,
      "loss": 0.7962,
      "num_input_tokens_seen": 25665096,
      "step": 44220
    },
    {
      "epoch": 6.586982424784034,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.025444643566854202,
      "loss": 0.806,
      "num_input_tokens_seen": 25668008,
      "step": 44225
    },
    {
      "epoch": 6.587727137324992,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02544324413169127,
      "loss": 0.8054,
      "num_input_tokens_seen": 25670728,
      "step": 44230
    },
    {
      "epoch": 6.588471849865952,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.02544184452010034,
      "loss": 0.8091,
      "num_input_tokens_seen": 25673672,
      "step": 44235
    },
    {
      "epoch": 6.589216562406911,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025440444732105057,
      "loss": 0.7932,
      "num_input_tokens_seen": 25676328,
      "step": 44240
    },
    {
      "epoch": 6.58996127494787,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.025439044767729072,
      "loss": 0.8162,
      "num_input_tokens_seen": 25679144,
      "step": 44245
    },
    {
      "epoch": 6.590705987488829,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.025437644626996034,
      "loss": 0.794,
      "num_input_tokens_seen": 25681896,
      "step": 44250
    },
    {
      "epoch": 6.591450700029789,
      "grad_norm": 0.0126953125,
      "learning_rate": 0.0254362443099296,
      "loss": 0.8334,
      "num_input_tokens_seen": 25684584,
      "step": 44255
    },
    {
      "epoch": 6.592195412570748,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.025434843816553424,
      "loss": 0.8001,
      "num_input_tokens_seen": 25687528,
      "step": 44260
    },
    {
      "epoch": 6.592940125111707,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02543344314689117,
      "loss": 0.8011,
      "num_input_tokens_seen": 25690536,
      "step": 44265
    },
    {
      "epoch": 6.593684837652666,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.0254320423009665,
      "loss": 0.8015,
      "num_input_tokens_seen": 25693544,
      "step": 44270
    },
    {
      "epoch": 6.594429550193626,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.025430641278803068,
      "loss": 0.8115,
      "num_input_tokens_seen": 25696104,
      "step": 44275
    },
    {
      "epoch": 6.595174262734584,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.025429240080424565,
      "loss": 0.8002,
      "num_input_tokens_seen": 25699080,
      "step": 44280
    },
    {
      "epoch": 6.595918975275543,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.025427838705854643,
      "loss": 0.8016,
      "num_input_tokens_seen": 25701896,
      "step": 44285
    },
    {
      "epoch": 6.596663687816503,
      "grad_norm": 0.01953125,
      "learning_rate": 0.025426437155116985,
      "loss": 0.7862,
      "num_input_tokens_seen": 25704968,
      "step": 44290
    },
    {
      "epoch": 6.5974084003574625,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02542503542823527,
      "loss": 0.7965,
      "num_input_tokens_seen": 25707912,
      "step": 44295
    },
    {
      "epoch": 6.598153112898421,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.025423633525233177,
      "loss": 0.8215,
      "num_input_tokens_seen": 25710824,
      "step": 44300
    },
    {
      "epoch": 6.59889782543938,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.025422231446134386,
      "loss": 0.7909,
      "num_input_tokens_seen": 25714408,
      "step": 44305
    },
    {
      "epoch": 6.59964253798034,
      "grad_norm": 0.021484375,
      "learning_rate": 0.025420829190962595,
      "loss": 0.8073,
      "num_input_tokens_seen": 25717064,
      "step": 44310
    },
    {
      "epoch": 6.600387250521299,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.025419426759741483,
      "loss": 0.8243,
      "num_input_tokens_seen": 25719816,
      "step": 44315
    },
    {
      "epoch": 6.601131963062258,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.02541802415249474,
      "loss": 0.7957,
      "num_input_tokens_seen": 25722760,
      "step": 44320
    },
    {
      "epoch": 6.601876675603217,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.025416621369246075,
      "loss": 0.8311,
      "num_input_tokens_seen": 25725960,
      "step": 44325
    },
    {
      "epoch": 6.602621388144176,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.025415218410019175,
      "loss": 0.8018,
      "num_input_tokens_seen": 25729096,
      "step": 44330
    },
    {
      "epoch": 6.603366100685135,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.025413815274837746,
      "loss": 0.8166,
      "num_input_tokens_seen": 25731816,
      "step": 44335
    },
    {
      "epoch": 6.604110813226095,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02541241196372549,
      "loss": 0.8146,
      "num_input_tokens_seen": 25734824,
      "step": 44340
    },
    {
      "epoch": 6.604855525767054,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02541100847670612,
      "loss": 0.7833,
      "num_input_tokens_seen": 25737704,
      "step": 44345
    },
    {
      "epoch": 6.605600238308013,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.025409604813803346,
      "loss": 0.7994,
      "num_input_tokens_seen": 25740488,
      "step": 44350
    },
    {
      "epoch": 6.606344950848972,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025408200975040873,
      "loss": 0.7931,
      "num_input_tokens_seen": 25743144,
      "step": 44355
    },
    {
      "epoch": 6.607089663389932,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.025406796960442426,
      "loss": 0.7891,
      "num_input_tokens_seen": 25746280,
      "step": 44360
    },
    {
      "epoch": 6.60783437593089,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.025405392770031718,
      "loss": 0.7944,
      "num_input_tokens_seen": 25749096,
      "step": 44365
    },
    {
      "epoch": 6.60857908847185,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.025403988403832477,
      "loss": 0.8132,
      "num_input_tokens_seen": 25752328,
      "step": 44370
    },
    {
      "epoch": 6.609323801012809,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.025402583861868427,
      "loss": 0.8073,
      "num_input_tokens_seen": 25755112,
      "step": 44375
    },
    {
      "epoch": 6.6100685135537685,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02540117914416329,
      "loss": 0.8074,
      "num_input_tokens_seen": 25757832,
      "step": 44380
    },
    {
      "epoch": 6.610813226094727,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.025399774250740807,
      "loss": 0.8336,
      "num_input_tokens_seen": 25760904,
      "step": 44385
    },
    {
      "epoch": 6.611557938635687,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02539836918162471,
      "loss": 0.8243,
      "num_input_tokens_seen": 25763880,
      "step": 44390
    },
    {
      "epoch": 6.612302651176646,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02539696393683873,
      "loss": 0.8087,
      "num_input_tokens_seen": 25766920,
      "step": 44395
    },
    {
      "epoch": 6.613047363717605,
      "grad_norm": 0.01129150390625,
      "learning_rate": 0.025395558516406614,
      "loss": 0.8106,
      "num_input_tokens_seen": 25769800,
      "step": 44400
    },
    {
      "epoch": 6.613792076258564,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.0253941529203521,
      "loss": 0.8007,
      "num_input_tokens_seen": 25772744,
      "step": 44405
    },
    {
      "epoch": 6.614536788799524,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.025392747148698935,
      "loss": 0.7935,
      "num_input_tokens_seen": 25775688,
      "step": 44410
    },
    {
      "epoch": 6.615281501340482,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.025391341201470876,
      "loss": 0.7986,
      "num_input_tokens_seen": 25778312,
      "step": 44415
    },
    {
      "epoch": 6.616026213881442,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.025389935078691663,
      "loss": 0.8045,
      "num_input_tokens_seen": 25781000,
      "step": 44420
    },
    {
      "epoch": 6.616770926422401,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02538852878038506,
      "loss": 0.8197,
      "num_input_tokens_seen": 25783816,
      "step": 44425
    },
    {
      "epoch": 6.6175156389633605,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.025387122306574823,
      "loss": 0.8129,
      "num_input_tokens_seen": 25786760,
      "step": 44430
    },
    {
      "epoch": 6.618260351504319,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02538571565728471,
      "loss": 0.8099,
      "num_input_tokens_seen": 25789512,
      "step": 44435
    },
    {
      "epoch": 6.619005064045279,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.025384308832538484,
      "loss": 0.8036,
      "num_input_tokens_seen": 25792200,
      "step": 44440
    },
    {
      "epoch": 6.619749776586238,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.025382901832359916,
      "loss": 0.7988,
      "num_input_tokens_seen": 25795048,
      "step": 44445
    },
    {
      "epoch": 6.620494489127196,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025381494656772773,
      "loss": 0.8153,
      "num_input_tokens_seen": 25798312,
      "step": 44450
    },
    {
      "epoch": 6.621239201668156,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02538008730580083,
      "loss": 0.8061,
      "num_input_tokens_seen": 25801096,
      "step": 44455
    },
    {
      "epoch": 6.621983914209116,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.025378679779467863,
      "loss": 0.816,
      "num_input_tokens_seen": 25803912,
      "step": 44460
    },
    {
      "epoch": 6.6227286267500745,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02537727207779765,
      "loss": 0.8019,
      "num_input_tokens_seen": 25807016,
      "step": 44465
    },
    {
      "epoch": 6.623473339291033,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02537586420081397,
      "loss": 0.7835,
      "num_input_tokens_seen": 25809864,
      "step": 44470
    },
    {
      "epoch": 6.624218051831993,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.025374456148540614,
      "loss": 0.7882,
      "num_input_tokens_seen": 25812552,
      "step": 44475
    },
    {
      "epoch": 6.6249627643729525,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02537304792100136,
      "loss": 0.8067,
      "num_input_tokens_seen": 25815432,
      "step": 44480
    },
    {
      "epoch": 6.625707476913911,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02537163951822001,
      "loss": 0.8024,
      "num_input_tokens_seen": 25818376,
      "step": 44485
    },
    {
      "epoch": 6.62645218945487,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.025370230940220343,
      "loss": 0.7895,
      "num_input_tokens_seen": 25821160,
      "step": 44490
    },
    {
      "epoch": 6.62719690199583,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02536882218702617,
      "loss": 0.8015,
      "num_input_tokens_seen": 25823976,
      "step": 44495
    },
    {
      "epoch": 6.627941614536788,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.025367413258661285,
      "loss": 0.8023,
      "num_input_tokens_seen": 25827176,
      "step": 44500
    },
    {
      "epoch": 6.628686327077748,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.025366004155149488,
      "loss": 0.805,
      "num_input_tokens_seen": 25830248,
      "step": 44505
    },
    {
      "epoch": 6.629431039618707,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02536459487651459,
      "loss": 0.807,
      "num_input_tokens_seen": 25833288,
      "step": 44510
    },
    {
      "epoch": 6.6301757521596665,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.025363185422780394,
      "loss": 0.8084,
      "num_input_tokens_seen": 25836296,
      "step": 44515
    },
    {
      "epoch": 6.630920464700625,
      "grad_norm": 0.03515625,
      "learning_rate": 0.025361775793970708,
      "loss": 0.7925,
      "num_input_tokens_seen": 25839112,
      "step": 44520
    },
    {
      "epoch": 6.631665177241585,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.025360365990109355,
      "loss": 0.7857,
      "num_input_tokens_seen": 25841992,
      "step": 44525
    },
    {
      "epoch": 6.632409889782544,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.025358956011220152,
      "loss": 0.8183,
      "num_input_tokens_seen": 25844872,
      "step": 44530
    },
    {
      "epoch": 6.633154602323503,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02535754585732691,
      "loss": 0.81,
      "num_input_tokens_seen": 25847752,
      "step": 44535
    },
    {
      "epoch": 6.633899314864462,
      "grad_norm": 0.013671875,
      "learning_rate": 0.025356135528453456,
      "loss": 0.8003,
      "num_input_tokens_seen": 25850664,
      "step": 44540
    },
    {
      "epoch": 6.634644027405422,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02535472502462362,
      "loss": 0.7964,
      "num_input_tokens_seen": 25853928,
      "step": 44545
    },
    {
      "epoch": 6.6353887399463805,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.025353314345861234,
      "loss": 0.8142,
      "num_input_tokens_seen": 25856840,
      "step": 44550
    },
    {
      "epoch": 6.63613345248734,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02535190349219012,
      "loss": 0.8123,
      "num_input_tokens_seen": 25859752,
      "step": 44555
    },
    {
      "epoch": 6.636878165028299,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.02535049246363412,
      "loss": 0.816,
      "num_input_tokens_seen": 25862888,
      "step": 44560
    },
    {
      "epoch": 6.6376228775692585,
      "grad_norm": 0.01953125,
      "learning_rate": 0.025349081260217066,
      "loss": 0.7901,
      "num_input_tokens_seen": 25865640,
      "step": 44565
    },
    {
      "epoch": 6.638367590110217,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.025347669881962805,
      "loss": 0.7988,
      "num_input_tokens_seen": 25868424,
      "step": 44570
    },
    {
      "epoch": 6.639112302651177,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.025346258328895182,
      "loss": 0.7977,
      "num_input_tokens_seen": 25871368,
      "step": 44575
    },
    {
      "epoch": 6.639857015192136,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.025344846601038033,
      "loss": 0.7889,
      "num_input_tokens_seen": 25874088,
      "step": 44580
    },
    {
      "epoch": 6.640601727733095,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.025343434698415216,
      "loss": 0.7794,
      "num_input_tokens_seen": 25877000,
      "step": 44585
    },
    {
      "epoch": 6.641346440274054,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.025342022621050585,
      "loss": 0.8133,
      "num_input_tokens_seen": 25879912,
      "step": 44590
    },
    {
      "epoch": 6.642091152815014,
      "grad_norm": 0.0234375,
      "learning_rate": 0.025340610368967995,
      "loss": 0.7915,
      "num_input_tokens_seen": 25883176,
      "step": 44595
    },
    {
      "epoch": 6.6428358653559725,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.025339197942191298,
      "loss": 0.7972,
      "num_input_tokens_seen": 25886216,
      "step": 44600
    },
    {
      "epoch": 6.643580577896932,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02533778534074436,
      "loss": 0.7978,
      "num_input_tokens_seen": 25888904,
      "step": 44605
    },
    {
      "epoch": 6.644325290437891,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.02533637256465105,
      "loss": 0.7991,
      "num_input_tokens_seen": 25891816,
      "step": 44610
    },
    {
      "epoch": 6.6450700029788505,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.025334959613935226,
      "loss": 0.8176,
      "num_input_tokens_seen": 25894984,
      "step": 44615
    },
    {
      "epoch": 6.645814715519809,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.025333546488620764,
      "loss": 0.8024,
      "num_input_tokens_seen": 25897832,
      "step": 44620
    },
    {
      "epoch": 6.646559428060769,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.025332133188731537,
      "loss": 0.8304,
      "num_input_tokens_seen": 25900712,
      "step": 44625
    },
    {
      "epoch": 6.647304140601728,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.025330719714291423,
      "loss": 0.8239,
      "num_input_tokens_seen": 25903432,
      "step": 44630
    },
    {
      "epoch": 6.6480488531426865,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.025329306065324297,
      "loss": 0.8244,
      "num_input_tokens_seen": 25906216,
      "step": 44635
    },
    {
      "epoch": 6.648793565683646,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02532789224185405,
      "loss": 0.8162,
      "num_input_tokens_seen": 25908872,
      "step": 44640
    },
    {
      "epoch": 6.649538278224606,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.025326478243904552,
      "loss": 0.7986,
      "num_input_tokens_seen": 25912008,
      "step": 44645
    },
    {
      "epoch": 6.6502829907655645,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0253250640714997,
      "loss": 0.7898,
      "num_input_tokens_seen": 25914856,
      "step": 44650
    },
    {
      "epoch": 6.651027703306523,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.025323649724663385,
      "loss": 0.7992,
      "num_input_tokens_seen": 25918056,
      "step": 44655
    },
    {
      "epoch": 6.651772415847483,
      "grad_norm": 0.021484375,
      "learning_rate": 0.025322235203419502,
      "loss": 0.8031,
      "num_input_tokens_seen": 25920936,
      "step": 44660
    },
    {
      "epoch": 6.652517128388443,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.025320820507791948,
      "loss": 0.8091,
      "num_input_tokens_seen": 25923816,
      "step": 44665
    },
    {
      "epoch": 6.653261840929401,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.025319405637804618,
      "loss": 0.8054,
      "num_input_tokens_seen": 25926696,
      "step": 44670
    },
    {
      "epoch": 6.65400655347036,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.025317990593481422,
      "loss": 0.808,
      "num_input_tokens_seen": 25929736,
      "step": 44675
    },
    {
      "epoch": 6.65475126601132,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.025316575374846256,
      "loss": 0.8037,
      "num_input_tokens_seen": 25932744,
      "step": 44680
    },
    {
      "epoch": 6.6554959785522785,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.025315159981923036,
      "loss": 0.8051,
      "num_input_tokens_seen": 25935752,
      "step": 44685
    },
    {
      "epoch": 6.656240691093238,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.025313744414735675,
      "loss": 0.7985,
      "num_input_tokens_seen": 25938440,
      "step": 44690
    },
    {
      "epoch": 6.656985403634197,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.025312328673308085,
      "loss": 0.8044,
      "num_input_tokens_seen": 25941448,
      "step": 44695
    },
    {
      "epoch": 6.6577301161751565,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02531091275766418,
      "loss": 0.823,
      "num_input_tokens_seen": 25944136,
      "step": 44700
    },
    {
      "epoch": 6.658474828716115,
      "grad_norm": 0.0107421875,
      "learning_rate": 0.02530949666782788,
      "loss": 0.7818,
      "num_input_tokens_seen": 25946984,
      "step": 44705
    },
    {
      "epoch": 6.659219541257075,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.025308080403823117,
      "loss": 0.8009,
      "num_input_tokens_seen": 25950280,
      "step": 44710
    },
    {
      "epoch": 6.659964253798034,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.025306663965673815,
      "loss": 0.8193,
      "num_input_tokens_seen": 25953288,
      "step": 44715
    },
    {
      "epoch": 6.660708966338993,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.025305247353403896,
      "loss": 0.8207,
      "num_input_tokens_seen": 25956072,
      "step": 44720
    },
    {
      "epoch": 6.661453678879952,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.025303830567037295,
      "loss": 0.7878,
      "num_input_tokens_seen": 25958792,
      "step": 44725
    },
    {
      "epoch": 6.662198391420912,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.025302413606597955,
      "loss": 0.8131,
      "num_input_tokens_seen": 25961800,
      "step": 44730
    },
    {
      "epoch": 6.6629431039618705,
      "grad_norm": 0.0115966796875,
      "learning_rate": 0.025300996472109805,
      "loss": 0.8172,
      "num_input_tokens_seen": 25964744,
      "step": 44735
    },
    {
      "epoch": 6.66368781650283,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.025299579163596785,
      "loss": 0.787,
      "num_input_tokens_seen": 25967496,
      "step": 44740
    },
    {
      "epoch": 6.664432529043789,
      "grad_norm": 0.0120849609375,
      "learning_rate": 0.025298161681082846,
      "loss": 0.7998,
      "num_input_tokens_seen": 25970728,
      "step": 44745
    },
    {
      "epoch": 6.665177241584749,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.025296744024591936,
      "loss": 0.8025,
      "num_input_tokens_seen": 25973608,
      "step": 44750
    },
    {
      "epoch": 6.665921954125707,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.025295326194147996,
      "loss": 0.7916,
      "num_input_tokens_seen": 25976616,
      "step": 44755
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.025293908189774984,
      "loss": 0.8102,
      "num_input_tokens_seen": 25979656,
      "step": 44760
    },
    {
      "epoch": 6.667411379207626,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02529249001149686,
      "loss": 0.8122,
      "num_input_tokens_seen": 25982504,
      "step": 44765
    },
    {
      "epoch": 6.668156091748585,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.025291071659337576,
      "loss": 0.7935,
      "num_input_tokens_seen": 25986152,
      "step": 44770
    },
    {
      "epoch": 6.668900804289544,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.025289653133321092,
      "loss": 0.7871,
      "num_input_tokens_seen": 25989672,
      "step": 44775
    },
    {
      "epoch": 6.669645516830504,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.025288234433471377,
      "loss": 0.8049,
      "num_input_tokens_seen": 25992584,
      "step": 44780
    },
    {
      "epoch": 6.6703902293714625,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.0252868155598124,
      "loss": 0.7884,
      "num_input_tokens_seen": 25995624,
      "step": 44785
    },
    {
      "epoch": 6.671134941912422,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02528539651236813,
      "loss": 0.8262,
      "num_input_tokens_seen": 25998536,
      "step": 44790
    },
    {
      "epoch": 6.671879654453381,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.025283977291162534,
      "loss": 0.8012,
      "num_input_tokens_seen": 26001288,
      "step": 44795
    },
    {
      "epoch": 6.67262436699434,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.0252825578962196,
      "loss": 0.8138,
      "num_input_tokens_seen": 26004232,
      "step": 44800
    },
    {
      "epoch": 6.673369079535299,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.0252811383275633,
      "loss": 0.8109,
      "num_input_tokens_seen": 26007272,
      "step": 44805
    },
    {
      "epoch": 6.674113792076259,
      "grad_norm": 0.01171875,
      "learning_rate": 0.025279718585217617,
      "loss": 0.7948,
      "num_input_tokens_seen": 26010024,
      "step": 44810
    },
    {
      "epoch": 6.674858504617218,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02527829866920653,
      "loss": 0.8038,
      "num_input_tokens_seen": 26012904,
      "step": 44815
    },
    {
      "epoch": 6.6756032171581765,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.025276878579554043,
      "loss": 0.7948,
      "num_input_tokens_seen": 26015624,
      "step": 44820
    },
    {
      "epoch": 6.676347929699136,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.025275458316284135,
      "loss": 0.8157,
      "num_input_tokens_seen": 26018600,
      "step": 44825
    },
    {
      "epoch": 6.677092642240096,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.0252740378794208,
      "loss": 0.8242,
      "num_input_tokens_seen": 26021576,
      "step": 44830
    },
    {
      "epoch": 6.677837354781055,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.025272617268988037,
      "loss": 0.7978,
      "num_input_tokens_seen": 26024456,
      "step": 44835
    },
    {
      "epoch": 6.678582067322013,
      "grad_norm": 0.013671875,
      "learning_rate": 0.025271196485009847,
      "loss": 0.7979,
      "num_input_tokens_seen": 26027400,
      "step": 44840
    },
    {
      "epoch": 6.679326779862973,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.02526977552751023,
      "loss": 0.8176,
      "num_input_tokens_seen": 26030056,
      "step": 44845
    },
    {
      "epoch": 6.680071492403932,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.025268354396513197,
      "loss": 0.7917,
      "num_input_tokens_seen": 26033224,
      "step": 44850
    },
    {
      "epoch": 6.680816204944891,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.025266933092042747,
      "loss": 0.7876,
      "num_input_tokens_seen": 26036136,
      "step": 44855
    },
    {
      "epoch": 6.68156091748585,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.0252655116141229,
      "loss": 0.8055,
      "num_input_tokens_seen": 26039368,
      "step": 44860
    },
    {
      "epoch": 6.68230563002681,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02526408996277767,
      "loss": 0.7837,
      "num_input_tokens_seen": 26042600,
      "step": 44865
    },
    {
      "epoch": 6.6830503425677685,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025262668138031068,
      "loss": 0.8172,
      "num_input_tokens_seen": 26045416,
      "step": 44870
    },
    {
      "epoch": 6.683795055108728,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.02526124613990712,
      "loss": 0.7885,
      "num_input_tokens_seen": 26048296,
      "step": 44875
    },
    {
      "epoch": 6.684539767649687,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.025259823968429847,
      "loss": 0.7781,
      "num_input_tokens_seen": 26051048,
      "step": 44880
    },
    {
      "epoch": 6.685284480190647,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02525840162362328,
      "loss": 0.7939,
      "num_input_tokens_seen": 26054120,
      "step": 44885
    },
    {
      "epoch": 6.686029192731605,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.025256979105511435,
      "loss": 0.8013,
      "num_input_tokens_seen": 26057128,
      "step": 44890
    },
    {
      "epoch": 6.686773905272565,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.025255556414118363,
      "loss": 0.8036,
      "num_input_tokens_seen": 26059880,
      "step": 44895
    },
    {
      "epoch": 6.687518617813524,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.02525413354946808,
      "loss": 0.8095,
      "num_input_tokens_seen": 26062920,
      "step": 44900
    },
    {
      "epoch": 6.688263330354483,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.025252710511584637,
      "loss": 0.813,
      "num_input_tokens_seen": 26065544,
      "step": 44905
    },
    {
      "epoch": 6.689008042895442,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02525128730049207,
      "loss": 0.8137,
      "num_input_tokens_seen": 26068168,
      "step": 44910
    },
    {
      "epoch": 6.689752755436402,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02524986391621442,
      "loss": 0.823,
      "num_input_tokens_seen": 26071048,
      "step": 44915
    },
    {
      "epoch": 6.690497467977361,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02524844035877574,
      "loss": 0.8073,
      "num_input_tokens_seen": 26074344,
      "step": 44920
    },
    {
      "epoch": 6.69124218051832,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.025247016628200076,
      "loss": 0.8124,
      "num_input_tokens_seen": 26077480,
      "step": 44925
    },
    {
      "epoch": 6.691986893059279,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.025245592724511483,
      "loss": 0.7958,
      "num_input_tokens_seen": 26080552,
      "step": 44930
    },
    {
      "epoch": 6.692731605600239,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.025244168647734006,
      "loss": 0.7913,
      "num_input_tokens_seen": 26083528,
      "step": 44935
    },
    {
      "epoch": 6.693476318141197,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.025242744397891717,
      "loss": 0.7885,
      "num_input_tokens_seen": 26086568,
      "step": 44940
    },
    {
      "epoch": 6.694221030682157,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.025241319975008673,
      "loss": 0.7913,
      "num_input_tokens_seen": 26089480,
      "step": 44945
    },
    {
      "epoch": 6.694965743223116,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.025239895379108934,
      "loss": 0.7889,
      "num_input_tokens_seen": 26092136,
      "step": 44950
    },
    {
      "epoch": 6.695710455764075,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025238470610216574,
      "loss": 0.7865,
      "num_input_tokens_seen": 26095144,
      "step": 44955
    },
    {
      "epoch": 6.696455168305034,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.025237045668355655,
      "loss": 0.8111,
      "num_input_tokens_seen": 26098184,
      "step": 44960
    },
    {
      "epoch": 6.697199880845994,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.025235620553550258,
      "loss": 0.8412,
      "num_input_tokens_seen": 26101096,
      "step": 44965
    },
    {
      "epoch": 6.697944593386953,
      "grad_norm": 0.025390625,
      "learning_rate": 0.025234195265824447,
      "loss": 0.8191,
      "num_input_tokens_seen": 26103976,
      "step": 44970
    },
    {
      "epoch": 6.698689305927912,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.025232769805202313,
      "loss": 0.8175,
      "num_input_tokens_seen": 26107016,
      "step": 44975
    },
    {
      "epoch": 6.699434018468871,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.025231344171707932,
      "loss": 0.7902,
      "num_input_tokens_seen": 26109768,
      "step": 44980
    },
    {
      "epoch": 6.70017873100983,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02522991836536539,
      "loss": 0.8268,
      "num_input_tokens_seen": 26112808,
      "step": 44985
    },
    {
      "epoch": 6.700923443550789,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02522849238619878,
      "loss": 0.8062,
      "num_input_tokens_seen": 26115688,
      "step": 44990
    },
    {
      "epoch": 6.701668156091749,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02522706623423218,
      "loss": 0.8009,
      "num_input_tokens_seen": 26118696,
      "step": 44995
    },
    {
      "epoch": 6.702412868632708,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.025225639909489696,
      "loss": 0.8009,
      "num_input_tokens_seen": 26121768,
      "step": 45000
    },
    {
      "epoch": 6.703157581173667,
      "grad_norm": 0.010498046875,
      "learning_rate": 0.025224213411995416,
      "loss": 0.7759,
      "num_input_tokens_seen": 26124360,
      "step": 45005
    },
    {
      "epoch": 6.703902293714626,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.025222786741773442,
      "loss": 0.8016,
      "num_input_tokens_seen": 26127400,
      "step": 45010
    },
    {
      "epoch": 6.704647006255585,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02522135989884787,
      "loss": 0.8391,
      "num_input_tokens_seen": 26130184,
      "step": 45015
    },
    {
      "epoch": 6.705391718796545,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.025219932883242816,
      "loss": 0.7985,
      "num_input_tokens_seen": 26133224,
      "step": 45020
    },
    {
      "epoch": 6.706136431337503,
      "grad_norm": 0.033203125,
      "learning_rate": 0.025218505694982386,
      "loss": 0.7809,
      "num_input_tokens_seen": 26136264,
      "step": 45025
    },
    {
      "epoch": 6.706881143878463,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.025217078334090683,
      "loss": 0.7902,
      "num_input_tokens_seen": 26138888,
      "step": 45030
    },
    {
      "epoch": 6.707625856419422,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02521565080059183,
      "loss": 0.8263,
      "num_input_tokens_seen": 26141640,
      "step": 45035
    },
    {
      "epoch": 6.708370568960381,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02521422309450994,
      "loss": 0.8052,
      "num_input_tokens_seen": 26144904,
      "step": 45040
    },
    {
      "epoch": 6.70911528150134,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.025212795215869128,
      "loss": 0.7736,
      "num_input_tokens_seen": 26147464,
      "step": 45045
    },
    {
      "epoch": 6.7098599940423,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.02521136716469352,
      "loss": 0.8259,
      "num_input_tokens_seen": 26150824,
      "step": 45050
    },
    {
      "epoch": 6.710604706583259,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.025209938941007246,
      "loss": 0.8141,
      "num_input_tokens_seen": 26153608,
      "step": 45055
    },
    {
      "epoch": 6.711349419124218,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02520851054483443,
      "loss": 0.8038,
      "num_input_tokens_seen": 26156552,
      "step": 45060
    },
    {
      "epoch": 6.712094131665177,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.0252070819761992,
      "loss": 0.8202,
      "num_input_tokens_seen": 26159272,
      "step": 45065
    },
    {
      "epoch": 6.712838844206137,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.02520565323512569,
      "loss": 0.812,
      "num_input_tokens_seen": 26162024,
      "step": 45070
    },
    {
      "epoch": 6.713583556747095,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.025204224321638054,
      "loss": 0.8002,
      "num_input_tokens_seen": 26164840,
      "step": 45075
    },
    {
      "epoch": 6.714328269288055,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.025202795235760415,
      "loss": 0.8203,
      "num_input_tokens_seen": 26167688,
      "step": 45080
    },
    {
      "epoch": 6.715072981829014,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.025201365977516918,
      "loss": 0.8152,
      "num_input_tokens_seen": 26170760,
      "step": 45085
    },
    {
      "epoch": 6.7158176943699734,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.025199936546931707,
      "loss": 0.7979,
      "num_input_tokens_seen": 26173896,
      "step": 45090
    },
    {
      "epoch": 6.716562406910932,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.025198506944028944,
      "loss": 0.7881,
      "num_input_tokens_seen": 26176488,
      "step": 45095
    },
    {
      "epoch": 6.717307119451892,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.025197077168832762,
      "loss": 0.7919,
      "num_input_tokens_seen": 26179432,
      "step": 45100
    },
    {
      "epoch": 6.718051831992851,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02519564722136733,
      "loss": 0.8102,
      "num_input_tokens_seen": 26182696,
      "step": 45105
    },
    {
      "epoch": 6.71879654453381,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.025194217101656802,
      "loss": 0.7983,
      "num_input_tokens_seen": 26185768,
      "step": 45110
    },
    {
      "epoch": 6.719541257074769,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.025192786809725334,
      "loss": 0.8069,
      "num_input_tokens_seen": 26189064,
      "step": 45115
    },
    {
      "epoch": 6.720285969615729,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.025191356345597094,
      "loss": 0.7971,
      "num_input_tokens_seen": 26191880,
      "step": 45120
    },
    {
      "epoch": 6.721030682156687,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02518992570929625,
      "loss": 0.8139,
      "num_input_tokens_seen": 26194728,
      "step": 45125
    },
    {
      "epoch": 6.721775394697647,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02518849490084696,
      "loss": 0.8095,
      "num_input_tokens_seen": 26197384,
      "step": 45130
    },
    {
      "epoch": 6.722520107238606,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.025187063920273405,
      "loss": 0.7979,
      "num_input_tokens_seen": 26200360,
      "step": 45135
    },
    {
      "epoch": 6.7232648197795655,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.02518563276759976,
      "loss": 0.8116,
      "num_input_tokens_seen": 26203400,
      "step": 45140
    },
    {
      "epoch": 6.724009532320524,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.025184201442850204,
      "loss": 0.8025,
      "num_input_tokens_seen": 26206152,
      "step": 45145
    },
    {
      "epoch": 6.724754244861483,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.025182769946048914,
      "loss": 0.8062,
      "num_input_tokens_seen": 26209064,
      "step": 45150
    },
    {
      "epoch": 6.725498957402443,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.025181338277220078,
      "loss": 0.7813,
      "num_input_tokens_seen": 26211816,
      "step": 45155
    },
    {
      "epoch": 6.726243669943402,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.025179906436387876,
      "loss": 0.8108,
      "num_input_tokens_seen": 26214888,
      "step": 45160
    },
    {
      "epoch": 6.726988382484361,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.025178474423576503,
      "loss": 0.8062,
      "num_input_tokens_seen": 26217672,
      "step": 45165
    },
    {
      "epoch": 6.72773309502532,
      "grad_norm": 0.01953125,
      "learning_rate": 0.025177042238810145,
      "loss": 0.7844,
      "num_input_tokens_seen": 26220840,
      "step": 45170
    },
    {
      "epoch": 6.7284778075662794,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.02517560988211301,
      "loss": 0.8159,
      "num_input_tokens_seen": 26223944,
      "step": 45175
    },
    {
      "epoch": 6.729222520107239,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.025174177353509283,
      "loss": 0.7959,
      "num_input_tokens_seen": 26226760,
      "step": 45180
    },
    {
      "epoch": 6.729967232648198,
      "grad_norm": 0.012451171875,
      "learning_rate": 0.025172744653023167,
      "loss": 0.7966,
      "num_input_tokens_seen": 26229736,
      "step": 45185
    },
    {
      "epoch": 6.730711945189157,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.025171311780678874,
      "loss": 0.7875,
      "num_input_tokens_seen": 26232552,
      "step": 45190
    },
    {
      "epoch": 6.731456657730116,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.025169878736500603,
      "loss": 0.8264,
      "num_input_tokens_seen": 26235176,
      "step": 45195
    },
    {
      "epoch": 6.732201370271075,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.025168445520512575,
      "loss": 0.8117,
      "num_input_tokens_seen": 26238216,
      "step": 45200
    },
    {
      "epoch": 6.732946082812035,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.025167012132738985,
      "loss": 0.79,
      "num_input_tokens_seen": 26240936,
      "step": 45205
    },
    {
      "epoch": 6.733690795352993,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.025165578573204065,
      "loss": 0.7875,
      "num_input_tokens_seen": 26243752,
      "step": 45210
    },
    {
      "epoch": 6.734435507893953,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.02516414484193202,
      "loss": 0.798,
      "num_input_tokens_seen": 26246696,
      "step": 45215
    },
    {
      "epoch": 6.735180220434912,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02516271093894708,
      "loss": 0.8045,
      "num_input_tokens_seen": 26249832,
      "step": 45220
    },
    {
      "epoch": 6.7359249329758715,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.025161276864273472,
      "loss": 0.8164,
      "num_input_tokens_seen": 26253032,
      "step": 45225
    },
    {
      "epoch": 6.73666964551683,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02515984261793542,
      "loss": 0.8133,
      "num_input_tokens_seen": 26255944,
      "step": 45230
    },
    {
      "epoch": 6.73741435805779,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.025158408199957145,
      "loss": 0.8248,
      "num_input_tokens_seen": 26258440,
      "step": 45235
    },
    {
      "epoch": 6.738159070598749,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.025156973610362895,
      "loss": 0.8101,
      "num_input_tokens_seen": 26261224,
      "step": 45240
    },
    {
      "epoch": 6.738903783139708,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.025155538849176897,
      "loss": 0.8028,
      "num_input_tokens_seen": 26264584,
      "step": 45245
    },
    {
      "epoch": 6.739648495680667,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02515410391642339,
      "loss": 0.7858,
      "num_input_tokens_seen": 26267208,
      "step": 45250
    },
    {
      "epoch": 6.740393208221627,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02515266881212662,
      "loss": 0.7957,
      "num_input_tokens_seen": 26270088,
      "step": 45255
    },
    {
      "epoch": 6.7411379207625854,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.025151233536310828,
      "loss": 0.8054,
      "num_input_tokens_seen": 26273000,
      "step": 45260
    },
    {
      "epoch": 6.741882633303545,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02514979808900026,
      "loss": 0.7907,
      "num_input_tokens_seen": 26275656,
      "step": 45265
    },
    {
      "epoch": 6.742627345844504,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.025148362470219173,
      "loss": 0.792,
      "num_input_tokens_seen": 26278536,
      "step": 45270
    },
    {
      "epoch": 6.7433720583854635,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.025146926679991816,
      "loss": 0.7999,
      "num_input_tokens_seen": 26281384,
      "step": 45275
    },
    {
      "epoch": 6.744116770926422,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.02514549071834244,
      "loss": 0.8197,
      "num_input_tokens_seen": 26284392,
      "step": 45280
    },
    {
      "epoch": 6.744861483467382,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.025144054585295318,
      "loss": 0.797,
      "num_input_tokens_seen": 26287080,
      "step": 45285
    },
    {
      "epoch": 6.745606196008341,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.0251426182808747,
      "loss": 0.8242,
      "num_input_tokens_seen": 26289928,
      "step": 45290
    },
    {
      "epoch": 6.7463509085493,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.025141181805104853,
      "loss": 0.7872,
      "num_input_tokens_seen": 26292776,
      "step": 45295
    },
    {
      "epoch": 6.747095621090259,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02513974515801005,
      "loss": 0.8046,
      "num_input_tokens_seen": 26295688,
      "step": 45300
    },
    {
      "epoch": 6.747840333631219,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.025138308339614557,
      "loss": 0.7981,
      "num_input_tokens_seen": 26298600,
      "step": 45305
    },
    {
      "epoch": 6.7485850461721775,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02513687134994265,
      "loss": 0.7925,
      "num_input_tokens_seen": 26301448,
      "step": 45310
    },
    {
      "epoch": 6.749329758713137,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.025135434189018598,
      "loss": 0.8136,
      "num_input_tokens_seen": 26304520,
      "step": 45315
    },
    {
      "epoch": 6.750074471254096,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02513399685686669,
      "loss": 0.8177,
      "num_input_tokens_seen": 26307944,
      "step": 45320
    },
    {
      "epoch": 6.7508191837950555,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.025132559353511205,
      "loss": 0.8147,
      "num_input_tokens_seen": 26310856,
      "step": 45325
    },
    {
      "epoch": 6.751563896336014,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02513112167897643,
      "loss": 0.7941,
      "num_input_tokens_seen": 26313832,
      "step": 45330
    },
    {
      "epoch": 6.752308608876973,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.025129683833286648,
      "loss": 0.785,
      "num_input_tokens_seen": 26316744,
      "step": 45335
    },
    {
      "epoch": 6.753053321417933,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02512824581646616,
      "loss": 0.7855,
      "num_input_tokens_seen": 26319784,
      "step": 45340
    },
    {
      "epoch": 6.753798033958892,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.025126807628539242,
      "loss": 0.7966,
      "num_input_tokens_seen": 26322728,
      "step": 45345
    },
    {
      "epoch": 6.754542746499851,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02512536926953021,
      "loss": 0.7936,
      "num_input_tokens_seen": 26325736,
      "step": 45350
    },
    {
      "epoch": 6.75528745904081,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02512393073946335,
      "loss": 0.7919,
      "num_input_tokens_seen": 26328744,
      "step": 45355
    },
    {
      "epoch": 6.7560321715817695,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.025122492038362972,
      "loss": 0.8122,
      "num_input_tokens_seen": 26331784,
      "step": 45360
    },
    {
      "epoch": 6.756776884122728,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02512105316625338,
      "loss": 0.8087,
      "num_input_tokens_seen": 26334696,
      "step": 45365
    },
    {
      "epoch": 6.757521596663688,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02511961412315888,
      "loss": 0.801,
      "num_input_tokens_seen": 26337512,
      "step": 45370
    },
    {
      "epoch": 6.758266309204647,
      "grad_norm": 0.01953125,
      "learning_rate": 0.025118174909103785,
      "loss": 0.7927,
      "num_input_tokens_seen": 26340456,
      "step": 45375
    },
    {
      "epoch": 6.759011021745606,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.025116735524112407,
      "loss": 0.7973,
      "num_input_tokens_seen": 26343336,
      "step": 45380
    },
    {
      "epoch": 6.759755734286565,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.025115295968209064,
      "loss": 0.7997,
      "num_input_tokens_seen": 26346120,
      "step": 45385
    },
    {
      "epoch": 6.760500446827525,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02511385624141808,
      "loss": 0.7959,
      "num_input_tokens_seen": 26348904,
      "step": 45390
    },
    {
      "epoch": 6.7612451593684835,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.025112416343763773,
      "loss": 0.8221,
      "num_input_tokens_seen": 26351816,
      "step": 45395
    },
    {
      "epoch": 6.761989871909443,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02511097627527047,
      "loss": 0.789,
      "num_input_tokens_seen": 26354536,
      "step": 45400
    },
    {
      "epoch": 6.762734584450402,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.025109536035962495,
      "loss": 0.8168,
      "num_input_tokens_seen": 26357288,
      "step": 45405
    },
    {
      "epoch": 6.7634792969913615,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.025108095625864187,
      "loss": 0.7877,
      "num_input_tokens_seen": 26360168,
      "step": 45410
    },
    {
      "epoch": 6.76422400953232,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.025106655044999877,
      "loss": 0.7624,
      "num_input_tokens_seen": 26362728,
      "step": 45415
    },
    {
      "epoch": 6.76496872207328,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.0251052142933939,
      "loss": 0.7801,
      "num_input_tokens_seen": 26365768,
      "step": 45420
    },
    {
      "epoch": 6.765713434614239,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.025103773371070604,
      "loss": 0.771,
      "num_input_tokens_seen": 26368680,
      "step": 45425
    },
    {
      "epoch": 6.766458147155198,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.025102332278054317,
      "loss": 0.79,
      "num_input_tokens_seen": 26371880,
      "step": 45430
    },
    {
      "epoch": 6.767202859696157,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.025100891014369398,
      "loss": 0.7928,
      "num_input_tokens_seen": 26374856,
      "step": 45435
    },
    {
      "epoch": 6.767947572237117,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.025099449580040196,
      "loss": 0.803,
      "num_input_tokens_seen": 26377768,
      "step": 45440
    },
    {
      "epoch": 6.7686922847780755,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.025098007975091055,
      "loss": 0.8051,
      "num_input_tokens_seen": 26380712,
      "step": 45445
    },
    {
      "epoch": 6.769436997319035,
      "grad_norm": 0.01220703125,
      "learning_rate": 0.025096566199546328,
      "loss": 0.7927,
      "num_input_tokens_seen": 26383624,
      "step": 45450
    },
    {
      "epoch": 6.770181709859994,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.025095124253430377,
      "loss": 0.7755,
      "num_input_tokens_seen": 26386568,
      "step": 45455
    },
    {
      "epoch": 6.7709264224009535,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.025093682136767565,
      "loss": 0.8103,
      "num_input_tokens_seen": 26389160,
      "step": 45460
    },
    {
      "epoch": 6.771671134941912,
      "grad_norm": 0.011962890625,
      "learning_rate": 0.02509223984958225,
      "loss": 0.8151,
      "num_input_tokens_seen": 26391880,
      "step": 45465
    },
    {
      "epoch": 6.772415847482872,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.025090797391898803,
      "loss": 0.8006,
      "num_input_tokens_seen": 26394664,
      "step": 45470
    },
    {
      "epoch": 6.773160560023831,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.025089354763741592,
      "loss": 0.8056,
      "num_input_tokens_seen": 26397512,
      "step": 45475
    },
    {
      "epoch": 6.77390527256479,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.025087911965134983,
      "loss": 0.7797,
      "num_input_tokens_seen": 26400328,
      "step": 45480
    },
    {
      "epoch": 6.774649985105749,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.025086468996103353,
      "loss": 0.8052,
      "num_input_tokens_seen": 26402952,
      "step": 45485
    },
    {
      "epoch": 6.775394697646709,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.025085025856671078,
      "loss": 0.83,
      "num_input_tokens_seen": 26405768,
      "step": 45490
    },
    {
      "epoch": 6.7761394101876675,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02508358254686255,
      "loss": 0.8199,
      "num_input_tokens_seen": 26408552,
      "step": 45495
    },
    {
      "epoch": 6.776884122728626,
      "grad_norm": 0.017578125,
      "learning_rate": 0.025082139066702137,
      "loss": 0.8016,
      "num_input_tokens_seen": 26411176,
      "step": 45500
    },
    {
      "epoch": 6.777628835269586,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.025080695416214233,
      "loss": 0.7774,
      "num_input_tokens_seen": 26413960,
      "step": 45505
    },
    {
      "epoch": 6.778373547810546,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.025079251595423225,
      "loss": 0.8155,
      "num_input_tokens_seen": 26416936,
      "step": 45510
    },
    {
      "epoch": 6.779118260351504,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.025077807604353505,
      "loss": 0.802,
      "num_input_tokens_seen": 26419944,
      "step": 45515
    },
    {
      "epoch": 6.779862972892463,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02507636344302947,
      "loss": 0.7969,
      "num_input_tokens_seen": 26422824,
      "step": 45520
    },
    {
      "epoch": 6.780607685433423,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.025074919111475517,
      "loss": 0.7975,
      "num_input_tokens_seen": 26425736,
      "step": 45525
    },
    {
      "epoch": 6.781352397974382,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.025073474609716043,
      "loss": 0.8111,
      "num_input_tokens_seen": 26428776,
      "step": 45530
    },
    {
      "epoch": 6.782097110515341,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02507202993777545,
      "loss": 0.8087,
      "num_input_tokens_seen": 26431496,
      "step": 45535
    },
    {
      "epoch": 6.7828418230563,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.025070585095678152,
      "loss": 0.7965,
      "num_input_tokens_seen": 26434600,
      "step": 45540
    },
    {
      "epoch": 6.7835865355972595,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.025069140083448557,
      "loss": 0.8453,
      "num_input_tokens_seen": 26437448,
      "step": 45545
    },
    {
      "epoch": 6.784331248138218,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02506769490111107,
      "loss": 0.7942,
      "num_input_tokens_seen": 26440328,
      "step": 45550
    },
    {
      "epoch": 6.785075960679178,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02506624954869011,
      "loss": 0.7807,
      "num_input_tokens_seen": 26443240,
      "step": 45555
    },
    {
      "epoch": 6.785820673220137,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.025064804026210096,
      "loss": 0.7852,
      "num_input_tokens_seen": 26445928,
      "step": 45560
    },
    {
      "epoch": 6.786565385761096,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.025063358333695447,
      "loss": 0.8108,
      "num_input_tokens_seen": 26448872,
      "step": 45565
    },
    {
      "epoch": 6.787310098302055,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.025061912471170584,
      "loss": 0.8035,
      "num_input_tokens_seen": 26451912,
      "step": 45570
    },
    {
      "epoch": 6.788054810843015,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02506046643865994,
      "loss": 0.7899,
      "num_input_tokens_seen": 26454664,
      "step": 45575
    },
    {
      "epoch": 6.7887995233839735,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.025059020236187942,
      "loss": 0.812,
      "num_input_tokens_seen": 26457448,
      "step": 45580
    },
    {
      "epoch": 6.789544235924933,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02505757386377902,
      "loss": 0.8178,
      "num_input_tokens_seen": 26460456,
      "step": 45585
    },
    {
      "epoch": 6.790288948465892,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.025056127321457608,
      "loss": 0.8128,
      "num_input_tokens_seen": 26463528,
      "step": 45590
    },
    {
      "epoch": 6.791033661006852,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.025054680609248144,
      "loss": 0.8133,
      "num_input_tokens_seen": 26466440,
      "step": 45595
    },
    {
      "epoch": 6.79177837354781,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02505323372717507,
      "loss": 0.8089,
      "num_input_tokens_seen": 26469160,
      "step": 45600
    },
    {
      "epoch": 6.79252308608877,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.025051786675262835,
      "loss": 0.7978,
      "num_input_tokens_seen": 26471688,
      "step": 45605
    },
    {
      "epoch": 6.793267798629729,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.025050339453535877,
      "loss": 0.8013,
      "num_input_tokens_seen": 26474504,
      "step": 45610
    },
    {
      "epoch": 6.794012511170688,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.025048892062018647,
      "loss": 0.7968,
      "num_input_tokens_seen": 26477480,
      "step": 45615
    },
    {
      "epoch": 6.794757223711647,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.0250474445007356,
      "loss": 0.7913,
      "num_input_tokens_seen": 26480200,
      "step": 45620
    },
    {
      "epoch": 6.795501936252607,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02504599676971119,
      "loss": 0.8045,
      "num_input_tokens_seen": 26483048,
      "step": 45625
    },
    {
      "epoch": 6.7962466487935655,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.025044548868969875,
      "loss": 0.7948,
      "num_input_tokens_seen": 26485768,
      "step": 45630
    },
    {
      "epoch": 6.796991361334525,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.025043100798536118,
      "loss": 0.8004,
      "num_input_tokens_seen": 26488872,
      "step": 45635
    },
    {
      "epoch": 6.797736073875484,
      "grad_norm": 0.02734375,
      "learning_rate": 0.025041652558434376,
      "loss": 0.7887,
      "num_input_tokens_seen": 26491784,
      "step": 45640
    },
    {
      "epoch": 6.798480786416444,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.025040204148689127,
      "loss": 0.8035,
      "num_input_tokens_seen": 26494696,
      "step": 45645
    },
    {
      "epoch": 6.799225498957402,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.025038755569324826,
      "loss": 0.7844,
      "num_input_tokens_seen": 26497384,
      "step": 45650
    },
    {
      "epoch": 6.799970211498362,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.025037306820365954,
      "loss": 0.7853,
      "num_input_tokens_seen": 26500392,
      "step": 45655
    },
    {
      "epoch": 6.800714924039321,
      "grad_norm": 0.02734375,
      "learning_rate": 0.025035857901836985,
      "loss": 0.8227,
      "num_input_tokens_seen": 26503176,
      "step": 45660
    },
    {
      "epoch": 6.8014596365802795,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.0250344088137624,
      "loss": 0.815,
      "num_input_tokens_seen": 26505992,
      "step": 45665
    },
    {
      "epoch": 6.802204349121239,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.025032959556166676,
      "loss": 0.8145,
      "num_input_tokens_seen": 26509064,
      "step": 45670
    },
    {
      "epoch": 6.802949061662199,
      "grad_norm": 0.01953125,
      "learning_rate": 0.025031510129074293,
      "loss": 0.7981,
      "num_input_tokens_seen": 26512168,
      "step": 45675
    },
    {
      "epoch": 6.803693774203158,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.025030060532509744,
      "loss": 0.822,
      "num_input_tokens_seen": 26514984,
      "step": 45680
    },
    {
      "epoch": 6.804438486744116,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.025028610766497518,
      "loss": 0.8053,
      "num_input_tokens_seen": 26518088,
      "step": 45685
    },
    {
      "epoch": 6.805183199285076,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.025027160831062104,
      "loss": 0.8013,
      "num_input_tokens_seen": 26520904,
      "step": 45690
    },
    {
      "epoch": 6.805927911826036,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.025025710726227997,
      "loss": 0.8116,
      "num_input_tokens_seen": 26523688,
      "step": 45695
    },
    {
      "epoch": 6.806672624366994,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.025024260452019702,
      "loss": 0.8031,
      "num_input_tokens_seen": 26526568,
      "step": 45700
    },
    {
      "epoch": 6.807417336907953,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.025022810008461713,
      "loss": 0.8114,
      "num_input_tokens_seen": 26529544,
      "step": 45705
    },
    {
      "epoch": 6.808162049448913,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.02502135939557853,
      "loss": 0.8193,
      "num_input_tokens_seen": 26532488,
      "step": 45710
    },
    {
      "epoch": 6.8089067619898715,
      "grad_norm": 0.02734375,
      "learning_rate": 0.025019908613394672,
      "loss": 0.7991,
      "num_input_tokens_seen": 26535464,
      "step": 45715
    },
    {
      "epoch": 6.809651474530831,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.025018457661934643,
      "loss": 0.7829,
      "num_input_tokens_seen": 26538536,
      "step": 45720
    },
    {
      "epoch": 6.81039618707179,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.025017006541222953,
      "loss": 0.8083,
      "num_input_tokens_seen": 26541128,
      "step": 45725
    },
    {
      "epoch": 6.81114089961275,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.025015555251284113,
      "loss": 0.8146,
      "num_input_tokens_seen": 26543976,
      "step": 45730
    },
    {
      "epoch": 6.811885612153708,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02501410379214265,
      "loss": 0.8243,
      "num_input_tokens_seen": 26546856,
      "step": 45735
    },
    {
      "epoch": 6.812630324694668,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02501265216382308,
      "loss": 0.7953,
      "num_input_tokens_seen": 26550024,
      "step": 45740
    },
    {
      "epoch": 6.813375037235627,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02501120036634993,
      "loss": 0.7914,
      "num_input_tokens_seen": 26553096,
      "step": 45745
    },
    {
      "epoch": 6.814119749776586,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.025009748399747723,
      "loss": 0.8168,
      "num_input_tokens_seen": 26555912,
      "step": 45750
    },
    {
      "epoch": 6.814864462317545,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.025008296264040993,
      "loss": 0.7933,
      "num_input_tokens_seen": 26558792,
      "step": 45755
    },
    {
      "epoch": 6.815609174858505,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02500684395925427,
      "loss": 0.795,
      "num_input_tokens_seen": 26561704,
      "step": 45760
    },
    {
      "epoch": 6.816353887399464,
      "grad_norm": 0.021484375,
      "learning_rate": 0.025005391485412084,
      "loss": 0.7932,
      "num_input_tokens_seen": 26564456,
      "step": 45765
    },
    {
      "epoch": 6.817098599940423,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02500393884253898,
      "loss": 0.785,
      "num_input_tokens_seen": 26567272,
      "step": 45770
    },
    {
      "epoch": 6.817843312481382,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.0250024860306595,
      "loss": 0.7613,
      "num_input_tokens_seen": 26570120,
      "step": 45775
    },
    {
      "epoch": 6.818588025022342,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.025001033049798182,
      "loss": 0.8091,
      "num_input_tokens_seen": 26573288,
      "step": 45780
    },
    {
      "epoch": 6.8193327375633,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.024999579899979574,
      "loss": 0.8033,
      "num_input_tokens_seen": 26576488,
      "step": 45785
    },
    {
      "epoch": 6.82007745010426,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.024998126581228228,
      "loss": 0.8062,
      "num_input_tokens_seen": 26579208,
      "step": 45790
    },
    {
      "epoch": 6.820822162645219,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02499667309356869,
      "loss": 0.8194,
      "num_input_tokens_seen": 26582248,
      "step": 45795
    },
    {
      "epoch": 6.821566875186178,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02499521943702553,
      "loss": 0.8056,
      "num_input_tokens_seen": 26585096,
      "step": 45800
    },
    {
      "epoch": 6.822311587727137,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02499376561162329,
      "loss": 0.8117,
      "num_input_tokens_seen": 26588168,
      "step": 45805
    },
    {
      "epoch": 6.823056300268097,
      "grad_norm": 0.0234375,
      "learning_rate": 0.024992311617386537,
      "loss": 0.8111,
      "num_input_tokens_seen": 26590984,
      "step": 45810
    },
    {
      "epoch": 6.823801012809056,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02499085745433984,
      "loss": 0.8141,
      "num_input_tokens_seen": 26593896,
      "step": 45815
    },
    {
      "epoch": 6.824545725350015,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.024989403122507754,
      "loss": 0.779,
      "num_input_tokens_seen": 26596904,
      "step": 45820
    },
    {
      "epoch": 6.825290437890974,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.02498794862191486,
      "loss": 0.7878,
      "num_input_tokens_seen": 26599592,
      "step": 45825
    },
    {
      "epoch": 6.826035150431934,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02498649395258572,
      "loss": 0.8013,
      "num_input_tokens_seen": 26602536,
      "step": 45830
    },
    {
      "epoch": 6.826779862972892,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.024985039114544923,
      "loss": 0.807,
      "num_input_tokens_seen": 26605352,
      "step": 45835
    },
    {
      "epoch": 6.827524575513852,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.024983584107817033,
      "loss": 0.8331,
      "num_input_tokens_seen": 26607912,
      "step": 45840
    },
    {
      "epoch": 6.828269288054811,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02498212893242664,
      "loss": 0.7891,
      "num_input_tokens_seen": 26610568,
      "step": 45845
    },
    {
      "epoch": 6.82901400059577,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.024980673588398318,
      "loss": 0.8285,
      "num_input_tokens_seen": 26613672,
      "step": 45850
    },
    {
      "epoch": 6.829758713136729,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.024979218075756662,
      "loss": 0.8057,
      "num_input_tokens_seen": 26616680,
      "step": 45855
    },
    {
      "epoch": 6.830503425677689,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.024977762394526256,
      "loss": 0.8095,
      "num_input_tokens_seen": 26619816,
      "step": 45860
    },
    {
      "epoch": 6.831248138218648,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.0249763065447317,
      "loss": 0.7951,
      "num_input_tokens_seen": 26622760,
      "step": 45865
    },
    {
      "epoch": 6.831992850759606,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.024974850526397585,
      "loss": 0.8263,
      "num_input_tokens_seen": 26625512,
      "step": 45870
    },
    {
      "epoch": 6.832737563300566,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.024973394339548503,
      "loss": 0.7981,
      "num_input_tokens_seen": 26630056,
      "step": 45875
    },
    {
      "epoch": 6.833482275841525,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.024971937984209065,
      "loss": 0.7883,
      "num_input_tokens_seen": 26632808,
      "step": 45880
    },
    {
      "epoch": 6.834226988382484,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02497048146040387,
      "loss": 0.7847,
      "num_input_tokens_seen": 26635560,
      "step": 45885
    },
    {
      "epoch": 6.834971700923443,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.024969024768157527,
      "loss": 0.8107,
      "num_input_tokens_seen": 26638472,
      "step": 45890
    },
    {
      "epoch": 6.835716413464403,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.024967567907494635,
      "loss": 0.7906,
      "num_input_tokens_seen": 26641448,
      "step": 45895
    },
    {
      "epoch": 6.836461126005362,
      "grad_norm": 0.01953125,
      "learning_rate": 0.024966110878439817,
      "loss": 0.7956,
      "num_input_tokens_seen": 26644264,
      "step": 45900
    },
    {
      "epoch": 6.837205838546321,
      "grad_norm": 0.01171875,
      "learning_rate": 0.024964653681017683,
      "loss": 0.8088,
      "num_input_tokens_seen": 26647016,
      "step": 45905
    },
    {
      "epoch": 6.83795055108728,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.024963196315252855,
      "loss": 0.7927,
      "num_input_tokens_seen": 26649832,
      "step": 45910
    },
    {
      "epoch": 6.83869526362824,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.024961738781169952,
      "loss": 0.8324,
      "num_input_tokens_seen": 26652776,
      "step": 45915
    },
    {
      "epoch": 6.839439976169198,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.024960281078793594,
      "loss": 0.8042,
      "num_input_tokens_seen": 26655592,
      "step": 45920
    },
    {
      "epoch": 6.840184688710158,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.024958823208148417,
      "loss": 0.8241,
      "num_input_tokens_seen": 26658824,
      "step": 45925
    },
    {
      "epoch": 6.840929401251117,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02495736516925904,
      "loss": 0.8133,
      "num_input_tokens_seen": 26661480,
      "step": 45930
    },
    {
      "epoch": 6.8416741137920765,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.024955906962150096,
      "loss": 0.8213,
      "num_input_tokens_seen": 26664328,
      "step": 45935
    },
    {
      "epoch": 6.842418826333035,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.024954448586846225,
      "loss": 0.823,
      "num_input_tokens_seen": 26667592,
      "step": 45940
    },
    {
      "epoch": 6.843163538873995,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02495299004337206,
      "loss": 0.7865,
      "num_input_tokens_seen": 26670504,
      "step": 45945
    },
    {
      "epoch": 6.843908251414954,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.024951531331752246,
      "loss": 0.7986,
      "num_input_tokens_seen": 26673448,
      "step": 45950
    },
    {
      "epoch": 6.844652963955913,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02495007245201142,
      "loss": 0.7962,
      "num_input_tokens_seen": 26676136,
      "step": 45955
    },
    {
      "epoch": 6.845397676496872,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.024948613404174237,
      "loss": 0.8022,
      "num_input_tokens_seen": 26679080,
      "step": 45960
    },
    {
      "epoch": 6.846142389037832,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.024947154188265343,
      "loss": 0.7916,
      "num_input_tokens_seen": 26681864,
      "step": 45965
    },
    {
      "epoch": 6.84688710157879,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02494569480430939,
      "loss": 0.8015,
      "num_input_tokens_seen": 26684744,
      "step": 45970
    },
    {
      "epoch": 6.84763181411975,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02494423525233103,
      "loss": 0.7908,
      "num_input_tokens_seen": 26687688,
      "step": 45975
    },
    {
      "epoch": 6.848376526660709,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.024942775532354917,
      "loss": 0.7934,
      "num_input_tokens_seen": 26690504,
      "step": 45980
    },
    {
      "epoch": 6.8491212392016685,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.024941315644405722,
      "loss": 0.7987,
      "num_input_tokens_seen": 26693224,
      "step": 45985
    },
    {
      "epoch": 6.849865951742627,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.024939855588508104,
      "loss": 0.7987,
      "num_input_tokens_seen": 26695976,
      "step": 45990
    },
    {
      "epoch": 6.850610664283587,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.024938395364686727,
      "loss": 0.7943,
      "num_input_tokens_seen": 26698760,
      "step": 45995
    },
    {
      "epoch": 6.851355376824546,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.02493693497296626,
      "loss": 0.7895,
      "num_input_tokens_seen": 26701448,
      "step": 46000
    },
    {
      "epoch": 6.852100089365505,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.024935474413371378,
      "loss": 0.7928,
      "num_input_tokens_seen": 26704520,
      "step": 46005
    },
    {
      "epoch": 6.852844801906464,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02493401368592675,
      "loss": 0.7995,
      "num_input_tokens_seen": 26707176,
      "step": 46010
    },
    {
      "epoch": 6.853589514447423,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.02493255279065706,
      "loss": 0.7719,
      "num_input_tokens_seen": 26710088,
      "step": 46015
    },
    {
      "epoch": 6.8543342269883825,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.024931091727586986,
      "loss": 0.7831,
      "num_input_tokens_seen": 26713000,
      "step": 46020
    },
    {
      "epoch": 6.855078939529342,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.02492963049674121,
      "loss": 0.7824,
      "num_input_tokens_seen": 26716232,
      "step": 46025
    },
    {
      "epoch": 6.855823652070301,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02492816909814442,
      "loss": 0.7868,
      "num_input_tokens_seen": 26719112,
      "step": 46030
    },
    {
      "epoch": 6.85656836461126,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.0249267075318213,
      "loss": 0.7897,
      "num_input_tokens_seen": 26721864,
      "step": 46035
    },
    {
      "epoch": 6.857313077152219,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02492524579779655,
      "loss": 0.7942,
      "num_input_tokens_seen": 26725256,
      "step": 46040
    },
    {
      "epoch": 6.858057789693179,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02492378389609486,
      "loss": 0.8176,
      "num_input_tokens_seen": 26727944,
      "step": 46045
    },
    {
      "epoch": 6.858802502234138,
      "grad_norm": 0.02734375,
      "learning_rate": 0.024922321826740923,
      "loss": 0.8038,
      "num_input_tokens_seen": 26730600,
      "step": 46050
    },
    {
      "epoch": 6.859547214775096,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.024920859589759443,
      "loss": 0.8003,
      "num_input_tokens_seen": 26733576,
      "step": 46055
    },
    {
      "epoch": 6.860291927316056,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.024919397185175128,
      "loss": 0.7897,
      "num_input_tokens_seen": 26736424,
      "step": 46060
    },
    {
      "epoch": 6.861036639857015,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.024917934613012675,
      "loss": 0.7939,
      "num_input_tokens_seen": 26739208,
      "step": 46065
    },
    {
      "epoch": 6.8617813523979745,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.0249164718732968,
      "loss": 0.8331,
      "num_input_tokens_seen": 26742056,
      "step": 46070
    },
    {
      "epoch": 6.862526064938933,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.024915008966052206,
      "loss": 0.801,
      "num_input_tokens_seen": 26744840,
      "step": 46075
    },
    {
      "epoch": 6.863270777479893,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02491354589130362,
      "loss": 0.8233,
      "num_input_tokens_seen": 26747688,
      "step": 46080
    },
    {
      "epoch": 6.864015490020852,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02491208264907575,
      "loss": 0.7941,
      "num_input_tokens_seen": 26750408,
      "step": 46085
    },
    {
      "epoch": 6.864760202561811,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.024910619239393316,
      "loss": 0.8061,
      "num_input_tokens_seen": 26753096,
      "step": 46090
    },
    {
      "epoch": 6.86550491510277,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.02490915566228104,
      "loss": 0.817,
      "num_input_tokens_seen": 26755848,
      "step": 46095
    },
    {
      "epoch": 6.86624962764373,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.024907691917763652,
      "loss": 0.7755,
      "num_input_tokens_seen": 26758632,
      "step": 46100
    },
    {
      "epoch": 6.8669943401846885,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.024906228005865883,
      "loss": 0.7728,
      "num_input_tokens_seen": 26761096,
      "step": 46105
    },
    {
      "epoch": 6.867739052725648,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.024904763926612457,
      "loss": 0.7817,
      "num_input_tokens_seen": 26764008,
      "step": 46110
    },
    {
      "epoch": 6.868483765266607,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.024903299680028107,
      "loss": 0.8034,
      "num_input_tokens_seen": 26766824,
      "step": 46115
    },
    {
      "epoch": 6.8692284778075665,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.024901835266137585,
      "loss": 0.7963,
      "num_input_tokens_seen": 26769544,
      "step": 46120
    },
    {
      "epoch": 6.869973190348525,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.02490037068496561,
      "loss": 0.8186,
      "num_input_tokens_seen": 26772200,
      "step": 46125
    },
    {
      "epoch": 6.870717902889485,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02489890593653694,
      "loss": 0.8274,
      "num_input_tokens_seen": 26775144,
      "step": 46130
    },
    {
      "epoch": 6.871462615430444,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.024897441020876314,
      "loss": 0.8194,
      "num_input_tokens_seen": 26778024,
      "step": 46135
    },
    {
      "epoch": 6.872207327971403,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.024895975938008477,
      "loss": 0.8013,
      "num_input_tokens_seen": 26781000,
      "step": 46140
    },
    {
      "epoch": 6.872952040512362,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.02489451068795819,
      "loss": 0.7849,
      "num_input_tokens_seen": 26783944,
      "step": 46145
    },
    {
      "epoch": 6.873696753053322,
      "grad_norm": 0.01953125,
      "learning_rate": 0.024893045270750196,
      "loss": 0.7826,
      "num_input_tokens_seen": 26786728,
      "step": 46150
    },
    {
      "epoch": 6.8744414655942805,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02489157968640926,
      "loss": 0.8108,
      "num_input_tokens_seen": 26789896,
      "step": 46155
    },
    {
      "epoch": 6.87518617813524,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02489011393496014,
      "loss": 0.7984,
      "num_input_tokens_seen": 26792648,
      "step": 46160
    },
    {
      "epoch": 6.875930890676199,
      "grad_norm": 0.03125,
      "learning_rate": 0.024888648016427595,
      "loss": 0.7661,
      "num_input_tokens_seen": 26795272,
      "step": 46165
    },
    {
      "epoch": 6.8766756032171585,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.024887181930836392,
      "loss": 0.818,
      "num_input_tokens_seen": 26798376,
      "step": 46170
    },
    {
      "epoch": 6.877420315758117,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.0248857156782113,
      "loss": 0.804,
      "num_input_tokens_seen": 26801416,
      "step": 46175
    },
    {
      "epoch": 6.878165028299077,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.024884249258577087,
      "loss": 0.7982,
      "num_input_tokens_seen": 26804104,
      "step": 46180
    },
    {
      "epoch": 6.878909740840036,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.024882782671958533,
      "loss": 0.8166,
      "num_input_tokens_seen": 26806984,
      "step": 46185
    },
    {
      "epoch": 6.879654453380995,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.024881315918380404,
      "loss": 0.7997,
      "num_input_tokens_seen": 26809768,
      "step": 46190
    },
    {
      "epoch": 6.880399165921954,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.024879848997867493,
      "loss": 0.8181,
      "num_input_tokens_seen": 26812744,
      "step": 46195
    },
    {
      "epoch": 6.881143878462913,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.024878381910444565,
      "loss": 0.7793,
      "num_input_tokens_seen": 26815496,
      "step": 46200
    },
    {
      "epoch": 6.8818885910038725,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.024876914656136426,
      "loss": 0.8026,
      "num_input_tokens_seen": 26818536,
      "step": 46205
    },
    {
      "epoch": 6.882633303544832,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.024875447234967842,
      "loss": 0.818,
      "num_input_tokens_seen": 26821352,
      "step": 46210
    },
    {
      "epoch": 6.883378016085791,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.024873979646963622,
      "loss": 0.7815,
      "num_input_tokens_seen": 26824072,
      "step": 46215
    },
    {
      "epoch": 6.88412272862675,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.024872511892148546,
      "loss": 0.8126,
      "num_input_tokens_seen": 26826984,
      "step": 46220
    },
    {
      "epoch": 6.884867441167709,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02487104397054742,
      "loss": 0.7929,
      "num_input_tokens_seen": 26829864,
      "step": 46225
    },
    {
      "epoch": 6.885612153708668,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.024869575882185035,
      "loss": 0.8286,
      "num_input_tokens_seen": 26832616,
      "step": 46230
    },
    {
      "epoch": 6.886356866249628,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.024868107627086196,
      "loss": 0.8495,
      "num_input_tokens_seen": 26835528,
      "step": 46235
    },
    {
      "epoch": 6.8871015787905865,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.02486663920527571,
      "loss": 0.8286,
      "num_input_tokens_seen": 26838632,
      "step": 46240
    },
    {
      "epoch": 6.887846291331546,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.024865170616778384,
      "loss": 0.7947,
      "num_input_tokens_seen": 26841608,
      "step": 46245
    },
    {
      "epoch": 6.888591003872505,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.02486370186161902,
      "loss": 0.8066,
      "num_input_tokens_seen": 26844616,
      "step": 46250
    },
    {
      "epoch": 6.8893357164134645,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.024862232939822447,
      "loss": 0.8061,
      "num_input_tokens_seen": 26847560,
      "step": 46255
    },
    {
      "epoch": 6.890080428954423,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.024860763851413465,
      "loss": 0.7806,
      "num_input_tokens_seen": 26850248,
      "step": 46260
    },
    {
      "epoch": 6.890825141495383,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.024859294596416903,
      "loss": 0.7814,
      "num_input_tokens_seen": 26853288,
      "step": 46265
    },
    {
      "epoch": 6.891569854036342,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.024857825174857583,
      "loss": 0.807,
      "num_input_tokens_seen": 26856040,
      "step": 46270
    },
    {
      "epoch": 6.892314566577301,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.02485635558676032,
      "loss": 0.8242,
      "num_input_tokens_seen": 26858888,
      "step": 46275
    },
    {
      "epoch": 6.89305927911826,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.024854885832149946,
      "loss": 0.7983,
      "num_input_tokens_seen": 26861896,
      "step": 46280
    },
    {
      "epoch": 6.89380399165922,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.024853415911051297,
      "loss": 0.7972,
      "num_input_tokens_seen": 26864520,
      "step": 46285
    },
    {
      "epoch": 6.8945487042001785,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.0248519458234892,
      "loss": 0.8001,
      "num_input_tokens_seen": 26867496,
      "step": 46290
    },
    {
      "epoch": 6.895293416741138,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.024850475569488488,
      "loss": 0.8029,
      "num_input_tokens_seen": 26870248,
      "step": 46295
    },
    {
      "epoch": 6.896038129282097,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.024849005149074007,
      "loss": 0.8155,
      "num_input_tokens_seen": 26873032,
      "step": 46300
    },
    {
      "epoch": 6.896782841823057,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02484753456227059,
      "loss": 0.7964,
      "num_input_tokens_seen": 26876200,
      "step": 46305
    },
    {
      "epoch": 6.897527554364015,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.024846063809103088,
      "loss": 0.7769,
      "num_input_tokens_seen": 26878952,
      "step": 46310
    },
    {
      "epoch": 6.898272266904975,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.024844592889596343,
      "loss": 0.7836,
      "num_input_tokens_seen": 26882024,
      "step": 46315
    },
    {
      "epoch": 6.899016979445934,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.024843121803775205,
      "loss": 0.8276,
      "num_input_tokens_seen": 26884936,
      "step": 46320
    },
    {
      "epoch": 6.899761691986893,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.024841650551664533,
      "loss": 0.8137,
      "num_input_tokens_seen": 26887944,
      "step": 46325
    },
    {
      "epoch": 6.900506404527852,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.024840179133289175,
      "loss": 0.7907,
      "num_input_tokens_seen": 26891304,
      "step": 46330
    },
    {
      "epoch": 6.901251117068812,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.024838707548673993,
      "loss": 0.8002,
      "num_input_tokens_seen": 26894152,
      "step": 46335
    },
    {
      "epoch": 6.9019958296097705,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.024837235797843846,
      "loss": 0.7823,
      "num_input_tokens_seen": 26896936,
      "step": 46340
    },
    {
      "epoch": 6.90274054215073,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.0248357638808236,
      "loss": 0.8009,
      "num_input_tokens_seen": 26899752,
      "step": 46345
    },
    {
      "epoch": 6.903485254691689,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.024834291797638113,
      "loss": 0.7843,
      "num_input_tokens_seen": 26902696,
      "step": 46350
    },
    {
      "epoch": 6.904229967232649,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02483281954831227,
      "loss": 0.7857,
      "num_input_tokens_seen": 26906856,
      "step": 46355
    },
    {
      "epoch": 6.904974679773607,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02483134713287093,
      "loss": 0.8153,
      "num_input_tokens_seen": 26909608,
      "step": 46360
    },
    {
      "epoch": 6.905719392314566,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.024829874551338973,
      "loss": 0.8014,
      "num_input_tokens_seen": 26912488,
      "step": 46365
    },
    {
      "epoch": 6.906464104855526,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.024828401803741275,
      "loss": 0.783,
      "num_input_tokens_seen": 26915464,
      "step": 46370
    },
    {
      "epoch": 6.907208817396485,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02482692889010272,
      "loss": 0.7912,
      "num_input_tokens_seen": 26918824,
      "step": 46375
    },
    {
      "epoch": 6.907953529937444,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.024825455810448187,
      "loss": 0.8109,
      "num_input_tokens_seen": 26921896,
      "step": 46380
    },
    {
      "epoch": 6.908698242478403,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.024823982564802564,
      "loss": 0.7991,
      "num_input_tokens_seen": 26924520,
      "step": 46385
    },
    {
      "epoch": 6.909442955019363,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02482250915319074,
      "loss": 0.7944,
      "num_input_tokens_seen": 26927016,
      "step": 46390
    },
    {
      "epoch": 6.910187667560322,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02482103557563761,
      "loss": 0.7928,
      "num_input_tokens_seen": 26929736,
      "step": 46395
    },
    {
      "epoch": 6.910932380101281,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.024819561832168064,
      "loss": 0.8264,
      "num_input_tokens_seen": 26932712,
      "step": 46400
    },
    {
      "epoch": 6.91167709264224,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.024818087922807,
      "loss": 0.8134,
      "num_input_tokens_seen": 26935720,
      "step": 46405
    },
    {
      "epoch": 6.912421805183199,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02481661384757932,
      "loss": 0.8281,
      "num_input_tokens_seen": 26938728,
      "step": 46410
    },
    {
      "epoch": 6.913166517724158,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.024815139606509926,
      "loss": 0.8056,
      "num_input_tokens_seen": 26941608,
      "step": 46415
    },
    {
      "epoch": 6.913911230265118,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.024813665199623726,
      "loss": 0.8037,
      "num_input_tokens_seen": 26944488,
      "step": 46420
    },
    {
      "epoch": 6.9146559428060765,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.024812190626945622,
      "loss": 0.7946,
      "num_input_tokens_seen": 26947272,
      "step": 46425
    },
    {
      "epoch": 6.915400655347036,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02481071588850053,
      "loss": 0.8141,
      "num_input_tokens_seen": 26950088,
      "step": 46430
    },
    {
      "epoch": 6.916145367887995,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02480924098431337,
      "loss": 0.7973,
      "num_input_tokens_seen": 26952904,
      "step": 46435
    },
    {
      "epoch": 6.916890080428955,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02480776591440905,
      "loss": 0.8011,
      "num_input_tokens_seen": 26956040,
      "step": 46440
    },
    {
      "epoch": 6.917634792969913,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.02480629067881249,
      "loss": 0.7758,
      "num_input_tokens_seen": 26958760,
      "step": 46445
    },
    {
      "epoch": 6.918379505510873,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.024804815277548624,
      "loss": 0.7878,
      "num_input_tokens_seen": 26962056,
      "step": 46450
    },
    {
      "epoch": 6.919124218051832,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02480333971064236,
      "loss": 0.8101,
      "num_input_tokens_seen": 26964936,
      "step": 46455
    },
    {
      "epoch": 6.919868930592791,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02480186397811864,
      "loss": 0.7988,
      "num_input_tokens_seen": 26967848,
      "step": 46460
    },
    {
      "epoch": 6.92061364313375,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.024800388080002393,
      "loss": 0.8045,
      "num_input_tokens_seen": 26970504,
      "step": 46465
    },
    {
      "epoch": 6.92135835567471,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.024798912016318547,
      "loss": 0.799,
      "num_input_tokens_seen": 26973384,
      "step": 46470
    },
    {
      "epoch": 6.922103068215669,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.024797435787092037,
      "loss": 0.7884,
      "num_input_tokens_seen": 26976392,
      "step": 46475
    },
    {
      "epoch": 6.922847780756628,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.024795959392347813,
      "loss": 0.8013,
      "num_input_tokens_seen": 26978984,
      "step": 46480
    },
    {
      "epoch": 6.923592493297587,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02479448283211081,
      "loss": 0.7969,
      "num_input_tokens_seen": 26981768,
      "step": 46485
    },
    {
      "epoch": 6.924337205838547,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.024793006106405974,
      "loss": 0.786,
      "num_input_tokens_seen": 26984744,
      "step": 46490
    },
    {
      "epoch": 6.925081918379505,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.024791529215258252,
      "loss": 0.826,
      "num_input_tokens_seen": 26987944,
      "step": 46495
    },
    {
      "epoch": 6.925826630920465,
      "grad_norm": 0.03125,
      "learning_rate": 0.024790052158692597,
      "loss": 0.8117,
      "num_input_tokens_seen": 26990696,
      "step": 46500
    },
    {
      "epoch": 6.926571343461424,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.024788574936733962,
      "loss": 0.7975,
      "num_input_tokens_seen": 26993640,
      "step": 46505
    },
    {
      "epoch": 6.927316056002383,
      "grad_norm": 0.01202392578125,
      "learning_rate": 0.0247870975494073,
      "loss": 0.7833,
      "num_input_tokens_seen": 26996392,
      "step": 46510
    },
    {
      "epoch": 6.928060768543342,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.02478561999673757,
      "loss": 0.8279,
      "num_input_tokens_seen": 26999272,
      "step": 46515
    },
    {
      "epoch": 6.928805481084302,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.024784142278749737,
      "loss": 0.8039,
      "num_input_tokens_seen": 27001960,
      "step": 46520
    },
    {
      "epoch": 6.929550193625261,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.024782664395468765,
      "loss": 0.7943,
      "num_input_tokens_seen": 27004872,
      "step": 46525
    },
    {
      "epoch": 6.930294906166219,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02478118634691962,
      "loss": 0.8041,
      "num_input_tokens_seen": 27007880,
      "step": 46530
    },
    {
      "epoch": 6.931039618707179,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.024779708133127278,
      "loss": 0.7891,
      "num_input_tokens_seen": 27010664,
      "step": 46535
    },
    {
      "epoch": 6.931784331248139,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.024778229754116702,
      "loss": 0.7906,
      "num_input_tokens_seen": 27013672,
      "step": 46540
    },
    {
      "epoch": 6.932529043789097,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02477675120991287,
      "loss": 0.7981,
      "num_input_tokens_seen": 27016648,
      "step": 46545
    },
    {
      "epoch": 6.933273756330056,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02477527250054077,
      "loss": 0.8045,
      "num_input_tokens_seen": 27019432,
      "step": 46550
    },
    {
      "epoch": 6.934018468871016,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.024773793626025373,
      "loss": 0.794,
      "num_input_tokens_seen": 27022472,
      "step": 46555
    },
    {
      "epoch": 6.9347631814119755,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.024772314586391665,
      "loss": 0.8171,
      "num_input_tokens_seen": 27025384,
      "step": 46560
    },
    {
      "epoch": 6.935507893952934,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.024770835381664632,
      "loss": 0.7931,
      "num_input_tokens_seen": 27028328,
      "step": 46565
    },
    {
      "epoch": 6.936252606493893,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.024769356011869272,
      "loss": 0.7889,
      "num_input_tokens_seen": 27030984,
      "step": 46570
    },
    {
      "epoch": 6.936997319034853,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02476787647703057,
      "loss": 0.7848,
      "num_input_tokens_seen": 27033960,
      "step": 46575
    },
    {
      "epoch": 6.937742031575811,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02476639677717352,
      "loss": 0.7757,
      "num_input_tokens_seen": 27036712,
      "step": 46580
    },
    {
      "epoch": 6.938486744116771,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.024764916912323125,
      "loss": 0.795,
      "num_input_tokens_seen": 27039368,
      "step": 46585
    },
    {
      "epoch": 6.93923145665773,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.024763436882504375,
      "loss": 0.7955,
      "num_input_tokens_seen": 27042280,
      "step": 46590
    },
    {
      "epoch": 6.939976169198689,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02476195668774229,
      "loss": 0.7927,
      "num_input_tokens_seen": 27045320,
      "step": 46595
    },
    {
      "epoch": 6.940720881739648,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02476047632806187,
      "loss": 0.8071,
      "num_input_tokens_seen": 27047912,
      "step": 46600
    },
    {
      "epoch": 6.941465594280608,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02475899580348812,
      "loss": 0.7865,
      "num_input_tokens_seen": 27050888,
      "step": 46605
    },
    {
      "epoch": 6.942210306821567,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.024757515114046055,
      "loss": 0.8065,
      "num_input_tokens_seen": 27053992,
      "step": 46610
    },
    {
      "epoch": 6.942955019362526,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.024756034259760685,
      "loss": 0.7589,
      "num_input_tokens_seen": 27057096,
      "step": 46615
    },
    {
      "epoch": 6.943699731903485,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.024754553240657038,
      "loss": 0.7887,
      "num_input_tokens_seen": 27060072,
      "step": 46620
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.024753072056760122,
      "loss": 0.8005,
      "num_input_tokens_seen": 27062856,
      "step": 46625
    },
    {
      "epoch": 6.945189156985403,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02475159070809497,
      "loss": 0.8194,
      "num_input_tokens_seen": 27065544,
      "step": 46630
    },
    {
      "epoch": 6.945933869526363,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.024750109194686603,
      "loss": 0.7843,
      "num_input_tokens_seen": 27068360,
      "step": 46635
    },
    {
      "epoch": 6.946678582067322,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.024748627516560054,
      "loss": 0.8289,
      "num_input_tokens_seen": 27071080,
      "step": 46640
    },
    {
      "epoch": 6.9474232946082815,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.024747145673740346,
      "loss": 0.776,
      "num_input_tokens_seen": 27073928,
      "step": 46645
    },
    {
      "epoch": 6.94816800714924,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.02474566366625252,
      "loss": 0.8064,
      "num_input_tokens_seen": 27077000,
      "step": 46650
    },
    {
      "epoch": 6.9489127196902,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.024744181494121612,
      "loss": 0.7652,
      "num_input_tokens_seen": 27079752,
      "step": 46655
    },
    {
      "epoch": 6.949657432231159,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.024742699157372663,
      "loss": 0.7825,
      "num_input_tokens_seen": 27082536,
      "step": 46660
    },
    {
      "epoch": 6.950402144772118,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.024741216656030712,
      "loss": 0.7823,
      "num_input_tokens_seen": 27085160,
      "step": 46665
    },
    {
      "epoch": 6.951146857313077,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.024739733990120805,
      "loss": 0.7911,
      "num_input_tokens_seen": 27088392,
      "step": 46670
    },
    {
      "epoch": 6.951891569854037,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.024738251159667993,
      "loss": 0.8407,
      "num_input_tokens_seen": 27091240,
      "step": 46675
    },
    {
      "epoch": 6.952636282394995,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.024736768164697324,
      "loss": 0.793,
      "num_input_tokens_seen": 27094344,
      "step": 46680
    },
    {
      "epoch": 6.953380994935955,
      "grad_norm": 0.015625,
      "learning_rate": 0.024735285005233854,
      "loss": 0.7862,
      "num_input_tokens_seen": 27097192,
      "step": 46685
    },
    {
      "epoch": 6.954125707476914,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02473380168130264,
      "loss": 0.823,
      "num_input_tokens_seen": 27099976,
      "step": 46690
    },
    {
      "epoch": 6.9548704200178735,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.024732318192928735,
      "loss": 0.7965,
      "num_input_tokens_seen": 27102792,
      "step": 46695
    },
    {
      "epoch": 6.955615132558832,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02473083454013721,
      "loss": 0.7945,
      "num_input_tokens_seen": 27105640,
      "step": 46700
    },
    {
      "epoch": 6.956359845099792,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.024729350722953118,
      "loss": 0.8382,
      "num_input_tokens_seen": 27108584,
      "step": 46705
    },
    {
      "epoch": 6.957104557640751,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.024727866741401545,
      "loss": 0.8016,
      "num_input_tokens_seen": 27111560,
      "step": 46710
    },
    {
      "epoch": 6.957849270181709,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02472638259550754,
      "loss": 0.8105,
      "num_input_tokens_seen": 27114472,
      "step": 46715
    },
    {
      "epoch": 6.958593982722669,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.024724898285296195,
      "loss": 0.8107,
      "num_input_tokens_seen": 27117000,
      "step": 46720
    },
    {
      "epoch": 6.959338695263629,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02472341381079257,
      "loss": 0.7848,
      "num_input_tokens_seen": 27119784,
      "step": 46725
    },
    {
      "epoch": 6.9600834078045875,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.024721929172021758,
      "loss": 0.8292,
      "num_input_tokens_seen": 27122760,
      "step": 46730
    },
    {
      "epoch": 6.960828120345546,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02472044436900883,
      "loss": 0.7836,
      "num_input_tokens_seen": 27125608,
      "step": 46735
    },
    {
      "epoch": 6.961572832886506,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.024718959401778876,
      "loss": 0.7822,
      "num_input_tokens_seen": 27128328,
      "step": 46740
    },
    {
      "epoch": 6.962317545427465,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.024717474270356976,
      "loss": 0.7966,
      "num_input_tokens_seen": 27131496,
      "step": 46745
    },
    {
      "epoch": 6.963062257968424,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.024715988974768226,
      "loss": 0.7834,
      "num_input_tokens_seen": 27134504,
      "step": 46750
    },
    {
      "epoch": 6.963806970509383,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02471450351503772,
      "loss": 0.7849,
      "num_input_tokens_seen": 27137480,
      "step": 46755
    },
    {
      "epoch": 6.964551683050343,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.024713017891190552,
      "loss": 0.8196,
      "num_input_tokens_seen": 27140680,
      "step": 46760
    },
    {
      "epoch": 6.965296395591301,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.024711532103251815,
      "loss": 0.7662,
      "num_input_tokens_seen": 27143560,
      "step": 46765
    },
    {
      "epoch": 6.966041108132261,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.024710046151246615,
      "loss": 0.838,
      "num_input_tokens_seen": 27146344,
      "step": 46770
    },
    {
      "epoch": 6.96678582067322,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.024708560035200052,
      "loss": 0.7875,
      "num_input_tokens_seen": 27149448,
      "step": 46775
    },
    {
      "epoch": 6.9675305332141795,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.024707073755137237,
      "loss": 0.7762,
      "num_input_tokens_seen": 27152200,
      "step": 46780
    },
    {
      "epoch": 6.968275245755138,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.024705587311083275,
      "loss": 0.7846,
      "num_input_tokens_seen": 27155048,
      "step": 46785
    },
    {
      "epoch": 6.969019958296098,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02470410070306328,
      "loss": 0.8056,
      "num_input_tokens_seen": 27157800,
      "step": 46790
    },
    {
      "epoch": 6.969764670837057,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.024702613931102364,
      "loss": 0.7811,
      "num_input_tokens_seen": 27160840,
      "step": 46795
    },
    {
      "epoch": 6.970509383378016,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.024701126995225652,
      "loss": 0.8296,
      "num_input_tokens_seen": 27163560,
      "step": 46800
    },
    {
      "epoch": 6.971254095918975,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02469963989545826,
      "loss": 0.8308,
      "num_input_tokens_seen": 27166312,
      "step": 46805
    },
    {
      "epoch": 6.971998808459935,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.02469815263182531,
      "loss": 0.8619,
      "num_input_tokens_seen": 27169448,
      "step": 46810
    },
    {
      "epoch": 6.9727435210008935,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02469666520435192,
      "loss": 0.7908,
      "num_input_tokens_seen": 27172168,
      "step": 46815
    },
    {
      "epoch": 6.973488233541853,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.024695177613063234,
      "loss": 0.8246,
      "num_input_tokens_seen": 27174984,
      "step": 46820
    },
    {
      "epoch": 6.974232946082812,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.024693689857984376,
      "loss": 0.8095,
      "num_input_tokens_seen": 27177896,
      "step": 46825
    },
    {
      "epoch": 6.9749776586237715,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.024692201939140474,
      "loss": 0.8212,
      "num_input_tokens_seen": 27180808,
      "step": 46830
    },
    {
      "epoch": 6.97572237116473,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02469071385655668,
      "loss": 0.7986,
      "num_input_tokens_seen": 27183496,
      "step": 46835
    },
    {
      "epoch": 6.97646708370569,
      "grad_norm": 0.01171875,
      "learning_rate": 0.024689225610258114,
      "loss": 0.7858,
      "num_input_tokens_seen": 27186440,
      "step": 46840
    },
    {
      "epoch": 6.977211796246649,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.024687737200269937,
      "loss": 0.7928,
      "num_input_tokens_seen": 27189352,
      "step": 46845
    },
    {
      "epoch": 6.977956508787608,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.024686248626617287,
      "loss": 0.7969,
      "num_input_tokens_seen": 27192072,
      "step": 46850
    },
    {
      "epoch": 6.978701221328567,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.024684759889325305,
      "loss": 0.8177,
      "num_input_tokens_seen": 27194984,
      "step": 46855
    },
    {
      "epoch": 6.979445933869527,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02468327098841915,
      "loss": 0.8073,
      "num_input_tokens_seen": 27197832,
      "step": 46860
    },
    {
      "epoch": 6.9801906464104855,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.024681781923923975,
      "loss": 0.7964,
      "num_input_tokens_seen": 27200712,
      "step": 46865
    },
    {
      "epoch": 6.980935358951445,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02468029269586493,
      "loss": 0.7975,
      "num_input_tokens_seen": 27203432,
      "step": 46870
    },
    {
      "epoch": 6.981680071492404,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.024678803304267184,
      "loss": 0.8022,
      "num_input_tokens_seen": 27206152,
      "step": 46875
    },
    {
      "epoch": 6.982424784033363,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.024677313749155893,
      "loss": 0.7957,
      "num_input_tokens_seen": 27209032,
      "step": 46880
    },
    {
      "epoch": 6.983169496574322,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.024675824030556216,
      "loss": 0.813,
      "num_input_tokens_seen": 27211752,
      "step": 46885
    },
    {
      "epoch": 6.983914209115282,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.024674334148493335,
      "loss": 0.8021,
      "num_input_tokens_seen": 27214600,
      "step": 46890
    },
    {
      "epoch": 6.984658921656241,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.024672844102992402,
      "loss": 0.8088,
      "num_input_tokens_seen": 27217864,
      "step": 46895
    },
    {
      "epoch": 6.9854036341971995,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02467135389407861,
      "loss": 0.8065,
      "num_input_tokens_seen": 27220712,
      "step": 46900
    },
    {
      "epoch": 6.986148346738159,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.02466986352177711,
      "loss": 0.7896,
      "num_input_tokens_seen": 27223976,
      "step": 46905
    },
    {
      "epoch": 6.986893059279119,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.024668372986113102,
      "loss": 0.8062,
      "num_input_tokens_seen": 27226728,
      "step": 46910
    },
    {
      "epoch": 6.9876377718200775,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.024666882287111756,
      "loss": 0.7815,
      "num_input_tokens_seen": 27229544,
      "step": 46915
    },
    {
      "epoch": 6.988382484361036,
      "grad_norm": 0.01165771484375,
      "learning_rate": 0.02466539142479826,
      "loss": 0.8142,
      "num_input_tokens_seen": 27232392,
      "step": 46920
    },
    {
      "epoch": 6.989127196901996,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.024663900399197802,
      "loss": 0.8046,
      "num_input_tokens_seen": 27235272,
      "step": 46925
    },
    {
      "epoch": 6.989871909442955,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.024662409210335565,
      "loss": 0.8133,
      "num_input_tokens_seen": 27238312,
      "step": 46930
    },
    {
      "epoch": 6.990616621983914,
      "grad_norm": 0.021484375,
      "learning_rate": 0.024660917858236747,
      "loss": 0.8031,
      "num_input_tokens_seen": 27241096,
      "step": 46935
    },
    {
      "epoch": 6.991361334524873,
      "grad_norm": 0.0234375,
      "learning_rate": 0.024659426342926536,
      "loss": 0.7813,
      "num_input_tokens_seen": 27244040,
      "step": 46940
    },
    {
      "epoch": 6.992106047065833,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02465793466443014,
      "loss": 0.7834,
      "num_input_tokens_seen": 27246760,
      "step": 46945
    },
    {
      "epoch": 6.9928507596067915,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.024656442822772753,
      "loss": 0.7749,
      "num_input_tokens_seen": 27249864,
      "step": 46950
    },
    {
      "epoch": 6.993595472147751,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02465495081797958,
      "loss": 0.7956,
      "num_input_tokens_seen": 27253000,
      "step": 46955
    },
    {
      "epoch": 6.99434018468871,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02465345865007582,
      "loss": 0.8189,
      "num_input_tokens_seen": 27255848,
      "step": 46960
    },
    {
      "epoch": 6.9950848972296695,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02465196631908669,
      "loss": 0.7817,
      "num_input_tokens_seen": 27258728,
      "step": 46965
    },
    {
      "epoch": 6.995829609770628,
      "grad_norm": 0.021484375,
      "learning_rate": 0.024650473825037402,
      "loss": 0.7814,
      "num_input_tokens_seen": 27261832,
      "step": 46970
    },
    {
      "epoch": 6.996574322311588,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.024648981167953162,
      "loss": 0.784,
      "num_input_tokens_seen": 27264776,
      "step": 46975
    },
    {
      "epoch": 6.997319034852547,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.024647488347859195,
      "loss": 0.8158,
      "num_input_tokens_seen": 27268008,
      "step": 46980
    },
    {
      "epoch": 6.998063747393506,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.02464599536478072,
      "loss": 0.8042,
      "num_input_tokens_seen": 27271176,
      "step": 46985
    },
    {
      "epoch": 6.998808459934465,
      "grad_norm": 0.02734375,
      "learning_rate": 0.024644502218742953,
      "loss": 0.8249,
      "num_input_tokens_seen": 27274408,
      "step": 46990
    },
    {
      "epoch": 6.999553172475425,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.024643008909771124,
      "loss": 0.8242,
      "num_input_tokens_seen": 27277416,
      "step": 46995
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.80185866355896,
      "eval_runtime": 70.7775,
      "eval_samples_per_second": 42.16,
      "eval_steps_per_second": 10.54,
      "num_input_tokens_seen": 27278720,
      "step": 46998
    },
    {
      "epoch": 7.0002978850163835,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.02464151543789046,
      "loss": 0.8143,
      "num_input_tokens_seen": 27279936,
      "step": 47000
    },
    {
      "epoch": 7.001042597557343,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.024640021803126196,
      "loss": 0.8251,
      "num_input_tokens_seen": 27282816,
      "step": 47005
    },
    {
      "epoch": 7.001787310098302,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.024638528005503556,
      "loss": 0.8191,
      "num_input_tokens_seen": 27285568,
      "step": 47010
    },
    {
      "epoch": 7.0025320226392616,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.024637034045047784,
      "loss": 0.7959,
      "num_input_tokens_seen": 27288736,
      "step": 47015
    },
    {
      "epoch": 7.00327673518022,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.024635539921784116,
      "loss": 0.8058,
      "num_input_tokens_seen": 27291392,
      "step": 47020
    },
    {
      "epoch": 7.00402144772118,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.024634045635737793,
      "loss": 0.8181,
      "num_input_tokens_seen": 27294080,
      "step": 47025
    },
    {
      "epoch": 7.004766160262139,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02463255118693406,
      "loss": 0.8061,
      "num_input_tokens_seen": 27297024,
      "step": 47030
    },
    {
      "epoch": 7.005510872803098,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.024631056575398172,
      "loss": 0.8015,
      "num_input_tokens_seen": 27300288,
      "step": 47035
    },
    {
      "epoch": 7.006255585344057,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.024629561801155362,
      "loss": 0.7943,
      "num_input_tokens_seen": 27303296,
      "step": 47040
    },
    {
      "epoch": 7.007000297885017,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.024628066864230896,
      "loss": 0.8165,
      "num_input_tokens_seen": 27306336,
      "step": 47045
    },
    {
      "epoch": 7.0077450104259755,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.024626571764650027,
      "loss": 0.8113,
      "num_input_tokens_seen": 27309280,
      "step": 47050
    },
    {
      "epoch": 7.008489722966935,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.024625076502438014,
      "loss": 0.8226,
      "num_input_tokens_seen": 27312096,
      "step": 47055
    },
    {
      "epoch": 7.009234435507894,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.024623581077620114,
      "loss": 0.815,
      "num_input_tokens_seen": 27314976,
      "step": 47060
    },
    {
      "epoch": 7.009979148048854,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02462208549022159,
      "loss": 0.7977,
      "num_input_tokens_seen": 27317824,
      "step": 47065
    },
    {
      "epoch": 7.010723860589812,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.024620589740267713,
      "loss": 0.8132,
      "num_input_tokens_seen": 27320608,
      "step": 47070
    },
    {
      "epoch": 7.011468573130771,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02461909382778375,
      "loss": 0.8209,
      "num_input_tokens_seen": 27324032,
      "step": 47075
    },
    {
      "epoch": 7.012213285671731,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.024617597752794974,
      "loss": 0.7982,
      "num_input_tokens_seen": 27327008,
      "step": 47080
    },
    {
      "epoch": 7.0129579982126895,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.024616101515326658,
      "loss": 0.7927,
      "num_input_tokens_seen": 27329664,
      "step": 47085
    },
    {
      "epoch": 7.013702710753649,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02461460511540408,
      "loss": 0.8121,
      "num_input_tokens_seen": 27332960,
      "step": 47090
    },
    {
      "epoch": 7.014447423294608,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.024613108553052522,
      "loss": 0.8016,
      "num_input_tokens_seen": 27336032,
      "step": 47095
    },
    {
      "epoch": 7.0151921358355676,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.024611611828297266,
      "loss": 0.8038,
      "num_input_tokens_seen": 27338944,
      "step": 47100
    },
    {
      "epoch": 7.015936848376526,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.024610114941163594,
      "loss": 0.8052,
      "num_input_tokens_seen": 27341632,
      "step": 47105
    },
    {
      "epoch": 7.016681560917486,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.024608617891676798,
      "loss": 0.7876,
      "num_input_tokens_seen": 27344352,
      "step": 47110
    },
    {
      "epoch": 7.017426273458445,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02460712067986217,
      "loss": 0.8065,
      "num_input_tokens_seen": 27347520,
      "step": 47115
    },
    {
      "epoch": 7.018170985999404,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.024605623305745006,
      "loss": 0.7948,
      "num_input_tokens_seen": 27350208,
      "step": 47120
    },
    {
      "epoch": 7.018915698540363,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.02460412576935059,
      "loss": 0.8037,
      "num_input_tokens_seen": 27352960,
      "step": 47125
    },
    {
      "epoch": 7.019660411081323,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.024602628070704236,
      "loss": 0.8054,
      "num_input_tokens_seen": 27355584,
      "step": 47130
    },
    {
      "epoch": 7.0204051236222815,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02460113020983124,
      "loss": 0.8102,
      "num_input_tokens_seen": 27358624,
      "step": 47135
    },
    {
      "epoch": 7.021149836163241,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02459963218675691,
      "loss": 0.8065,
      "num_input_tokens_seen": 27361440,
      "step": 47140
    },
    {
      "epoch": 7.0218945487042,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.024598134001506546,
      "loss": 0.8133,
      "num_input_tokens_seen": 27364320,
      "step": 47145
    },
    {
      "epoch": 7.02263926124516,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.024596635654105463,
      "loss": 0.7903,
      "num_input_tokens_seen": 27367136,
      "step": 47150
    },
    {
      "epoch": 7.023383973786118,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.024595137144578973,
      "loss": 0.7849,
      "num_input_tokens_seen": 27370432,
      "step": 47155
    },
    {
      "epoch": 7.024128686327078,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.024593638472952397,
      "loss": 0.82,
      "num_input_tokens_seen": 27373664,
      "step": 47160
    },
    {
      "epoch": 7.024873398868037,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.02459213963925105,
      "loss": 0.7968,
      "num_input_tokens_seen": 27376448,
      "step": 47165
    },
    {
      "epoch": 7.025618111408996,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02459064064350025,
      "loss": 0.8018,
      "num_input_tokens_seen": 27379392,
      "step": 47170
    },
    {
      "epoch": 7.026362823949955,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.024589141485725326,
      "loss": 0.8048,
      "num_input_tokens_seen": 27382240,
      "step": 47175
    },
    {
      "epoch": 7.027107536490915,
      "grad_norm": 0.01092529296875,
      "learning_rate": 0.024587642165951603,
      "loss": 0.8081,
      "num_input_tokens_seen": 27385472,
      "step": 47180
    },
    {
      "epoch": 7.0278522490318736,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.024586142684204407,
      "loss": 0.8034,
      "num_input_tokens_seen": 27388256,
      "step": 47185
    },
    {
      "epoch": 7.028596961572833,
      "grad_norm": 0.037109375,
      "learning_rate": 0.024584643040509075,
      "loss": 0.8001,
      "num_input_tokens_seen": 27391072,
      "step": 47190
    },
    {
      "epoch": 7.029341674113792,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.024583143234890945,
      "loss": 0.7864,
      "num_input_tokens_seen": 27394048,
      "step": 47195
    },
    {
      "epoch": 7.030086386654752,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.024581643267375344,
      "loss": 0.7895,
      "num_input_tokens_seen": 27397024,
      "step": 47200
    },
    {
      "epoch": 7.03083109919571,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.024580143137987617,
      "loss": 0.7925,
      "num_input_tokens_seen": 27400096,
      "step": 47205
    },
    {
      "epoch": 7.03157581173667,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.024578642846753113,
      "loss": 0.7995,
      "num_input_tokens_seen": 27402784,
      "step": 47210
    },
    {
      "epoch": 7.032320524277629,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.024577142393697172,
      "loss": 0.7944,
      "num_input_tokens_seen": 27405280,
      "step": 47215
    },
    {
      "epoch": 7.033065236818588,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02457564177884514,
      "loss": 0.7842,
      "num_input_tokens_seen": 27408704,
      "step": 47220
    },
    {
      "epoch": 7.033809949359547,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.024574141002222376,
      "loss": 0.8129,
      "num_input_tokens_seen": 27411392,
      "step": 47225
    },
    {
      "epoch": 7.034554661900507,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02457264006385423,
      "loss": 0.8168,
      "num_input_tokens_seen": 27414208,
      "step": 47230
    },
    {
      "epoch": 7.035299374441466,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02457113896376606,
      "loss": 0.8118,
      "num_input_tokens_seen": 27416992,
      "step": 47235
    },
    {
      "epoch": 7.036044086982425,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.024569637701983223,
      "loss": 0.803,
      "num_input_tokens_seen": 27419840,
      "step": 47240
    },
    {
      "epoch": 7.036788799523384,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.024568136278531084,
      "loss": 0.806,
      "num_input_tokens_seen": 27422752,
      "step": 47245
    },
    {
      "epoch": 7.037533512064343,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02456663469343501,
      "loss": 0.7955,
      "num_input_tokens_seen": 27425504,
      "step": 47250
    },
    {
      "epoch": 7.038278224605302,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.024565132946720365,
      "loss": 0.8046,
      "num_input_tokens_seen": 27428384,
      "step": 47255
    },
    {
      "epoch": 7.039022937146261,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02456363103841252,
      "loss": 0.7827,
      "num_input_tokens_seen": 27431360,
      "step": 47260
    },
    {
      "epoch": 7.039767649687221,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.024562128968536847,
      "loss": 0.7984,
      "num_input_tokens_seen": 27434208,
      "step": 47265
    },
    {
      "epoch": 7.0405123622281796,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.024560626737118722,
      "loss": 0.8018,
      "num_input_tokens_seen": 27436864,
      "step": 47270
    },
    {
      "epoch": 7.041257074769139,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.024559124344183527,
      "loss": 0.8205,
      "num_input_tokens_seen": 27439712,
      "step": 47275
    },
    {
      "epoch": 7.042001787310098,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.024557621789756644,
      "loss": 0.7816,
      "num_input_tokens_seen": 27442464,
      "step": 47280
    },
    {
      "epoch": 7.042746499851058,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.024556119073863455,
      "loss": 0.8123,
      "num_input_tokens_seen": 27445184,
      "step": 47285
    },
    {
      "epoch": 7.043491212392016,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.02455461619652934,
      "loss": 0.8067,
      "num_input_tokens_seen": 27447840,
      "step": 47290
    },
    {
      "epoch": 7.044235924932976,
      "grad_norm": 0.01953125,
      "learning_rate": 0.024553113157779707,
      "loss": 0.7933,
      "num_input_tokens_seen": 27451008,
      "step": 47295
    },
    {
      "epoch": 7.044980637473935,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02455160995763993,
      "loss": 0.7961,
      "num_input_tokens_seen": 27454400,
      "step": 47300
    },
    {
      "epoch": 7.045725350014894,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.024550106596135404,
      "loss": 0.8049,
      "num_input_tokens_seen": 27457280,
      "step": 47305
    },
    {
      "epoch": 7.046470062555853,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02454860307329154,
      "loss": 0.7932,
      "num_input_tokens_seen": 27460256,
      "step": 47310
    },
    {
      "epoch": 7.047214775096813,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.024547099389133735,
      "loss": 0.8189,
      "num_input_tokens_seen": 27463072,
      "step": 47315
    },
    {
      "epoch": 7.047959487637772,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.024545595543687387,
      "loss": 0.7998,
      "num_input_tokens_seen": 27465728,
      "step": 47320
    },
    {
      "epoch": 7.048704200178731,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.024544091536977897,
      "loss": 0.7797,
      "num_input_tokens_seen": 27469024,
      "step": 47325
    },
    {
      "epoch": 7.04944891271969,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02454258736903069,
      "loss": 0.7843,
      "num_input_tokens_seen": 27471904,
      "step": 47330
    },
    {
      "epoch": 7.05019362526065,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.024541083039871167,
      "loss": 0.7986,
      "num_input_tokens_seen": 27474880,
      "step": 47335
    },
    {
      "epoch": 7.050938337801608,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.024539578549524734,
      "loss": 0.8196,
      "num_input_tokens_seen": 27477888,
      "step": 47340
    },
    {
      "epoch": 7.051683050342568,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.024538073898016828,
      "loss": 0.7845,
      "num_input_tokens_seen": 27480672,
      "step": 47345
    },
    {
      "epoch": 7.052427762883527,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.024536569085372852,
      "loss": 0.8197,
      "num_input_tokens_seen": 27483424,
      "step": 47350
    },
    {
      "epoch": 7.053172475424486,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.024535064111618234,
      "loss": 0.7975,
      "num_input_tokens_seen": 27486176,
      "step": 47355
    },
    {
      "epoch": 7.053917187965445,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.024533558976778398,
      "loss": 0.8059,
      "num_input_tokens_seen": 27489120,
      "step": 47360
    },
    {
      "epoch": 7.054661900506405,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02453205368087878,
      "loss": 0.8052,
      "num_input_tokens_seen": 27491968,
      "step": 47365
    },
    {
      "epoch": 7.055406613047364,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02453054822394479,
      "loss": 0.7862,
      "num_input_tokens_seen": 27494816,
      "step": 47370
    },
    {
      "epoch": 7.056151325588323,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.024529042606001885,
      "loss": 0.7822,
      "num_input_tokens_seen": 27497792,
      "step": 47375
    },
    {
      "epoch": 7.056896038129282,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.024527536827075487,
      "loss": 0.8239,
      "num_input_tokens_seen": 27500576,
      "step": 47380
    },
    {
      "epoch": 7.057640750670242,
      "grad_norm": 0.021484375,
      "learning_rate": 0.024526030887191033,
      "loss": 0.7952,
      "num_input_tokens_seen": 27503456,
      "step": 47385
    },
    {
      "epoch": 7.0583854632112,
      "grad_norm": 0.03515625,
      "learning_rate": 0.024524524786373975,
      "loss": 0.7915,
      "num_input_tokens_seen": 27506272,
      "step": 47390
    },
    {
      "epoch": 7.05913017575216,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.02452301852464975,
      "loss": 0.7891,
      "num_input_tokens_seen": 27509184,
      "step": 47395
    },
    {
      "epoch": 7.059874888293119,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.024521512102043803,
      "loss": 0.8332,
      "num_input_tokens_seen": 27511808,
      "step": 47400
    },
    {
      "epoch": 7.0606196008340785,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02452000551858159,
      "loss": 0.7989,
      "num_input_tokens_seen": 27514848,
      "step": 47405
    },
    {
      "epoch": 7.061364313375037,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.024518498774288553,
      "loss": 0.7862,
      "num_input_tokens_seen": 27517632,
      "step": 47410
    },
    {
      "epoch": 7.062109025915996,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.024516991869190166,
      "loss": 0.7874,
      "num_input_tokens_seen": 27520640,
      "step": 47415
    },
    {
      "epoch": 7.062853738456956,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.024515484803311862,
      "loss": 0.8035,
      "num_input_tokens_seen": 27523648,
      "step": 47420
    },
    {
      "epoch": 7.063598450997914,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.024513977576679117,
      "loss": 0.8158,
      "num_input_tokens_seen": 27526400,
      "step": 47425
    },
    {
      "epoch": 7.064343163538874,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.024512470189317393,
      "loss": 0.8021,
      "num_input_tokens_seen": 27529376,
      "step": 47430
    },
    {
      "epoch": 7.065087876079833,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.02451096264125215,
      "loss": 0.7759,
      "num_input_tokens_seen": 27531968,
      "step": 47435
    },
    {
      "epoch": 7.065832588620792,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02450945493250886,
      "loss": 0.7985,
      "num_input_tokens_seen": 27534784,
      "step": 47440
    },
    {
      "epoch": 7.066577301161751,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.024507947063112993,
      "loss": 0.824,
      "num_input_tokens_seen": 27537920,
      "step": 47445
    },
    {
      "epoch": 7.067322013702711,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02450643903309003,
      "loss": 0.8186,
      "num_input_tokens_seen": 27540800,
      "step": 47450
    },
    {
      "epoch": 7.06806672624367,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.024504930842465432,
      "loss": 0.7912,
      "num_input_tokens_seen": 27543360,
      "step": 47455
    },
    {
      "epoch": 7.068811438784629,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.024503422491264696,
      "loss": 0.8061,
      "num_input_tokens_seen": 27546240,
      "step": 47460
    },
    {
      "epoch": 7.069556151325588,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02450191397951329,
      "loss": 0.806,
      "num_input_tokens_seen": 27549248,
      "step": 47465
    },
    {
      "epoch": 7.070300863866548,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.024500405307236708,
      "loss": 0.7849,
      "num_input_tokens_seen": 27552512,
      "step": 47470
    },
    {
      "epoch": 7.071045576407506,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.024498896474460432,
      "loss": 0.7811,
      "num_input_tokens_seen": 27555360,
      "step": 47475
    },
    {
      "epoch": 7.071790288948466,
      "grad_norm": 0.017578125,
      "learning_rate": 0.024497387481209956,
      "loss": 0.7837,
      "num_input_tokens_seen": 27558432,
      "step": 47480
    },
    {
      "epoch": 7.072535001489425,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.024495878327510767,
      "loss": 0.791,
      "num_input_tokens_seen": 27561184,
      "step": 47485
    },
    {
      "epoch": 7.0732797140303845,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.024494369013388372,
      "loss": 0.7767,
      "num_input_tokens_seen": 27564288,
      "step": 47490
    },
    {
      "epoch": 7.074024426571343,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02449285953886826,
      "loss": 0.7883,
      "num_input_tokens_seen": 27567008,
      "step": 47495
    },
    {
      "epoch": 7.074769139112303,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.024491349903975935,
      "loss": 0.7779,
      "num_input_tokens_seen": 27569920,
      "step": 47500
    },
    {
      "epoch": 7.075513851653262,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.024489840108736897,
      "loss": 0.8087,
      "num_input_tokens_seen": 27572864,
      "step": 47505
    },
    {
      "epoch": 7.076258564194221,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.024488330153176655,
      "loss": 0.7846,
      "num_input_tokens_seen": 27575776,
      "step": 47510
    },
    {
      "epoch": 7.07700327673518,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.024486820037320722,
      "loss": 0.83,
      "num_input_tokens_seen": 27578560,
      "step": 47515
    },
    {
      "epoch": 7.07774798927614,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.0244853097611946,
      "loss": 0.8349,
      "num_input_tokens_seen": 27581600,
      "step": 47520
    },
    {
      "epoch": 7.078492701817098,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.024483799324823818,
      "loss": 0.8004,
      "num_input_tokens_seen": 27584640,
      "step": 47525
    },
    {
      "epoch": 7.079237414358058,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.024482288728233877,
      "loss": 0.8156,
      "num_input_tokens_seen": 27587616,
      "step": 47530
    },
    {
      "epoch": 7.079982126899017,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.024480777971450315,
      "loss": 0.8018,
      "num_input_tokens_seen": 27590368,
      "step": 47535
    },
    {
      "epoch": 7.0807268394399765,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.024479267054498637,
      "loss": 0.8187,
      "num_input_tokens_seen": 27593152,
      "step": 47540
    },
    {
      "epoch": 7.081471551980935,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02447775597740438,
      "loss": 0.778,
      "num_input_tokens_seen": 27596064,
      "step": 47545
    },
    {
      "epoch": 7.082216264521895,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.024476244740193068,
      "loss": 0.7723,
      "num_input_tokens_seen": 27598912,
      "step": 47550
    },
    {
      "epoch": 7.082960977062854,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.024474733342890232,
      "loss": 0.7967,
      "num_input_tokens_seen": 27602016,
      "step": 47555
    },
    {
      "epoch": 7.083705689603813,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.024473221785521405,
      "loss": 0.7984,
      "num_input_tokens_seen": 27604608,
      "step": 47560
    },
    {
      "epoch": 7.084450402144772,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.024471710068112124,
      "loss": 0.791,
      "num_input_tokens_seen": 27607520,
      "step": 47565
    },
    {
      "epoch": 7.085195114685732,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02447019819068793,
      "loss": 0.833,
      "num_input_tokens_seen": 27610784,
      "step": 47570
    },
    {
      "epoch": 7.0859398272266905,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.024468686153274365,
      "loss": 0.8182,
      "num_input_tokens_seen": 27613632,
      "step": 47575
    },
    {
      "epoch": 7.08668453976765,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.024467173955896968,
      "loss": 0.7821,
      "num_input_tokens_seen": 27616448,
      "step": 47580
    },
    {
      "epoch": 7.087429252308609,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02446566159858129,
      "loss": 0.7851,
      "num_input_tokens_seen": 27619392,
      "step": 47585
    },
    {
      "epoch": 7.088173964849568,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.024464149081352877,
      "loss": 0.7726,
      "num_input_tokens_seen": 27622304,
      "step": 47590
    },
    {
      "epoch": 7.088918677390527,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.024462636404237285,
      "loss": 0.7994,
      "num_input_tokens_seen": 27625120,
      "step": 47595
    },
    {
      "epoch": 7.089663389931486,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.024461123567260067,
      "loss": 0.8103,
      "num_input_tokens_seen": 27628576,
      "step": 47600
    },
    {
      "epoch": 7.090408102472446,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.024459610570446783,
      "loss": 0.7908,
      "num_input_tokens_seen": 27631296,
      "step": 47605
    },
    {
      "epoch": 7.091152815013404,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.024458097413822998,
      "loss": 0.7591,
      "num_input_tokens_seen": 27634048,
      "step": 47610
    },
    {
      "epoch": 7.091897527554364,
      "grad_norm": 0.01171875,
      "learning_rate": 0.024456584097414264,
      "loss": 0.8001,
      "num_input_tokens_seen": 27637216,
      "step": 47615
    },
    {
      "epoch": 7.092642240095323,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.024455070621246154,
      "loss": 0.8099,
      "num_input_tokens_seen": 27640160,
      "step": 47620
    },
    {
      "epoch": 7.0933869526362825,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.024453556985344238,
      "loss": 0.7644,
      "num_input_tokens_seen": 27642944,
      "step": 47625
    },
    {
      "epoch": 7.094131665177241,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.024452043189734082,
      "loss": 0.7787,
      "num_input_tokens_seen": 27645856,
      "step": 47630
    },
    {
      "epoch": 7.094876377718201,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.024450529234441264,
      "loss": 0.8402,
      "num_input_tokens_seen": 27648800,
      "step": 47635
    },
    {
      "epoch": 7.09562109025916,
      "grad_norm": 0.025390625,
      "learning_rate": 0.024449015119491362,
      "loss": 0.8302,
      "num_input_tokens_seen": 27651680,
      "step": 47640
    },
    {
      "epoch": 7.096365802800119,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.024447500844909948,
      "loss": 0.771,
      "num_input_tokens_seen": 27654624,
      "step": 47645
    },
    {
      "epoch": 7.097110515341078,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.024445986410722615,
      "loss": 0.7844,
      "num_input_tokens_seen": 27657504,
      "step": 47650
    },
    {
      "epoch": 7.097855227882038,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.024444471816954937,
      "loss": 0.7874,
      "num_input_tokens_seen": 27660224,
      "step": 47655
    },
    {
      "epoch": 7.0985999404229965,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.02444295706363251,
      "loss": 0.7836,
      "num_input_tokens_seen": 27663232,
      "step": 47660
    },
    {
      "epoch": 7.099344652963956,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02444144215078092,
      "loss": 0.7897,
      "num_input_tokens_seen": 27665984,
      "step": 47665
    },
    {
      "epoch": 7.100089365504915,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02443992707842576,
      "loss": 0.7976,
      "num_input_tokens_seen": 27669152,
      "step": 47670
    },
    {
      "epoch": 7.1008340780458745,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02443841184659263,
      "loss": 0.8127,
      "num_input_tokens_seen": 27672032,
      "step": 47675
    },
    {
      "epoch": 7.101578790586833,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.024436896455307118,
      "loss": 0.8191,
      "num_input_tokens_seen": 27674720,
      "step": 47680
    },
    {
      "epoch": 7.102323503127793,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02443538090459484,
      "loss": 0.7961,
      "num_input_tokens_seen": 27677984,
      "step": 47685
    },
    {
      "epoch": 7.103068215668752,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.024433865194481385,
      "loss": 0.7926,
      "num_input_tokens_seen": 27680768,
      "step": 47690
    },
    {
      "epoch": 7.103812928209711,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.024432349324992365,
      "loss": 0.8227,
      "num_input_tokens_seen": 27683264,
      "step": 47695
    },
    {
      "epoch": 7.10455764075067,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.024430833296153392,
      "loss": 0.7816,
      "num_input_tokens_seen": 27686304,
      "step": 47700
    },
    {
      "epoch": 7.10530235329163,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.024429317107990078,
      "loss": 0.8005,
      "num_input_tokens_seen": 27689280,
      "step": 47705
    },
    {
      "epoch": 7.1060470658325885,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.024427800760528032,
      "loss": 0.799,
      "num_input_tokens_seen": 27692160,
      "step": 47710
    },
    {
      "epoch": 7.106791778373548,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02442628425379287,
      "loss": 0.8642,
      "num_input_tokens_seen": 27694816,
      "step": 47715
    },
    {
      "epoch": 7.107536490914507,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02442476758781022,
      "loss": 0.7983,
      "num_input_tokens_seen": 27697760,
      "step": 47720
    },
    {
      "epoch": 7.1082812034554665,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.024423250762605705,
      "loss": 0.7937,
      "num_input_tokens_seen": 27701024,
      "step": 47725
    },
    {
      "epoch": 7.109025915996425,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02442173377820494,
      "loss": 0.7896,
      "num_input_tokens_seen": 27703712,
      "step": 47730
    },
    {
      "epoch": 7.109770628537385,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.02442021663463356,
      "loss": 0.7984,
      "num_input_tokens_seen": 27706592,
      "step": 47735
    },
    {
      "epoch": 7.110515341078344,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.024418699331917197,
      "loss": 0.8064,
      "num_input_tokens_seen": 27709408,
      "step": 47740
    },
    {
      "epoch": 7.111260053619303,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02441718187008148,
      "loss": 0.8,
      "num_input_tokens_seen": 27712160,
      "step": 47745
    },
    {
      "epoch": 7.112004766160262,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.024415664249152043,
      "loss": 0.8316,
      "num_input_tokens_seen": 27715232,
      "step": 47750
    },
    {
      "epoch": 7.112749478701222,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.024414146469154532,
      "loss": 0.8125,
      "num_input_tokens_seen": 27718208,
      "step": 47755
    },
    {
      "epoch": 7.1134941912421805,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.024412628530114583,
      "loss": 0.8178,
      "num_input_tokens_seen": 27721056,
      "step": 47760
    },
    {
      "epoch": 7.114238903783139,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02441111043205784,
      "loss": 0.8416,
      "num_input_tokens_seen": 27724000,
      "step": 47765
    },
    {
      "epoch": 7.114983616324099,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.024409592175009952,
      "loss": 0.8186,
      "num_input_tokens_seen": 27727008,
      "step": 47770
    },
    {
      "epoch": 7.115728328865058,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.024408073758996573,
      "loss": 0.8156,
      "num_input_tokens_seen": 27730176,
      "step": 47775
    },
    {
      "epoch": 7.116473041406017,
      "grad_norm": 0.01953125,
      "learning_rate": 0.024406555184043346,
      "loss": 0.798,
      "num_input_tokens_seen": 27732928,
      "step": 47780
    },
    {
      "epoch": 7.117217753946976,
      "grad_norm": 0.01324462890625,
      "learning_rate": 0.024405036450175926,
      "loss": 0.8133,
      "num_input_tokens_seen": 27736160,
      "step": 47785
    },
    {
      "epoch": 7.117962466487936,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.024403517557419978,
      "loss": 0.8037,
      "num_input_tokens_seen": 27739168,
      "step": 47790
    },
    {
      "epoch": 7.1187071790288945,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02440199850580116,
      "loss": 0.8138,
      "num_input_tokens_seen": 27742304,
      "step": 47795
    },
    {
      "epoch": 7.119451891569854,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02440047929534513,
      "loss": 0.7851,
      "num_input_tokens_seen": 27745120,
      "step": 47800
    },
    {
      "epoch": 7.120196604110813,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.024398959926077556,
      "loss": 0.7869,
      "num_input_tokens_seen": 27747968,
      "step": 47805
    },
    {
      "epoch": 7.1209413166517725,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.024397440398024106,
      "loss": 0.7918,
      "num_input_tokens_seen": 27750816,
      "step": 47810
    },
    {
      "epoch": 7.121686029192731,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.024395920711210453,
      "loss": 0.7971,
      "num_input_tokens_seen": 27753728,
      "step": 47815
    },
    {
      "epoch": 7.122430741733691,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02439440086566227,
      "loss": 0.8057,
      "num_input_tokens_seen": 27756544,
      "step": 47820
    },
    {
      "epoch": 7.12317545427465,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.024392880861405234,
      "loss": 0.8026,
      "num_input_tokens_seen": 27759424,
      "step": 47825
    },
    {
      "epoch": 7.123920166815609,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02439136069846502,
      "loss": 0.7985,
      "num_input_tokens_seen": 27762112,
      "step": 47830
    },
    {
      "epoch": 7.124664879356568,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02438984037686731,
      "loss": 0.7899,
      "num_input_tokens_seen": 27764896,
      "step": 47835
    },
    {
      "epoch": 7.125409591897528,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02438831989663779,
      "loss": 0.798,
      "num_input_tokens_seen": 27767808,
      "step": 47840
    },
    {
      "epoch": 7.1261543044384865,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.024386799257802152,
      "loss": 0.7796,
      "num_input_tokens_seen": 27770656,
      "step": 47845
    },
    {
      "epoch": 7.126899016979446,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.024385278460386076,
      "loss": 0.7863,
      "num_input_tokens_seen": 27773472,
      "step": 47850
    },
    {
      "epoch": 7.127643729520405,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.024383757504415265,
      "loss": 0.8001,
      "num_input_tokens_seen": 27776448,
      "step": 47855
    },
    {
      "epoch": 7.128388442061365,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.024382236389915404,
      "loss": 0.7893,
      "num_input_tokens_seen": 27779040,
      "step": 47860
    },
    {
      "epoch": 7.129133154602323,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.0243807151169122,
      "loss": 0.8586,
      "num_input_tokens_seen": 27781664,
      "step": 47865
    },
    {
      "epoch": 7.129877867143283,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.024379193685431337,
      "loss": 0.8167,
      "num_input_tokens_seen": 27784448,
      "step": 47870
    },
    {
      "epoch": 7.130622579684242,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.02437767209549854,
      "loss": 0.7899,
      "num_input_tokens_seen": 27787296,
      "step": 47875
    },
    {
      "epoch": 7.131367292225201,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.024376150347139497,
      "loss": 0.8284,
      "num_input_tokens_seen": 27790208,
      "step": 47880
    },
    {
      "epoch": 7.13211200476616,
      "grad_norm": 0.01177978515625,
      "learning_rate": 0.02437462844037993,
      "loss": 0.793,
      "num_input_tokens_seen": 27793216,
      "step": 47885
    },
    {
      "epoch": 7.13285671730712,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.024373106375245538,
      "loss": 0.8321,
      "num_input_tokens_seen": 27796256,
      "step": 47890
    },
    {
      "epoch": 7.1336014298480785,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.024371584151762042,
      "loss": 0.8027,
      "num_input_tokens_seen": 27799104,
      "step": 47895
    },
    {
      "epoch": 7.134346142389038,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.024370061769955157,
      "loss": 0.7963,
      "num_input_tokens_seen": 27801824,
      "step": 47900
    },
    {
      "epoch": 7.135090854929997,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.024368539229850605,
      "loss": 0.7956,
      "num_input_tokens_seen": 27804608,
      "step": 47905
    },
    {
      "epoch": 7.135835567470957,
      "grad_norm": 0.029296875,
      "learning_rate": 0.0243670165314741,
      "loss": 0.7958,
      "num_input_tokens_seen": 27807648,
      "step": 47910
    },
    {
      "epoch": 7.136580280011915,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.02436549367485137,
      "loss": 0.7782,
      "num_input_tokens_seen": 27810240,
      "step": 47915
    },
    {
      "epoch": 7.137324992552875,
      "grad_norm": 0.017578125,
      "learning_rate": 0.024363970660008144,
      "loss": 0.8311,
      "num_input_tokens_seen": 27813216,
      "step": 47920
    },
    {
      "epoch": 7.138069705093834,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.024362447486970154,
      "loss": 0.7999,
      "num_input_tokens_seen": 27816352,
      "step": 47925
    },
    {
      "epoch": 7.1388144176347925,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.024360924155763124,
      "loss": 0.7914,
      "num_input_tokens_seen": 27819360,
      "step": 47930
    },
    {
      "epoch": 7.139559130175752,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.0243594006664128,
      "loss": 0.7969,
      "num_input_tokens_seen": 27822432,
      "step": 47935
    },
    {
      "epoch": 7.140303842716711,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02435787701894491,
      "loss": 0.8001,
      "num_input_tokens_seen": 27825376,
      "step": 47940
    },
    {
      "epoch": 7.141048555257671,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.024356353213385203,
      "loss": 0.7955,
      "num_input_tokens_seen": 27828384,
      "step": 47945
    },
    {
      "epoch": 7.141793267798629,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.024354829249759416,
      "loss": 0.8046,
      "num_input_tokens_seen": 27831328,
      "step": 47950
    },
    {
      "epoch": 7.142537980339589,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.024353305128093295,
      "loss": 0.792,
      "num_input_tokens_seen": 27834016,
      "step": 47955
    },
    {
      "epoch": 7.143282692880548,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02435178084841259,
      "loss": 0.771,
      "num_input_tokens_seen": 27836672,
      "step": 47960
    },
    {
      "epoch": 7.144027405421507,
      "grad_norm": 0.013671875,
      "learning_rate": 0.024350256410743055,
      "loss": 0.8145,
      "num_input_tokens_seen": 27839296,
      "step": 47965
    },
    {
      "epoch": 7.144772117962466,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.024348731815110432,
      "loss": 0.7786,
      "num_input_tokens_seen": 27842112,
      "step": 47970
    },
    {
      "epoch": 7.145516830503426,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.024347207061540498,
      "loss": 0.8198,
      "num_input_tokens_seen": 27845024,
      "step": 47975
    },
    {
      "epoch": 7.1462615430443845,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.024345682150058997,
      "loss": 0.8129,
      "num_input_tokens_seen": 27847776,
      "step": 47980
    },
    {
      "epoch": 7.147006255585344,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.024344157080691698,
      "loss": 0.7813,
      "num_input_tokens_seen": 27851008,
      "step": 47985
    },
    {
      "epoch": 7.147750968126303,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.02434263185346435,
      "loss": 0.7906,
      "num_input_tokens_seen": 27854240,
      "step": 47990
    },
    {
      "epoch": 7.148495680667263,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.024341106468402745,
      "loss": 0.8022,
      "num_input_tokens_seen": 27857312,
      "step": 47995
    },
    {
      "epoch": 7.149240393208221,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.024339580925532636,
      "loss": 0.788,
      "num_input_tokens_seen": 27860032,
      "step": 48000
    },
    {
      "epoch": 7.149985105749181,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.0243380552248798,
      "loss": 0.7933,
      "num_input_tokens_seen": 27863008,
      "step": 48005
    },
    {
      "epoch": 7.15072981829014,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02433652936647001,
      "loss": 0.8063,
      "num_input_tokens_seen": 27865920,
      "step": 48010
    },
    {
      "epoch": 7.151474530831099,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.024335003350329044,
      "loss": 0.7824,
      "num_input_tokens_seen": 27869120,
      "step": 48015
    },
    {
      "epoch": 7.152219243372058,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.024333477176482692,
      "loss": 0.7878,
      "num_input_tokens_seen": 27872128,
      "step": 48020
    },
    {
      "epoch": 7.152963955913018,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02433195084495672,
      "loss": 0.8079,
      "num_input_tokens_seen": 27874912,
      "step": 48025
    },
    {
      "epoch": 7.153708668453977,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02433042435577693,
      "loss": 0.7944,
      "num_input_tokens_seen": 27877344,
      "step": 48030
    },
    {
      "epoch": 7.154453380994936,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.024328897708969097,
      "loss": 0.8439,
      "num_input_tokens_seen": 27880128,
      "step": 48035
    },
    {
      "epoch": 7.155198093535895,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.024327370904559022,
      "loss": 0.8265,
      "num_input_tokens_seen": 27882816,
      "step": 48040
    },
    {
      "epoch": 7.155942806076855,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.0243258439425725,
      "loss": 0.8253,
      "num_input_tokens_seen": 27885504,
      "step": 48045
    },
    {
      "epoch": 7.156687518617813,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02432431682303532,
      "loss": 0.802,
      "num_input_tokens_seen": 27888480,
      "step": 48050
    },
    {
      "epoch": 7.157432231158773,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.024322789545973278,
      "loss": 0.8101,
      "num_input_tokens_seen": 27891328,
      "step": 48055
    },
    {
      "epoch": 7.158176943699732,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.024321262111412188,
      "loss": 0.7853,
      "num_input_tokens_seen": 27894112,
      "step": 48060
    },
    {
      "epoch": 7.158921656240691,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.024319734519377845,
      "loss": 0.7636,
      "num_input_tokens_seen": 27897216,
      "step": 48065
    },
    {
      "epoch": 7.15966636878165,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.024318206769896066,
      "loss": 0.7874,
      "num_input_tokens_seen": 27900064,
      "step": 48070
    },
    {
      "epoch": 7.16041108132261,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.024316678862992647,
      "loss": 0.8162,
      "num_input_tokens_seen": 27903072,
      "step": 48075
    },
    {
      "epoch": 7.161155793863569,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02431515079869341,
      "loss": 0.7988,
      "num_input_tokens_seen": 27905984,
      "step": 48080
    },
    {
      "epoch": 7.161900506404528,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.024313622577024168,
      "loss": 0.7971,
      "num_input_tokens_seen": 27908736,
      "step": 48085
    },
    {
      "epoch": 7.162645218945487,
      "grad_norm": 0.01092529296875,
      "learning_rate": 0.02431209419801074,
      "loss": 0.7973,
      "num_input_tokens_seen": 27911552,
      "step": 48090
    },
    {
      "epoch": 7.163389931486447,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.024310565661678944,
      "loss": 0.8247,
      "num_input_tokens_seen": 27914400,
      "step": 48095
    },
    {
      "epoch": 7.164134644027405,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.024309036968054604,
      "loss": 0.78,
      "num_input_tokens_seen": 27917408,
      "step": 48100
    },
    {
      "epoch": 7.164879356568365,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.024307508117163545,
      "loss": 0.788,
      "num_input_tokens_seen": 27920896,
      "step": 48105
    },
    {
      "epoch": 7.165624069109324,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.024305979109031595,
      "loss": 0.7773,
      "num_input_tokens_seen": 27923712,
      "step": 48110
    },
    {
      "epoch": 7.166368781650283,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.024304449943684588,
      "loss": 0.8037,
      "num_input_tokens_seen": 27926304,
      "step": 48115
    },
    {
      "epoch": 7.167113494191242,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.024302920621148354,
      "loss": 0.8104,
      "num_input_tokens_seen": 27929216,
      "step": 48120
    },
    {
      "epoch": 7.167858206732201,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02430139114144873,
      "loss": 0.8286,
      "num_input_tokens_seen": 27931872,
      "step": 48125
    },
    {
      "epoch": 7.168602919273161,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02429986150461156,
      "loss": 0.787,
      "num_input_tokens_seen": 27934784,
      "step": 48130
    },
    {
      "epoch": 7.169347631814119,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.024298331710662677,
      "loss": 0.7928,
      "num_input_tokens_seen": 27937536,
      "step": 48135
    },
    {
      "epoch": 7.170092344355079,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.024296801759627937,
      "loss": 0.8014,
      "num_input_tokens_seen": 27940352,
      "step": 48140
    },
    {
      "epoch": 7.170837056896038,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02429527165153318,
      "loss": 0.7842,
      "num_input_tokens_seen": 27943488,
      "step": 48145
    },
    {
      "epoch": 7.171581769436997,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02429374138640425,
      "loss": 0.7911,
      "num_input_tokens_seen": 27946560,
      "step": 48150
    },
    {
      "epoch": 7.172326481977956,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.024292210964267008,
      "loss": 0.7834,
      "num_input_tokens_seen": 27949472,
      "step": 48155
    },
    {
      "epoch": 7.173071194518916,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.024290680385147305,
      "loss": 0.8138,
      "num_input_tokens_seen": 27952288,
      "step": 48160
    },
    {
      "epoch": 7.173815907059875,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.024289149649070998,
      "loss": 0.7984,
      "num_input_tokens_seen": 27955200,
      "step": 48165
    },
    {
      "epoch": 7.174560619600834,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02428761875606395,
      "loss": 0.8177,
      "num_input_tokens_seen": 27957952,
      "step": 48170
    },
    {
      "epoch": 7.175305332141793,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.024286087706152024,
      "loss": 0.8021,
      "num_input_tokens_seen": 27960768,
      "step": 48175
    },
    {
      "epoch": 7.176050044682753,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.024284556499361083,
      "loss": 0.8014,
      "num_input_tokens_seen": 27963520,
      "step": 48180
    },
    {
      "epoch": 7.176794757223711,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.024283025135716996,
      "loss": 0.8047,
      "num_input_tokens_seen": 27966336,
      "step": 48185
    },
    {
      "epoch": 7.177539469764671,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.024281493615245637,
      "loss": 0.7837,
      "num_input_tokens_seen": 27969408,
      "step": 48190
    },
    {
      "epoch": 7.17828418230563,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.024279961937972872,
      "loss": 0.8455,
      "num_input_tokens_seen": 27972608,
      "step": 48195
    },
    {
      "epoch": 7.1790288948465895,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.024278430103924584,
      "loss": 0.8049,
      "num_input_tokens_seen": 27975328,
      "step": 48200
    },
    {
      "epoch": 7.179773607387548,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.024276898113126652,
      "loss": 0.7827,
      "num_input_tokens_seen": 27978240,
      "step": 48205
    },
    {
      "epoch": 7.180518319928508,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.024275365965604954,
      "loss": 0.7946,
      "num_input_tokens_seen": 27981088,
      "step": 48210
    },
    {
      "epoch": 7.181263032469467,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.024273833661385375,
      "loss": 0.7915,
      "num_input_tokens_seen": 27983872,
      "step": 48215
    },
    {
      "epoch": 7.182007745010426,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.0242723012004938,
      "loss": 0.7823,
      "num_input_tokens_seen": 27986752,
      "step": 48220
    },
    {
      "epoch": 7.182752457551385,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.024270768582956122,
      "loss": 0.7777,
      "num_input_tokens_seen": 27989504,
      "step": 48225
    },
    {
      "epoch": 7.183497170092345,
      "grad_norm": 0.0234375,
      "learning_rate": 0.024269235808798233,
      "loss": 0.8114,
      "num_input_tokens_seen": 27992256,
      "step": 48230
    },
    {
      "epoch": 7.184241882633303,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.024267702878046016,
      "loss": 0.8213,
      "num_input_tokens_seen": 27995136,
      "step": 48235
    },
    {
      "epoch": 7.184986595174263,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.024266169790725392,
      "loss": 0.8059,
      "num_input_tokens_seen": 27998048,
      "step": 48240
    },
    {
      "epoch": 7.185731307715222,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02426463654686224,
      "loss": 0.7853,
      "num_input_tokens_seen": 28000896,
      "step": 48245
    },
    {
      "epoch": 7.1864760202561815,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.024263103146482466,
      "loss": 0.7959,
      "num_input_tokens_seen": 28004000,
      "step": 48250
    },
    {
      "epoch": 7.18722073279714,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.02426156958961198,
      "loss": 0.8394,
      "num_input_tokens_seen": 28006752,
      "step": 48255
    },
    {
      "epoch": 7.1879654453381,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.024260035876276696,
      "loss": 0.8025,
      "num_input_tokens_seen": 28009760,
      "step": 48260
    },
    {
      "epoch": 7.188710157879059,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02425850200650251,
      "loss": 0.7864,
      "num_input_tokens_seen": 28012512,
      "step": 48265
    },
    {
      "epoch": 7.189454870420018,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.024256967980315347,
      "loss": 0.7764,
      "num_input_tokens_seen": 28015328,
      "step": 48270
    },
    {
      "epoch": 7.190199582960977,
      "grad_norm": 0.02734375,
      "learning_rate": 0.024255433797741116,
      "loss": 0.8011,
      "num_input_tokens_seen": 28018048,
      "step": 48275
    },
    {
      "epoch": 7.190944295501936,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.024253899458805742,
      "loss": 0.7845,
      "num_input_tokens_seen": 28020832,
      "step": 48280
    },
    {
      "epoch": 7.1916890080428955,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.024252364963535136,
      "loss": 0.7827,
      "num_input_tokens_seen": 28023840,
      "step": 48285
    },
    {
      "epoch": 7.192433720583854,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02425083031195523,
      "loss": 0.8192,
      "num_input_tokens_seen": 28026624,
      "step": 48290
    },
    {
      "epoch": 7.193178433124814,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02424929550409195,
      "loss": 0.7687,
      "num_input_tokens_seen": 28029376,
      "step": 48295
    },
    {
      "epoch": 7.193923145665773,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02424776053997122,
      "loss": 0.7854,
      "num_input_tokens_seen": 28032704,
      "step": 48300
    },
    {
      "epoch": 7.194667858206732,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.024246225419618978,
      "loss": 0.7663,
      "num_input_tokens_seen": 28035392,
      "step": 48305
    },
    {
      "epoch": 7.195412570747691,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.024244690143061152,
      "loss": 0.8231,
      "num_input_tokens_seen": 28038080,
      "step": 48310
    },
    {
      "epoch": 7.196157283288651,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02424315471032368,
      "loss": 0.7907,
      "num_input_tokens_seen": 28041056,
      "step": 48315
    },
    {
      "epoch": 7.196901995829609,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02424161912143251,
      "loss": 0.8158,
      "num_input_tokens_seen": 28044000,
      "step": 48320
    },
    {
      "epoch": 7.197646708370569,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.024240083376413573,
      "loss": 0.8297,
      "num_input_tokens_seen": 28046720,
      "step": 48325
    },
    {
      "epoch": 7.198391420911528,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.024238547475292818,
      "loss": 0.8205,
      "num_input_tokens_seen": 28049600,
      "step": 48330
    },
    {
      "epoch": 7.1991361334524875,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.024237011418096196,
      "loss": 0.819,
      "num_input_tokens_seen": 28052544,
      "step": 48335
    },
    {
      "epoch": 7.199880845993446,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02423547520484965,
      "loss": 0.8289,
      "num_input_tokens_seen": 28055680,
      "step": 48340
    },
    {
      "epoch": 7.200625558534406,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.024233938835579144,
      "loss": 0.7781,
      "num_input_tokens_seen": 28058592,
      "step": 48345
    },
    {
      "epoch": 7.201370271075365,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.02423240231031062,
      "loss": 0.7972,
      "num_input_tokens_seen": 28061504,
      "step": 48350
    },
    {
      "epoch": 7.202114983616324,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.024230865629070047,
      "loss": 0.8221,
      "num_input_tokens_seen": 28064320,
      "step": 48355
    },
    {
      "epoch": 7.202859696157283,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.024229328791883376,
      "loss": 0.8061,
      "num_input_tokens_seen": 28067232,
      "step": 48360
    },
    {
      "epoch": 7.203604408698243,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.024227791798776578,
      "loss": 0.8202,
      "num_input_tokens_seen": 28070400,
      "step": 48365
    },
    {
      "epoch": 7.2043491212392015,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.024226254649775614,
      "loss": 0.8049,
      "num_input_tokens_seen": 28073152,
      "step": 48370
    },
    {
      "epoch": 7.205093833780161,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.024224717344906457,
      "loss": 0.8119,
      "num_input_tokens_seen": 28075808,
      "step": 48375
    },
    {
      "epoch": 7.20583854632112,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.024223179884195076,
      "loss": 0.798,
      "num_input_tokens_seen": 28079008,
      "step": 48380
    },
    {
      "epoch": 7.2065832588620795,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.024221642267667445,
      "loss": 0.7915,
      "num_input_tokens_seen": 28081664,
      "step": 48385
    },
    {
      "epoch": 7.207327971403038,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.024220104495349542,
      "loss": 0.8018,
      "num_input_tokens_seen": 28084256,
      "step": 48390
    },
    {
      "epoch": 7.208072683943998,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02421856656726734,
      "loss": 0.7945,
      "num_input_tokens_seen": 28087136,
      "step": 48395
    },
    {
      "epoch": 7.208817396484957,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.024217028483446833,
      "loss": 0.8266,
      "num_input_tokens_seen": 28090112,
      "step": 48400
    },
    {
      "epoch": 7.209562109025916,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02421549024391399,
      "loss": 0.8306,
      "num_input_tokens_seen": 28092736,
      "step": 48405
    },
    {
      "epoch": 7.210306821566875,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.024213951848694806,
      "loss": 0.8212,
      "num_input_tokens_seen": 28095616,
      "step": 48410
    },
    {
      "epoch": 7.211051534107835,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02421241329781527,
      "loss": 0.8037,
      "num_input_tokens_seen": 28098400,
      "step": 48415
    },
    {
      "epoch": 7.2117962466487935,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02421087459130138,
      "loss": 0.8105,
      "num_input_tokens_seen": 28101280,
      "step": 48420
    },
    {
      "epoch": 7.212540959189753,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.024209335729179122,
      "loss": 0.8044,
      "num_input_tokens_seen": 28104416,
      "step": 48425
    },
    {
      "epoch": 7.213285671730712,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.024207796711474498,
      "loss": 0.8155,
      "num_input_tokens_seen": 28107040,
      "step": 48430
    },
    {
      "epoch": 7.2140303842716715,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.024206257538213503,
      "loss": 0.7955,
      "num_input_tokens_seen": 28109760,
      "step": 48435
    },
    {
      "epoch": 7.21477509681263,
      "grad_norm": 0.0234375,
      "learning_rate": 0.024204718209422146,
      "loss": 0.8017,
      "num_input_tokens_seen": 28112608,
      "step": 48440
    },
    {
      "epoch": 7.21551980935359,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02420317872512643,
      "loss": 0.7947,
      "num_input_tokens_seen": 28115616,
      "step": 48445
    },
    {
      "epoch": 7.216264521894549,
      "grad_norm": 0.025390625,
      "learning_rate": 0.024201639085352365,
      "loss": 0.8145,
      "num_input_tokens_seen": 28118464,
      "step": 48450
    },
    {
      "epoch": 7.217009234435508,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.024200099290125957,
      "loss": 0.8182,
      "num_input_tokens_seen": 28121312,
      "step": 48455
    },
    {
      "epoch": 7.217753946976467,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.024198559339473225,
      "loss": 0.8174,
      "num_input_tokens_seen": 28124256,
      "step": 48460
    },
    {
      "epoch": 7.218498659517426,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02419701923342018,
      "loss": 0.8235,
      "num_input_tokens_seen": 28127040,
      "step": 48465
    },
    {
      "epoch": 7.2192433720583855,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02419547897199284,
      "loss": 0.8118,
      "num_input_tokens_seen": 28129920,
      "step": 48470
    },
    {
      "epoch": 7.219988084599344,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02419393855521723,
      "loss": 0.7999,
      "num_input_tokens_seen": 28132672,
      "step": 48475
    },
    {
      "epoch": 7.220732797140304,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.024192397983119375,
      "loss": 0.7975,
      "num_input_tokens_seen": 28135392,
      "step": 48480
    },
    {
      "epoch": 7.221477509681263,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.024190857255725298,
      "loss": 0.7946,
      "num_input_tokens_seen": 28138144,
      "step": 48485
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02418931637306103,
      "loss": 0.7971,
      "num_input_tokens_seen": 28140992,
      "step": 48490
    },
    {
      "epoch": 7.222966934763181,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.0241877753351526,
      "loss": 0.8048,
      "num_input_tokens_seen": 28143936,
      "step": 48495
    },
    {
      "epoch": 7.223711647304141,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02418623414202604,
      "loss": 0.8105,
      "num_input_tokens_seen": 28146912,
      "step": 48500
    },
    {
      "epoch": 7.2244563598450995,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.024184692793707396,
      "loss": 0.7864,
      "num_input_tokens_seen": 28149792,
      "step": 48505
    },
    {
      "epoch": 7.225201072386059,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.024183151290222703,
      "loss": 0.7977,
      "num_input_tokens_seen": 28152864,
      "step": 48510
    },
    {
      "epoch": 7.225945784927018,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.024181609631597996,
      "loss": 0.7838,
      "num_input_tokens_seen": 28155744,
      "step": 48515
    },
    {
      "epoch": 7.2266904974679775,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02418006781785933,
      "loss": 0.8019,
      "num_input_tokens_seen": 28158816,
      "step": 48520
    },
    {
      "epoch": 7.227435210008936,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02417852584903275,
      "loss": 0.8058,
      "num_input_tokens_seen": 28161696,
      "step": 48525
    },
    {
      "epoch": 7.228179922549896,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.0241769837251443,
      "loss": 0.8109,
      "num_input_tokens_seen": 28164448,
      "step": 48530
    },
    {
      "epoch": 7.228924635090855,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02417544144622004,
      "loss": 0.7948,
      "num_input_tokens_seen": 28167264,
      "step": 48535
    },
    {
      "epoch": 7.229669347631814,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.024173899012286025,
      "loss": 0.8304,
      "num_input_tokens_seen": 28170176,
      "step": 48540
    },
    {
      "epoch": 7.230414060172773,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.024172356423368308,
      "loss": 0.8222,
      "num_input_tokens_seen": 28173056,
      "step": 48545
    },
    {
      "epoch": 7.231158772713733,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.024170813679492947,
      "loss": 0.8009,
      "num_input_tokens_seen": 28176224,
      "step": 48550
    },
    {
      "epoch": 7.2319034852546915,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.024169270780686015,
      "loss": 0.7992,
      "num_input_tokens_seen": 28179104,
      "step": 48555
    },
    {
      "epoch": 7.232648197795651,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02416772772697357,
      "loss": 0.7933,
      "num_input_tokens_seen": 28182016,
      "step": 48560
    },
    {
      "epoch": 7.23339291033661,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02416618451838168,
      "loss": 0.791,
      "num_input_tokens_seen": 28184832,
      "step": 48565
    },
    {
      "epoch": 7.23413762287757,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.024164641154936425,
      "loss": 0.807,
      "num_input_tokens_seen": 28187616,
      "step": 48570
    },
    {
      "epoch": 7.234882335418528,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02416309763666387,
      "loss": 0.8177,
      "num_input_tokens_seen": 28190464,
      "step": 48575
    },
    {
      "epoch": 7.235627047959488,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.02416155396359009,
      "loss": 0.7994,
      "num_input_tokens_seen": 28193248,
      "step": 48580
    },
    {
      "epoch": 7.236371760500447,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02416001013574117,
      "loss": 0.8038,
      "num_input_tokens_seen": 28196128,
      "step": 48585
    },
    {
      "epoch": 7.237116473041406,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02415846615314319,
      "loss": 0.7874,
      "num_input_tokens_seen": 28199040,
      "step": 48590
    },
    {
      "epoch": 7.237861185582365,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02415692201582223,
      "loss": 0.7998,
      "num_input_tokens_seen": 28202368,
      "step": 48595
    },
    {
      "epoch": 7.238605898123325,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02415537772380438,
      "loss": 0.7911,
      "num_input_tokens_seen": 28205312,
      "step": 48600
    },
    {
      "epoch": 7.2393506106642835,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.024153833277115733,
      "loss": 0.8056,
      "num_input_tokens_seen": 28208320,
      "step": 48605
    },
    {
      "epoch": 7.240095323205243,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02415228867578237,
      "loss": 0.7978,
      "num_input_tokens_seen": 28211168,
      "step": 48610
    },
    {
      "epoch": 7.240840035746202,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02415074391983039,
      "loss": 0.803,
      "num_input_tokens_seen": 28214080,
      "step": 48615
    },
    {
      "epoch": 7.241584748287162,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.024149199009285898,
      "loss": 0.8114,
      "num_input_tokens_seen": 28216832,
      "step": 48620
    },
    {
      "epoch": 7.24232946082812,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02414765394417499,
      "loss": 0.7871,
      "num_input_tokens_seen": 28219648,
      "step": 48625
    },
    {
      "epoch": 7.243074173369079,
      "grad_norm": 0.021484375,
      "learning_rate": 0.024146108724523756,
      "loss": 0.7966,
      "num_input_tokens_seen": 28222496,
      "step": 48630
    },
    {
      "epoch": 7.243818885910039,
      "grad_norm": 0.05078125,
      "learning_rate": 0.024144563350358316,
      "loss": 0.844,
      "num_input_tokens_seen": 28225984,
      "step": 48635
    },
    {
      "epoch": 7.2445635984509975,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.024143017821704773,
      "loss": 0.7991,
      "num_input_tokens_seen": 28228992,
      "step": 48640
    },
    {
      "epoch": 7.245308310991957,
      "grad_norm": 0.03125,
      "learning_rate": 0.024141472138589238,
      "loss": 0.7967,
      "num_input_tokens_seen": 28232288,
      "step": 48645
    },
    {
      "epoch": 7.246053023532916,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02413992630103782,
      "loss": 0.8152,
      "num_input_tokens_seen": 28235168,
      "step": 48650
    },
    {
      "epoch": 7.246797736073876,
      "grad_norm": 0.01092529296875,
      "learning_rate": 0.024138380309076633,
      "loss": 0.8004,
      "num_input_tokens_seen": 28238368,
      "step": 48655
    },
    {
      "epoch": 7.247542448614834,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.0241368341627318,
      "loss": 0.8029,
      "num_input_tokens_seen": 28241376,
      "step": 48660
    },
    {
      "epoch": 7.248287161155794,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.024135287862029446,
      "loss": 0.7949,
      "num_input_tokens_seen": 28244384,
      "step": 48665
    },
    {
      "epoch": 7.249031873696753,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.024133741406995684,
      "loss": 0.8155,
      "num_input_tokens_seen": 28247168,
      "step": 48670
    },
    {
      "epoch": 7.249776586237712,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.024132194797656645,
      "loss": 0.7929,
      "num_input_tokens_seen": 28250048,
      "step": 48675
    },
    {
      "epoch": 7.250521298778671,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.024130648034038454,
      "loss": 0.793,
      "num_input_tokens_seen": 28253120,
      "step": 48680
    },
    {
      "epoch": 7.251266011319631,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.024129101116167246,
      "loss": 0.8232,
      "num_input_tokens_seen": 28256064,
      "step": 48685
    },
    {
      "epoch": 7.2520107238605895,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.024127554044069154,
      "loss": 0.8074,
      "num_input_tokens_seen": 28258912,
      "step": 48690
    },
    {
      "epoch": 7.252755436401549,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02412600681777031,
      "loss": 0.8031,
      "num_input_tokens_seen": 28261792,
      "step": 48695
    },
    {
      "epoch": 7.253500148942508,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02412445943729686,
      "loss": 0.7954,
      "num_input_tokens_seen": 28264768,
      "step": 48700
    },
    {
      "epoch": 7.254244861483468,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02412291190267494,
      "loss": 0.788,
      "num_input_tokens_seen": 28268160,
      "step": 48705
    },
    {
      "epoch": 7.254989574024426,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.024121364213930695,
      "loss": 0.7911,
      "num_input_tokens_seen": 28270784,
      "step": 48710
    },
    {
      "epoch": 7.255734286565386,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.024119816371090275,
      "loss": 0.8473,
      "num_input_tokens_seen": 28273536,
      "step": 48715
    },
    {
      "epoch": 7.256478999106345,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.024118268374179824,
      "loss": 0.8196,
      "num_input_tokens_seen": 28276512,
      "step": 48720
    },
    {
      "epoch": 7.257223711647304,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.024116720223225495,
      "loss": 0.8212,
      "num_input_tokens_seen": 28279552,
      "step": 48725
    },
    {
      "epoch": 7.257968424188263,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.024115171918253443,
      "loss": 0.7929,
      "num_input_tokens_seen": 28282240,
      "step": 48730
    },
    {
      "epoch": 7.258713136729223,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02411362345928983,
      "loss": 0.7884,
      "num_input_tokens_seen": 28284896,
      "step": 48735
    },
    {
      "epoch": 7.259457849270182,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.024112074846360812,
      "loss": 0.7848,
      "num_input_tokens_seen": 28287808,
      "step": 48740
    },
    {
      "epoch": 7.260202561811141,
      "grad_norm": 0.01153564453125,
      "learning_rate": 0.024110526079492545,
      "loss": 0.7903,
      "num_input_tokens_seen": 28290528,
      "step": 48745
    },
    {
      "epoch": 7.2609472743521,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.0241089771587112,
      "loss": 0.7919,
      "num_input_tokens_seen": 28293664,
      "step": 48750
    },
    {
      "epoch": 7.26169198689306,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02410742808404295,
      "loss": 0.7718,
      "num_input_tokens_seen": 28296640,
      "step": 48755
    },
    {
      "epoch": 7.262436699434018,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.024105878855513952,
      "loss": 0.7918,
      "num_input_tokens_seen": 28299584,
      "step": 48760
    },
    {
      "epoch": 7.263181411974978,
      "grad_norm": 0.01055908203125,
      "learning_rate": 0.024104329473150388,
      "loss": 0.8108,
      "num_input_tokens_seen": 28302400,
      "step": 48765
    },
    {
      "epoch": 7.263926124515937,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02410277993697843,
      "loss": 0.7773,
      "num_input_tokens_seen": 28305504,
      "step": 48770
    },
    {
      "epoch": 7.264670837056896,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.024101230247024256,
      "loss": 0.7928,
      "num_input_tokens_seen": 28308480,
      "step": 48775
    },
    {
      "epoch": 7.265415549597855,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02409968040331405,
      "loss": 0.7929,
      "num_input_tokens_seen": 28311584,
      "step": 48780
    },
    {
      "epoch": 7.266160262138815,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.024098130405873993,
      "loss": 0.8005,
      "num_input_tokens_seen": 28315168,
      "step": 48785
    },
    {
      "epoch": 7.266904974679774,
      "grad_norm": 0.01953125,
      "learning_rate": 0.024096580254730268,
      "loss": 0.789,
      "num_input_tokens_seen": 28318048,
      "step": 48790
    },
    {
      "epoch": 7.267649687220732,
      "grad_norm": 0.0128173828125,
      "learning_rate": 0.02409502994990906,
      "loss": 0.8124,
      "num_input_tokens_seen": 28320800,
      "step": 48795
    },
    {
      "epoch": 7.268394399761692,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.024093479491436573,
      "loss": 0.7739,
      "num_input_tokens_seen": 28323584,
      "step": 48800
    },
    {
      "epoch": 7.269139112302652,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02409192887933899,
      "loss": 0.8233,
      "num_input_tokens_seen": 28326368,
      "step": 48805
    },
    {
      "epoch": 7.26988382484361,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.024090378113642508,
      "loss": 0.8069,
      "num_input_tokens_seen": 28328960,
      "step": 48810
    },
    {
      "epoch": 7.270628537384569,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.024088827194373323,
      "loss": 0.789,
      "num_input_tokens_seen": 28331936,
      "step": 48815
    },
    {
      "epoch": 7.271373249925529,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.024087276121557644,
      "loss": 0.7986,
      "num_input_tokens_seen": 28334976,
      "step": 48820
    },
    {
      "epoch": 7.272117962466488,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02408572489522167,
      "loss": 0.8023,
      "num_input_tokens_seen": 28337888,
      "step": 48825
    },
    {
      "epoch": 7.272862675007447,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.024084173515391608,
      "loss": 0.7983,
      "num_input_tokens_seen": 28340896,
      "step": 48830
    },
    {
      "epoch": 7.273607387548406,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02408262198209367,
      "loss": 0.7889,
      "num_input_tokens_seen": 28343616,
      "step": 48835
    },
    {
      "epoch": 7.274352100089366,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.024081070295354064,
      "loss": 0.8436,
      "num_input_tokens_seen": 28346368,
      "step": 48840
    },
    {
      "epoch": 7.275096812630324,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.024079518455199,
      "loss": 0.7781,
      "num_input_tokens_seen": 28349344,
      "step": 48845
    },
    {
      "epoch": 7.275841525171284,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.024077966461654703,
      "loss": 0.7979,
      "num_input_tokens_seen": 28352768,
      "step": 48850
    },
    {
      "epoch": 7.276586237712243,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02407641431474739,
      "loss": 0.795,
      "num_input_tokens_seen": 28355360,
      "step": 48855
    },
    {
      "epoch": 7.277330950253202,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02407486201450328,
      "loss": 0.792,
      "num_input_tokens_seen": 28358400,
      "step": 48860
    },
    {
      "epoch": 7.278075662794161,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.0240733095609486,
      "loss": 0.8319,
      "num_input_tokens_seen": 28361344,
      "step": 48865
    },
    {
      "epoch": 7.278820375335121,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02407175695410958,
      "loss": 0.8017,
      "num_input_tokens_seen": 28364512,
      "step": 48870
    },
    {
      "epoch": 7.27956508787608,
      "grad_norm": 0.01507568359375,
      "learning_rate": 0.024070204194012446,
      "loss": 0.7858,
      "num_input_tokens_seen": 28367264,
      "step": 48875
    },
    {
      "epoch": 7.280309800417039,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.024068651280683428,
      "loss": 0.8181,
      "num_input_tokens_seen": 28370240,
      "step": 48880
    },
    {
      "epoch": 7.281054512957998,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02406709821414876,
      "loss": 0.7912,
      "num_input_tokens_seen": 28373216,
      "step": 48885
    },
    {
      "epoch": 7.281799225498958,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02406554499443469,
      "loss": 0.8048,
      "num_input_tokens_seen": 28375872,
      "step": 48890
    },
    {
      "epoch": 7.282543938039916,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.024063991621567447,
      "loss": 0.7816,
      "num_input_tokens_seen": 28378912,
      "step": 48895
    },
    {
      "epoch": 7.283288650580876,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.02406243809557328,
      "loss": 0.7933,
      "num_input_tokens_seen": 28381728,
      "step": 48900
    },
    {
      "epoch": 7.284033363121835,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02406088441647843,
      "loss": 0.808,
      "num_input_tokens_seen": 28384832,
      "step": 48905
    },
    {
      "epoch": 7.2847780756627944,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.024059330584309147,
      "loss": 0.8318,
      "num_input_tokens_seen": 28387520,
      "step": 48910
    },
    {
      "epoch": 7.285522788203753,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.024057776599091682,
      "loss": 0.791,
      "num_input_tokens_seen": 28390816,
      "step": 48915
    },
    {
      "epoch": 7.286267500744713,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02405622246085229,
      "loss": 0.7923,
      "num_input_tokens_seen": 28393984,
      "step": 48920
    },
    {
      "epoch": 7.287012213285672,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.02405466816961722,
      "loss": 0.8113,
      "num_input_tokens_seen": 28396736,
      "step": 48925
    },
    {
      "epoch": 7.287756925826631,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.024053113725412734,
      "loss": 0.7927,
      "num_input_tokens_seen": 28399680,
      "step": 48930
    },
    {
      "epoch": 7.28850163836759,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.024051559128265097,
      "loss": 0.8276,
      "num_input_tokens_seen": 28402336,
      "step": 48935
    },
    {
      "epoch": 7.28924635090855,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.024050004378200564,
      "loss": 0.8083,
      "num_input_tokens_seen": 28404928,
      "step": 48940
    },
    {
      "epoch": 7.289991063449508,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.024048449475245405,
      "loss": 0.8049,
      "num_input_tokens_seen": 28407936,
      "step": 48945
    },
    {
      "epoch": 7.290735775990468,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.024046894419425893,
      "loss": 0.788,
      "num_input_tokens_seen": 28410816,
      "step": 48950
    },
    {
      "epoch": 7.291480488531427,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.02404533921076829,
      "loss": 0.8277,
      "num_input_tokens_seen": 28413824,
      "step": 48955
    },
    {
      "epoch": 7.292225201072386,
      "grad_norm": 0.01953125,
      "learning_rate": 0.024043783849298877,
      "loss": 0.7953,
      "num_input_tokens_seen": 28416928,
      "step": 48960
    },
    {
      "epoch": 7.292969913613345,
      "grad_norm": 0.0234375,
      "learning_rate": 0.024042228335043922,
      "loss": 0.8081,
      "num_input_tokens_seen": 28420000,
      "step": 48965
    },
    {
      "epoch": 7.293714626154305,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.024040672668029717,
      "loss": 0.8024,
      "num_input_tokens_seen": 28422976,
      "step": 48970
    },
    {
      "epoch": 7.294459338695264,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.024039116848282534,
      "loss": 0.7952,
      "num_input_tokens_seen": 28426208,
      "step": 48975
    },
    {
      "epoch": 7.295204051236222,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.024037560875828656,
      "loss": 0.7926,
      "num_input_tokens_seen": 28428992,
      "step": 48980
    },
    {
      "epoch": 7.295948763777182,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.024036004750694374,
      "loss": 0.7877,
      "num_input_tokens_seen": 28431840,
      "step": 48985
    },
    {
      "epoch": 7.296693476318141,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.024034448472905975,
      "loss": 0.7881,
      "num_input_tokens_seen": 28435008,
      "step": 48990
    },
    {
      "epoch": 7.2974381888591004,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02403289204248975,
      "loss": 0.7915,
      "num_input_tokens_seen": 28438176,
      "step": 48995
    },
    {
      "epoch": 7.298182901400059,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.024031335459471996,
      "loss": 0.8158,
      "num_input_tokens_seen": 28441472,
      "step": 49000
    },
    {
      "epoch": 7.298927613941019,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02402977872387901,
      "loss": 0.7826,
      "num_input_tokens_seen": 28444288,
      "step": 49005
    },
    {
      "epoch": 7.299672326481978,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.024028221835737086,
      "loss": 0.8027,
      "num_input_tokens_seen": 28446944,
      "step": 49010
    },
    {
      "epoch": 7.300417039022937,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.024026664795072532,
      "loss": 0.7947,
      "num_input_tokens_seen": 28449888,
      "step": 49015
    },
    {
      "epoch": 7.301161751563896,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.024025107601911653,
      "loss": 0.8005,
      "num_input_tokens_seen": 28452768,
      "step": 49020
    },
    {
      "epoch": 7.301906464104856,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.02402355025628075,
      "loss": 0.7862,
      "num_input_tokens_seen": 28455872,
      "step": 49025
    },
    {
      "epoch": 7.302651176645814,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02402199275820614,
      "loss": 0.8009,
      "num_input_tokens_seen": 28458880,
      "step": 49030
    },
    {
      "epoch": 7.303395889186774,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.02402043510771413,
      "loss": 0.8101,
      "num_input_tokens_seen": 28461696,
      "step": 49035
    },
    {
      "epoch": 7.304140601727733,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.024018877304831036,
      "loss": 0.8164,
      "num_input_tokens_seen": 28464768,
      "step": 49040
    },
    {
      "epoch": 7.3048853142686925,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.024017319349583177,
      "loss": 0.7996,
      "num_input_tokens_seen": 28467552,
      "step": 49045
    },
    {
      "epoch": 7.305630026809651,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.024015761241996874,
      "loss": 0.8005,
      "num_input_tokens_seen": 28470592,
      "step": 49050
    },
    {
      "epoch": 7.306374739350611,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02401420298209845,
      "loss": 0.7676,
      "num_input_tokens_seen": 28473376,
      "step": 49055
    },
    {
      "epoch": 7.30711945189157,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.024012644569914223,
      "loss": 0.8038,
      "num_input_tokens_seen": 28476160,
      "step": 49060
    },
    {
      "epoch": 7.307864164432529,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.024011086005470524,
      "loss": 0.7956,
      "num_input_tokens_seen": 28478816,
      "step": 49065
    },
    {
      "epoch": 7.308608876973488,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.024009527288793693,
      "loss": 0.8065,
      "num_input_tokens_seen": 28481920,
      "step": 49070
    },
    {
      "epoch": 7.309353589514448,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02400796841991005,
      "loss": 0.7997,
      "num_input_tokens_seen": 28484704,
      "step": 49075
    },
    {
      "epoch": 7.3100983020554064,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02400640939884594,
      "loss": 0.8052,
      "num_input_tokens_seen": 28487616,
      "step": 49080
    },
    {
      "epoch": 7.310843014596366,
      "grad_norm": 0.013671875,
      "learning_rate": 0.024004850225627695,
      "loss": 0.8219,
      "num_input_tokens_seen": 28490656,
      "step": 49085
    },
    {
      "epoch": 7.311587727137325,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.02400329090028166,
      "loss": 0.782,
      "num_input_tokens_seen": 28493888,
      "step": 49090
    },
    {
      "epoch": 7.3123324396782845,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.024001731422834165,
      "loss": 0.7885,
      "num_input_tokens_seen": 28496864,
      "step": 49095
    },
    {
      "epoch": 7.313077152219243,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.024000171793311577,
      "loss": 0.8253,
      "num_input_tokens_seen": 28499648,
      "step": 49100
    },
    {
      "epoch": 7.313821864760203,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02399861201174023,
      "loss": 0.7977,
      "num_input_tokens_seen": 28502592,
      "step": 49105
    },
    {
      "epoch": 7.314566577301162,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.023997052078146484,
      "loss": 0.8129,
      "num_input_tokens_seen": 28505344,
      "step": 49110
    },
    {
      "epoch": 7.315311289842121,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.023995491992556683,
      "loss": 0.7907,
      "num_input_tokens_seen": 28508032,
      "step": 49115
    },
    {
      "epoch": 7.31605600238308,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.023993931754997186,
      "loss": 0.8018,
      "num_input_tokens_seen": 28511136,
      "step": 49120
    },
    {
      "epoch": 7.31680071492404,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.023992371365494355,
      "loss": 0.8233,
      "num_input_tokens_seen": 28513888,
      "step": 49125
    },
    {
      "epoch": 7.3175454274649985,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.023990810824074546,
      "loss": 0.7832,
      "num_input_tokens_seen": 28517024,
      "step": 49130
    },
    {
      "epoch": 7.318290140005958,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02398925013076413,
      "loss": 0.8214,
      "num_input_tokens_seen": 28519680,
      "step": 49135
    },
    {
      "epoch": 7.319034852546917,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02398768928558947,
      "loss": 0.777,
      "num_input_tokens_seen": 28522752,
      "step": 49140
    },
    {
      "epoch": 7.319779565087876,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02398612828857693,
      "loss": 0.8062,
      "num_input_tokens_seen": 28525824,
      "step": 49145
    },
    {
      "epoch": 7.320524277628835,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.023984567139752888,
      "loss": 0.78,
      "num_input_tokens_seen": 28528864,
      "step": 49150
    },
    {
      "epoch": 7.321268990169794,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.023983005839143718,
      "loss": 0.8115,
      "num_input_tokens_seen": 28531680,
      "step": 49155
    },
    {
      "epoch": 7.322013702710754,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.023981444386775793,
      "loss": 0.7872,
      "num_input_tokens_seen": 28534432,
      "step": 49160
    },
    {
      "epoch": 7.3227584152517124,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.023979882782675488,
      "loss": 0.8102,
      "num_input_tokens_seen": 28537440,
      "step": 49165
    },
    {
      "epoch": 7.323503127792672,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.023978321026869198,
      "loss": 0.8012,
      "num_input_tokens_seen": 28540608,
      "step": 49170
    },
    {
      "epoch": 7.324247840333631,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.023976759119383295,
      "loss": 0.7925,
      "num_input_tokens_seen": 28543520,
      "step": 49175
    },
    {
      "epoch": 7.3249925528745905,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02397519706024417,
      "loss": 0.8288,
      "num_input_tokens_seen": 28546400,
      "step": 49180
    },
    {
      "epoch": 7.325737265415549,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.023973634849478218,
      "loss": 0.8127,
      "num_input_tokens_seen": 28549408,
      "step": 49185
    },
    {
      "epoch": 7.326481977956509,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.023972072487111824,
      "loss": 0.8017,
      "num_input_tokens_seen": 28552416,
      "step": 49190
    },
    {
      "epoch": 7.327226690497468,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.023970509973171378,
      "loss": 0.812,
      "num_input_tokens_seen": 28555392,
      "step": 49195
    },
    {
      "epoch": 7.327971403038427,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.023968947307683286,
      "loss": 0.8013,
      "num_input_tokens_seen": 28558336,
      "step": 49200
    },
    {
      "epoch": 7.328716115579386,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02396738449067395,
      "loss": 0.7954,
      "num_input_tokens_seen": 28561056,
      "step": 49205
    },
    {
      "epoch": 7.329460828120346,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.023965821522169764,
      "loss": 0.8092,
      "num_input_tokens_seen": 28563840,
      "step": 49210
    },
    {
      "epoch": 7.3302055406613045,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.023964258402197132,
      "loss": 0.7977,
      "num_input_tokens_seen": 28566624,
      "step": 49215
    },
    {
      "epoch": 7.330950253202264,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.023962695130782467,
      "loss": 0.7891,
      "num_input_tokens_seen": 28569344,
      "step": 49220
    },
    {
      "epoch": 7.331694965743223,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.023961131707952176,
      "loss": 0.833,
      "num_input_tokens_seen": 28572096,
      "step": 49225
    },
    {
      "epoch": 7.3324396782841825,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.023959568133732678,
      "loss": 0.7908,
      "num_input_tokens_seen": 28574944,
      "step": 49230
    },
    {
      "epoch": 7.333184390825141,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.023958004408150376,
      "loss": 0.7992,
      "num_input_tokens_seen": 28577664,
      "step": 49235
    },
    {
      "epoch": 7.333929103366101,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.0239564405312317,
      "loss": 0.7955,
      "num_input_tokens_seen": 28580448,
      "step": 49240
    },
    {
      "epoch": 7.33467381590706,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02395487650300306,
      "loss": 0.81,
      "num_input_tokens_seen": 28583328,
      "step": 49245
    },
    {
      "epoch": 7.335418528448019,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02395331232349088,
      "loss": 0.7882,
      "num_input_tokens_seen": 28586080,
      "step": 49250
    },
    {
      "epoch": 7.336163240988978,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.023951747992721595,
      "loss": 0.8103,
      "num_input_tokens_seen": 28589280,
      "step": 49255
    },
    {
      "epoch": 7.336907953529938,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.023950183510721625,
      "loss": 0.8046,
      "num_input_tokens_seen": 28592224,
      "step": 49260
    },
    {
      "epoch": 7.3376526660708965,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.023948618877517397,
      "loss": 0.8033,
      "num_input_tokens_seen": 28595168,
      "step": 49265
    },
    {
      "epoch": 7.338397378611856,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.02394705409313535,
      "loss": 0.8097,
      "num_input_tokens_seen": 28598112,
      "step": 49270
    },
    {
      "epoch": 7.339142091152815,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.023945489157601918,
      "loss": 0.8229,
      "num_input_tokens_seen": 28601152,
      "step": 49275
    },
    {
      "epoch": 7.3398868036937746,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.023943924070943538,
      "loss": 0.7804,
      "num_input_tokens_seen": 28604192,
      "step": 49280
    },
    {
      "epoch": 7.340631516234733,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.023942358833186652,
      "loss": 0.8068,
      "num_input_tokens_seen": 28606976,
      "step": 49285
    },
    {
      "epoch": 7.341376228775693,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.0239407934443577,
      "loss": 0.8313,
      "num_input_tokens_seen": 28609632,
      "step": 49290
    },
    {
      "epoch": 7.342120941316652,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.023939227904483128,
      "loss": 0.7952,
      "num_input_tokens_seen": 28612256,
      "step": 49295
    },
    {
      "epoch": 7.342865653857611,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02393766221358939,
      "loss": 0.8048,
      "num_input_tokens_seen": 28615072,
      "step": 49300
    },
    {
      "epoch": 7.34361036639857,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.023936096371702933,
      "loss": 0.8179,
      "num_input_tokens_seen": 28617856,
      "step": 49305
    },
    {
      "epoch": 7.344355078939529,
      "grad_norm": 0.015625,
      "learning_rate": 0.02393453037885021,
      "loss": 0.8007,
      "num_input_tokens_seen": 28620736,
      "step": 49310
    },
    {
      "epoch": 7.3450997914804885,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.023932964235057672,
      "loss": 0.7869,
      "num_input_tokens_seen": 28623520,
      "step": 49315
    },
    {
      "epoch": 7.345844504021448,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02393139794035179,
      "loss": 0.7811,
      "num_input_tokens_seen": 28628000,
      "step": 49320
    },
    {
      "epoch": 7.346589216562407,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02392983149475901,
      "loss": 0.8042,
      "num_input_tokens_seen": 28631072,
      "step": 49325
    },
    {
      "epoch": 7.347333929103366,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.023928264898305806,
      "loss": 0.7981,
      "num_input_tokens_seen": 28633984,
      "step": 49330
    },
    {
      "epoch": 7.348078641644325,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02392669815101864,
      "loss": 0.8146,
      "num_input_tokens_seen": 28636768,
      "step": 49335
    },
    {
      "epoch": 7.348823354185284,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.023925131252923986,
      "loss": 0.791,
      "num_input_tokens_seen": 28639488,
      "step": 49340
    },
    {
      "epoch": 7.349568066726244,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.023923564204048304,
      "loss": 0.7956,
      "num_input_tokens_seen": 28642720,
      "step": 49345
    },
    {
      "epoch": 7.3503127792672025,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.023921997004418077,
      "loss": 0.7905,
      "num_input_tokens_seen": 28645504,
      "step": 49350
    },
    {
      "epoch": 7.351057491808162,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02392042965405978,
      "loss": 0.8073,
      "num_input_tokens_seen": 28648544,
      "step": 49355
    },
    {
      "epoch": 7.351802204349121,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02391886215299989,
      "loss": 0.8032,
      "num_input_tokens_seen": 28651744,
      "step": 49360
    },
    {
      "epoch": 7.3525469168900806,
      "grad_norm": 0.0234375,
      "learning_rate": 0.023917294501264886,
      "loss": 0.7765,
      "num_input_tokens_seen": 28654528,
      "step": 49365
    },
    {
      "epoch": 7.353291629431039,
      "grad_norm": 0.01422119140625,
      "learning_rate": 0.023915726698881253,
      "loss": 0.8044,
      "num_input_tokens_seen": 28657312,
      "step": 49370
    },
    {
      "epoch": 7.354036341971999,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02391415874587548,
      "loss": 0.8131,
      "num_input_tokens_seen": 28660544,
      "step": 49375
    },
    {
      "epoch": 7.354781054512958,
      "grad_norm": 0.0390625,
      "learning_rate": 0.02391259064227406,
      "loss": 0.8304,
      "num_input_tokens_seen": 28663488,
      "step": 49380
    },
    {
      "epoch": 7.355525767053917,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.023911022388103473,
      "loss": 0.7888,
      "num_input_tokens_seen": 28666304,
      "step": 49385
    },
    {
      "epoch": 7.356270479594876,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.023909453983390224,
      "loss": 0.8029,
      "num_input_tokens_seen": 28669408,
      "step": 49390
    },
    {
      "epoch": 7.357015192135836,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.02390788542816081,
      "loss": 0.7706,
      "num_input_tokens_seen": 28672160,
      "step": 49395
    },
    {
      "epoch": 7.3577599046767945,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.023906316722441718,
      "loss": 0.7856,
      "num_input_tokens_seen": 28675008,
      "step": 49400
    },
    {
      "epoch": 7.358504617217754,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.023904747866259454,
      "loss": 0.8254,
      "num_input_tokens_seen": 28677728,
      "step": 49405
    },
    {
      "epoch": 7.359249329758713,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02390317885964053,
      "loss": 0.8047,
      "num_input_tokens_seen": 28680352,
      "step": 49410
    },
    {
      "epoch": 7.359994042299673,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02390160970261145,
      "loss": 0.7892,
      "num_input_tokens_seen": 28683264,
      "step": 49415
    },
    {
      "epoch": 7.360738754840631,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.02390004039519872,
      "loss": 0.7971,
      "num_input_tokens_seen": 28685952,
      "step": 49420
    },
    {
      "epoch": 7.361483467381591,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.023898470937428844,
      "loss": 0.8007,
      "num_input_tokens_seen": 28688864,
      "step": 49425
    },
    {
      "epoch": 7.36222817992255,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.023896901329328358,
      "loss": 0.8159,
      "num_input_tokens_seen": 28691776,
      "step": 49430
    },
    {
      "epoch": 7.362972892463509,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.023895331570923762,
      "loss": 0.7983,
      "num_input_tokens_seen": 28694464,
      "step": 49435
    },
    {
      "epoch": 7.363717605004468,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02389376166224158,
      "loss": 0.7869,
      "num_input_tokens_seen": 28697440,
      "step": 49440
    },
    {
      "epoch": 7.364462317545428,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02389219160330833,
      "loss": 0.8146,
      "num_input_tokens_seen": 28700352,
      "step": 49445
    },
    {
      "epoch": 7.3652070300863866,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.023890621394150546,
      "loss": 0.7984,
      "num_input_tokens_seen": 28703552,
      "step": 49450
    },
    {
      "epoch": 7.365951742627346,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02388905103479475,
      "loss": 0.7933,
      "num_input_tokens_seen": 28706496,
      "step": 49455
    },
    {
      "epoch": 7.366696455168305,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.023887480525267464,
      "loss": 0.8105,
      "num_input_tokens_seen": 28709312,
      "step": 49460
    },
    {
      "epoch": 7.367441167709265,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.023885909865595235,
      "loss": 0.8114,
      "num_input_tokens_seen": 28712224,
      "step": 49465
    },
    {
      "epoch": 7.368185880250223,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.023884339055804584,
      "loss": 0.8099,
      "num_input_tokens_seen": 28714880,
      "step": 49470
    },
    {
      "epoch": 7.368930592791183,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02388276809592206,
      "loss": 0.8026,
      "num_input_tokens_seen": 28717952,
      "step": 49475
    },
    {
      "epoch": 7.369675305332142,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02388119698597419,
      "loss": 0.8141,
      "num_input_tokens_seen": 28720576,
      "step": 49480
    },
    {
      "epoch": 7.370420017873101,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.023879625725987532,
      "loss": 0.8016,
      "num_input_tokens_seen": 28723488,
      "step": 49485
    },
    {
      "epoch": 7.37116473041406,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.02387805431598862,
      "loss": 0.8117,
      "num_input_tokens_seen": 28726176,
      "step": 49490
    },
    {
      "epoch": 7.371909442955019,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.023876482756004,
      "loss": 0.8287,
      "num_input_tokens_seen": 28728992,
      "step": 49495
    },
    {
      "epoch": 7.372654155495979,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.023874911046060227,
      "loss": 0.7999,
      "num_input_tokens_seen": 28732032,
      "step": 49500
    },
    {
      "epoch": 7.373398868036937,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02387333918618385,
      "loss": 0.8039,
      "num_input_tokens_seen": 28734944,
      "step": 49505
    },
    {
      "epoch": 7.374143580577897,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.023871767176401427,
      "loss": 0.7979,
      "num_input_tokens_seen": 28737920,
      "step": 49510
    },
    {
      "epoch": 7.374888293118856,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02387019501673952,
      "loss": 0.7841,
      "num_input_tokens_seen": 28740896,
      "step": 49515
    },
    {
      "epoch": 7.375633005659815,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02386862270722468,
      "loss": 0.8099,
      "num_input_tokens_seen": 28743616,
      "step": 49520
    },
    {
      "epoch": 7.376377718200774,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.023867050247883468,
      "loss": 0.7997,
      "num_input_tokens_seen": 28746560,
      "step": 49525
    },
    {
      "epoch": 7.377122430741734,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02386547763874246,
      "loss": 0.8006,
      "num_input_tokens_seen": 28749504,
      "step": 49530
    },
    {
      "epoch": 7.3778671432826926,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.023863904879828216,
      "loss": 0.7919,
      "num_input_tokens_seen": 28752128,
      "step": 49535
    },
    {
      "epoch": 7.378611855823652,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.023862331971167303,
      "loss": 0.8137,
      "num_input_tokens_seen": 28755328,
      "step": 49540
    },
    {
      "epoch": 7.379356568364611,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.023860758912786308,
      "loss": 0.803,
      "num_input_tokens_seen": 28758240,
      "step": 49545
    },
    {
      "epoch": 7.380101280905571,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02385918570471179,
      "loss": 0.791,
      "num_input_tokens_seen": 28761088,
      "step": 49550
    },
    {
      "epoch": 7.380845993446529,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.023857612346970338,
      "loss": 0.818,
      "num_input_tokens_seen": 28763904,
      "step": 49555
    },
    {
      "epoch": 7.381590705987489,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.023856038839588527,
      "loss": 0.7979,
      "num_input_tokens_seen": 28766656,
      "step": 49560
    },
    {
      "epoch": 7.382335418528448,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.023854465182592936,
      "loss": 0.785,
      "num_input_tokens_seen": 28769280,
      "step": 49565
    },
    {
      "epoch": 7.383080131069407,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02385289137601016,
      "loss": 0.8,
      "num_input_tokens_seen": 28771872,
      "step": 49570
    },
    {
      "epoch": 7.383824843610366,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.02385131741986678,
      "loss": 0.8143,
      "num_input_tokens_seen": 28774464,
      "step": 49575
    },
    {
      "epoch": 7.384569556151326,
      "grad_norm": 0.05078125,
      "learning_rate": 0.02384974331418939,
      "loss": 0.8225,
      "num_input_tokens_seen": 28777280,
      "step": 49580
    },
    {
      "epoch": 7.385314268692285,
      "grad_norm": 0.07421875,
      "learning_rate": 0.023848169059004578,
      "loss": 0.8093,
      "num_input_tokens_seen": 28780352,
      "step": 49585
    },
    {
      "epoch": 7.386058981233244,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02384659465433894,
      "loss": 0.8041,
      "num_input_tokens_seen": 28783328,
      "step": 49590
    },
    {
      "epoch": 7.386803693774203,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.023845020100219088,
      "loss": 0.812,
      "num_input_tokens_seen": 28786208,
      "step": 49595
    },
    {
      "epoch": 7.387548406315163,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.023843445396671603,
      "loss": 0.7896,
      "num_input_tokens_seen": 28789088,
      "step": 49600
    },
    {
      "epoch": 7.388293118856121,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.023841870543723092,
      "loss": 0.8013,
      "num_input_tokens_seen": 28792064,
      "step": 49605
    },
    {
      "epoch": 7.389037831397081,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.023840295541400172,
      "loss": 0.809,
      "num_input_tokens_seen": 28794912,
      "step": 49610
    },
    {
      "epoch": 7.38978254393804,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02383872038972944,
      "loss": 0.8082,
      "num_input_tokens_seen": 28798016,
      "step": 49615
    },
    {
      "epoch": 7.390527256478999,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.023837145088737514,
      "loss": 0.7958,
      "num_input_tokens_seen": 28800960,
      "step": 49620
    },
    {
      "epoch": 7.391271969019958,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.023835569638451,
      "loss": 0.805,
      "num_input_tokens_seen": 28803872,
      "step": 49625
    },
    {
      "epoch": 7.392016681560918,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02383399403889652,
      "loss": 0.801,
      "num_input_tokens_seen": 28806976,
      "step": 49630
    },
    {
      "epoch": 7.392761394101877,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.023832418290100688,
      "loss": 0.8072,
      "num_input_tokens_seen": 28809856,
      "step": 49635
    },
    {
      "epoch": 7.393506106642836,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.023830842392090122,
      "loss": 0.8041,
      "num_input_tokens_seen": 28812672,
      "step": 49640
    },
    {
      "epoch": 7.394250819183795,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.023829266344891454,
      "loss": 0.805,
      "num_input_tokens_seen": 28815584,
      "step": 49645
    },
    {
      "epoch": 7.394995531724755,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.023827690148531304,
      "loss": 0.8054,
      "num_input_tokens_seen": 28818656,
      "step": 49650
    },
    {
      "epoch": 7.395740244265713,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.023826113803036297,
      "loss": 0.8133,
      "num_input_tokens_seen": 28821280,
      "step": 49655
    },
    {
      "epoch": 7.396484956806672,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.023824537308433074,
      "loss": 0.7955,
      "num_input_tokens_seen": 28824192,
      "step": 49660
    },
    {
      "epoch": 7.397229669347632,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.023822960664748257,
      "loss": 0.7991,
      "num_input_tokens_seen": 28827104,
      "step": 49665
    },
    {
      "epoch": 7.3979743818885915,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.023821383872008488,
      "loss": 0.7924,
      "num_input_tokens_seen": 28829920,
      "step": 49670
    },
    {
      "epoch": 7.39871909442955,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.023819806930240408,
      "loss": 0.812,
      "num_input_tokens_seen": 28832704,
      "step": 49675
    },
    {
      "epoch": 7.399463806970509,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.02381822983947065,
      "loss": 0.8063,
      "num_input_tokens_seen": 28835616,
      "step": 49680
    },
    {
      "epoch": 7.400208519511469,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.02381665259972586,
      "loss": 0.7918,
      "num_input_tokens_seen": 28838560,
      "step": 49685
    },
    {
      "epoch": 7.400953232052427,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02381507521103269,
      "loss": 0.8058,
      "num_input_tokens_seen": 28841344,
      "step": 49690
    },
    {
      "epoch": 7.401697944593387,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.023813497673417776,
      "loss": 0.8237,
      "num_input_tokens_seen": 28844000,
      "step": 49695
    },
    {
      "epoch": 7.402442657134346,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.023811919986907782,
      "loss": 0.8105,
      "num_input_tokens_seen": 28847008,
      "step": 49700
    },
    {
      "epoch": 7.403187369675305,
      "grad_norm": 0.021484375,
      "learning_rate": 0.023810342151529357,
      "loss": 0.7946,
      "num_input_tokens_seen": 28849760,
      "step": 49705
    },
    {
      "epoch": 7.403932082216264,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.023808764167309152,
      "loss": 0.7949,
      "num_input_tokens_seen": 28852896,
      "step": 49710
    },
    {
      "epoch": 7.404676794757224,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.023807186034273835,
      "loss": 0.7943,
      "num_input_tokens_seen": 28856096,
      "step": 49715
    },
    {
      "epoch": 7.405421507298183,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.023805607752450053,
      "loss": 0.8122,
      "num_input_tokens_seen": 28859136,
      "step": 49720
    },
    {
      "epoch": 7.406166219839142,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.023804029321864484,
      "loss": 0.7863,
      "num_input_tokens_seen": 28861984,
      "step": 49725
    },
    {
      "epoch": 7.406910932380101,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.023802450742543786,
      "loss": 0.8086,
      "num_input_tokens_seen": 28865088,
      "step": 49730
    },
    {
      "epoch": 7.407655644921061,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.023800872014514626,
      "loss": 0.7982,
      "num_input_tokens_seen": 28868064,
      "step": 49735
    },
    {
      "epoch": 7.408400357462019,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02379929313780368,
      "loss": 0.8233,
      "num_input_tokens_seen": 28870976,
      "step": 49740
    },
    {
      "epoch": 7.409145070002979,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.023797714112437623,
      "loss": 0.7876,
      "num_input_tokens_seen": 28873568,
      "step": 49745
    },
    {
      "epoch": 7.409889782543938,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.023796134938443122,
      "loss": 0.7946,
      "num_input_tokens_seen": 28876576,
      "step": 49750
    },
    {
      "epoch": 7.4106344950848975,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02379455561584687,
      "loss": 0.7935,
      "num_input_tokens_seen": 28879360,
      "step": 49755
    },
    {
      "epoch": 7.411379207625856,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.023792976144675532,
      "loss": 0.7943,
      "num_input_tokens_seen": 28882336,
      "step": 49760
    },
    {
      "epoch": 7.412123920166816,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.0237913965249558,
      "loss": 0.7883,
      "num_input_tokens_seen": 28884992,
      "step": 49765
    },
    {
      "epoch": 7.412868632707775,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02378981675671436,
      "loss": 0.8177,
      "num_input_tokens_seen": 28887840,
      "step": 49770
    },
    {
      "epoch": 7.413613345248734,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.023788236839977894,
      "loss": 0.8131,
      "num_input_tokens_seen": 28890848,
      "step": 49775
    },
    {
      "epoch": 7.414358057789693,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.023786656774773103,
      "loss": 0.7937,
      "num_input_tokens_seen": 28893728,
      "step": 49780
    },
    {
      "epoch": 7.415102770330653,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.023785076561126675,
      "loss": 0.7903,
      "num_input_tokens_seen": 28896704,
      "step": 49785
    },
    {
      "epoch": 7.415847482871611,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02378349619906531,
      "loss": 0.8078,
      "num_input_tokens_seen": 28899712,
      "step": 49790
    },
    {
      "epoch": 7.416592195412571,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.023781915688615703,
      "loss": 0.8133,
      "num_input_tokens_seen": 28902592,
      "step": 49795
    },
    {
      "epoch": 7.41733690795353,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.023780335029804558,
      "loss": 0.8201,
      "num_input_tokens_seen": 28905440,
      "step": 49800
    },
    {
      "epoch": 7.4180816204944895,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02377875422265857,
      "loss": 0.8063,
      "num_input_tokens_seen": 28908160,
      "step": 49805
    },
    {
      "epoch": 7.418826333035448,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02377717326720446,
      "loss": 0.8067,
      "num_input_tokens_seen": 28910848,
      "step": 49810
    },
    {
      "epoch": 7.419571045576408,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.023775592163468924,
      "loss": 0.8152,
      "num_input_tokens_seen": 28913664,
      "step": 49815
    },
    {
      "epoch": 7.420315758117367,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.023774010911478675,
      "loss": 0.7967,
      "num_input_tokens_seen": 28916512,
      "step": 49820
    },
    {
      "epoch": 7.421060470658326,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.02377242951126043,
      "loss": 0.8117,
      "num_input_tokens_seen": 28919648,
      "step": 49825
    },
    {
      "epoch": 7.421805183199285,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02377084796284091,
      "loss": 0.8007,
      "num_input_tokens_seen": 28922656,
      "step": 49830
    },
    {
      "epoch": 7.422549895740245,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.023769266266246824,
      "loss": 0.8064,
      "num_input_tokens_seen": 28925376,
      "step": 49835
    },
    {
      "epoch": 7.4232946082812035,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.023767684421504898,
      "loss": 0.8033,
      "num_input_tokens_seen": 28928384,
      "step": 49840
    },
    {
      "epoch": 7.424039320822162,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.023766102428641855,
      "loss": 0.7887,
      "num_input_tokens_seen": 28931328,
      "step": 49845
    },
    {
      "epoch": 7.424784033363122,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02376452028768442,
      "loss": 0.8062,
      "num_input_tokens_seen": 28934272,
      "step": 49850
    },
    {
      "epoch": 7.425528745904081,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02376293799865932,
      "loss": 0.8114,
      "num_input_tokens_seen": 28937056,
      "step": 49855
    },
    {
      "epoch": 7.42627345844504,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02376135556159329,
      "loss": 0.8084,
      "num_input_tokens_seen": 28940032,
      "step": 49860
    },
    {
      "epoch": 7.427018170985999,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.023759772976513063,
      "loss": 0.8116,
      "num_input_tokens_seen": 28942944,
      "step": 49865
    },
    {
      "epoch": 7.427762883526959,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.023758190243445377,
      "loss": 0.7927,
      "num_input_tokens_seen": 28946080,
      "step": 49870
    },
    {
      "epoch": 7.428507596067917,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.023756607362416967,
      "loss": 0.8084,
      "num_input_tokens_seen": 28948896,
      "step": 49875
    },
    {
      "epoch": 7.429252308608877,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.023755024333454576,
      "loss": 0.8124,
      "num_input_tokens_seen": 28951488,
      "step": 49880
    },
    {
      "epoch": 7.429997021149836,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.023753441156584947,
      "loss": 0.7945,
      "num_input_tokens_seen": 28954432,
      "step": 49885
    },
    {
      "epoch": 7.4307417336907955,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.023751857831834824,
      "loss": 0.8195,
      "num_input_tokens_seen": 28957408,
      "step": 49890
    },
    {
      "epoch": 7.431486446231754,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.023750274359230963,
      "loss": 0.8248,
      "num_input_tokens_seen": 28960160,
      "step": 49895
    },
    {
      "epoch": 7.432231158772714,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.023748690738800104,
      "loss": 0.7992,
      "num_input_tokens_seen": 28963040,
      "step": 49900
    },
    {
      "epoch": 7.432975871313673,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.023747106970569008,
      "loss": 0.7924,
      "num_input_tokens_seen": 28965856,
      "step": 49905
    },
    {
      "epoch": 7.433720583854632,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02374552305456443,
      "loss": 0.8058,
      "num_input_tokens_seen": 28968800,
      "step": 49910
    },
    {
      "epoch": 7.434465296395591,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.023743938990813132,
      "loss": 0.8007,
      "num_input_tokens_seen": 28972000,
      "step": 49915
    },
    {
      "epoch": 7.435210008936551,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.023742354779341866,
      "loss": 0.8003,
      "num_input_tokens_seen": 28974688,
      "step": 49920
    },
    {
      "epoch": 7.4359547214775095,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.023740770420177408,
      "loss": 0.7994,
      "num_input_tokens_seen": 28977504,
      "step": 49925
    },
    {
      "epoch": 7.436699434018469,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.023739185913346512,
      "loss": 0.8023,
      "num_input_tokens_seen": 28980672,
      "step": 49930
    },
    {
      "epoch": 7.437444146559428,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.023737601258875956,
      "loss": 0.8081,
      "num_input_tokens_seen": 28984160,
      "step": 49935
    },
    {
      "epoch": 7.4381888591003875,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.023736016456792505,
      "loss": 0.797,
      "num_input_tokens_seen": 28987200,
      "step": 49940
    },
    {
      "epoch": 7.438933571641346,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.023734431507122938,
      "loss": 0.7916,
      "num_input_tokens_seen": 28989984,
      "step": 49945
    },
    {
      "epoch": 7.439678284182306,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.023732846409894023,
      "loss": 0.8126,
      "num_input_tokens_seen": 28992928,
      "step": 49950
    },
    {
      "epoch": 7.440422996723265,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.023731261165132543,
      "loss": 0.802,
      "num_input_tokens_seen": 28995808,
      "step": 49955
    },
    {
      "epoch": 7.441167709264224,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.023729675772865284,
      "loss": 0.7878,
      "num_input_tokens_seen": 28998848,
      "step": 49960
    },
    {
      "epoch": 7.441912421805183,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.023728090233119028,
      "loss": 0.7993,
      "num_input_tokens_seen": 29001536,
      "step": 49965
    },
    {
      "epoch": 7.442657134346143,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02372650454592055,
      "loss": 0.8064,
      "num_input_tokens_seen": 29004256,
      "step": 49970
    },
    {
      "epoch": 7.4434018468871015,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.023724918711296648,
      "loss": 0.796,
      "num_input_tokens_seen": 29007232,
      "step": 49975
    },
    {
      "epoch": 7.444146559428061,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.023723332729274117,
      "loss": 0.7855,
      "num_input_tokens_seen": 29010272,
      "step": 49980
    },
    {
      "epoch": 7.44489127196902,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.023721746599879745,
      "loss": 0.8134,
      "num_input_tokens_seen": 29013280,
      "step": 49985
    },
    {
      "epoch": 7.4456359845099795,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.023720160323140324,
      "loss": 0.8015,
      "num_input_tokens_seen": 29016160,
      "step": 49990
    },
    {
      "epoch": 7.446380697050938,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.023718573899082662,
      "loss": 0.8122,
      "num_input_tokens_seen": 29019072,
      "step": 49995
    },
    {
      "epoch": 7.447125409591898,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.023716987327733554,
      "loss": 0.81,
      "num_input_tokens_seen": 29021920,
      "step": 50000
    },
    {
      "epoch": 7.447870122132857,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.0237154006091198,
      "loss": 0.7989,
      "num_input_tokens_seen": 29024832,
      "step": 50005
    },
    {
      "epoch": 7.4486148346738155,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.02371381374326822,
      "loss": 0.8243,
      "num_input_tokens_seen": 29027712,
      "step": 50010
    },
    {
      "epoch": 7.449359547214775,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02371222673020561,
      "loss": 0.8067,
      "num_input_tokens_seen": 29030624,
      "step": 50015
    },
    {
      "epoch": 7.450104259755734,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.02371063956995878,
      "loss": 0.7918,
      "num_input_tokens_seen": 29033824,
      "step": 50020
    },
    {
      "epoch": 7.4508489722966935,
      "grad_norm": 0.025390625,
      "learning_rate": 0.023709052262554545,
      "loss": 0.7881,
      "num_input_tokens_seen": 29036704,
      "step": 50025
    },
    {
      "epoch": 7.451593684837652,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02370746480801973,
      "loss": 0.7918,
      "num_input_tokens_seen": 29039712,
      "step": 50030
    },
    {
      "epoch": 7.452338397378612,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.023705877206381145,
      "loss": 0.7848,
      "num_input_tokens_seen": 29042688,
      "step": 50035
    },
    {
      "epoch": 7.453083109919571,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.023704289457665617,
      "loss": 0.8058,
      "num_input_tokens_seen": 29045344,
      "step": 50040
    },
    {
      "epoch": 7.45382782246053,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.023702701561899957,
      "loss": 0.8302,
      "num_input_tokens_seen": 29048160,
      "step": 50045
    },
    {
      "epoch": 7.454572535001489,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.023701113519111004,
      "loss": 0.7984,
      "num_input_tokens_seen": 29050752,
      "step": 50050
    },
    {
      "epoch": 7.455317247542449,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02369952532932558,
      "loss": 0.8238,
      "num_input_tokens_seen": 29053792,
      "step": 50055
    },
    {
      "epoch": 7.4560619600834075,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02369793699257052,
      "loss": 0.8087,
      "num_input_tokens_seen": 29056576,
      "step": 50060
    },
    {
      "epoch": 7.456806672624367,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02369634850887265,
      "loss": 0.8046,
      "num_input_tokens_seen": 29059328,
      "step": 50065
    },
    {
      "epoch": 7.457551385165326,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.023694759878258812,
      "loss": 0.8105,
      "num_input_tokens_seen": 29062240,
      "step": 50070
    },
    {
      "epoch": 7.4582960977062855,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.023693171100755844,
      "loss": 0.799,
      "num_input_tokens_seen": 29064928,
      "step": 50075
    },
    {
      "epoch": 7.459040810247244,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02369158217639059,
      "loss": 0.8089,
      "num_input_tokens_seen": 29067840,
      "step": 50080
    },
    {
      "epoch": 7.459785522788204,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.023689993105189884,
      "loss": 0.807,
      "num_input_tokens_seen": 29070528,
      "step": 50085
    },
    {
      "epoch": 7.460530235329163,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.023688403887180576,
      "loss": 0.7911,
      "num_input_tokens_seen": 29073280,
      "step": 50090
    },
    {
      "epoch": 7.461274947870122,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.023686814522389518,
      "loss": 0.8004,
      "num_input_tokens_seen": 29076128,
      "step": 50095
    },
    {
      "epoch": 7.462019660411081,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.023685225010843552,
      "loss": 0.8006,
      "num_input_tokens_seen": 29079392,
      "step": 50100
    },
    {
      "epoch": 7.462764372952041,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02368363535256954,
      "loss": 0.8073,
      "num_input_tokens_seen": 29082272,
      "step": 50105
    },
    {
      "epoch": 7.4635090854929995,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.023682045547594337,
      "loss": 0.8079,
      "num_input_tokens_seen": 29085056,
      "step": 50110
    },
    {
      "epoch": 7.464253798033959,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.023680455595944793,
      "loss": 0.7911,
      "num_input_tokens_seen": 29088032,
      "step": 50115
    },
    {
      "epoch": 7.464998510574918,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02367886549764778,
      "loss": 0.7973,
      "num_input_tokens_seen": 29090880,
      "step": 50120
    },
    {
      "epoch": 7.465743223115878,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.02367727525273015,
      "loss": 0.7976,
      "num_input_tokens_seen": 29093824,
      "step": 50125
    },
    {
      "epoch": 7.466487935656836,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02367568486121878,
      "loss": 0.7899,
      "num_input_tokens_seen": 29096736,
      "step": 50130
    },
    {
      "epoch": 7.467232648197796,
      "grad_norm": 0.04296875,
      "learning_rate": 0.023674094323140528,
      "loss": 0.7886,
      "num_input_tokens_seen": 29099776,
      "step": 50135
    },
    {
      "epoch": 7.467977360738755,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.023672503638522264,
      "loss": 0.8028,
      "num_input_tokens_seen": 29103008,
      "step": 50140
    },
    {
      "epoch": 7.468722073279714,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.02367091280739087,
      "loss": 0.7853,
      "num_input_tokens_seen": 29106144,
      "step": 50145
    },
    {
      "epoch": 7.469466785820673,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.02366932182977322,
      "loss": 0.7698,
      "num_input_tokens_seen": 29108800,
      "step": 50150
    },
    {
      "epoch": 7.470211498361633,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.02366773070569618,
      "loss": 0.814,
      "num_input_tokens_seen": 29111872,
      "step": 50155
    },
    {
      "epoch": 7.4709562109025915,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.023666139435186646,
      "loss": 0.8459,
      "num_input_tokens_seen": 29114688,
      "step": 50160
    },
    {
      "epoch": 7.471700923443551,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.023664548018271496,
      "loss": 0.7979,
      "num_input_tokens_seen": 29117856,
      "step": 50165
    },
    {
      "epoch": 7.47244563598451,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.02366295645497761,
      "loss": 0.8093,
      "num_input_tokens_seen": 29120640,
      "step": 50170
    },
    {
      "epoch": 7.473190348525469,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02366136474533188,
      "loss": 0.8203,
      "num_input_tokens_seen": 29123712,
      "step": 50175
    },
    {
      "epoch": 7.473935061066428,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.023659772889361195,
      "loss": 0.7945,
      "num_input_tokens_seen": 29126656,
      "step": 50180
    },
    {
      "epoch": 7.474679773607388,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02365818088709245,
      "loss": 0.7947,
      "num_input_tokens_seen": 29129344,
      "step": 50185
    },
    {
      "epoch": 7.475424486148347,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.023656588738552534,
      "loss": 0.8341,
      "num_input_tokens_seen": 29131968,
      "step": 50190
    },
    {
      "epoch": 7.4761691986893055,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.023654996443768353,
      "loss": 0.8421,
      "num_input_tokens_seen": 29134880,
      "step": 50195
    },
    {
      "epoch": 7.476913911230265,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.023653404002766807,
      "loss": 0.7963,
      "num_input_tokens_seen": 29137664,
      "step": 50200
    },
    {
      "epoch": 7.477658623771224,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.023651811415574788,
      "loss": 0.8241,
      "num_input_tokens_seen": 29140416,
      "step": 50205
    },
    {
      "epoch": 7.478403336312184,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.023650218682219212,
      "loss": 0.7906,
      "num_input_tokens_seen": 29143648,
      "step": 50210
    },
    {
      "epoch": 7.479148048853142,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.023648625802726985,
      "loss": 0.7988,
      "num_input_tokens_seen": 29146464,
      "step": 50215
    },
    {
      "epoch": 7.479892761394102,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.023647032777125012,
      "loss": 0.7906,
      "num_input_tokens_seen": 29148896,
      "step": 50220
    },
    {
      "epoch": 7.480637473935061,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02364543960544021,
      "loss": 0.7894,
      "num_input_tokens_seen": 29151840,
      "step": 50225
    },
    {
      "epoch": 7.48138218647602,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.023643846287699492,
      "loss": 0.7769,
      "num_input_tokens_seen": 29154720,
      "step": 50230
    },
    {
      "epoch": 7.482126899016979,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.023642252823929778,
      "loss": 0.7829,
      "num_input_tokens_seen": 29157504,
      "step": 50235
    },
    {
      "epoch": 7.482871611557939,
      "grad_norm": 0.02734375,
      "learning_rate": 0.023640659214157988,
      "loss": 0.7871,
      "num_input_tokens_seen": 29160480,
      "step": 50240
    },
    {
      "epoch": 7.4836163240988975,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02363906545841104,
      "loss": 0.8022,
      "num_input_tokens_seen": 29163552,
      "step": 50245
    },
    {
      "epoch": 7.484361036639857,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.023637471556715858,
      "loss": 0.7919,
      "num_input_tokens_seen": 29166336,
      "step": 50250
    },
    {
      "epoch": 7.485105749180816,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.023635877509099377,
      "loss": 0.7731,
      "num_input_tokens_seen": 29169216,
      "step": 50255
    },
    {
      "epoch": 7.485850461721776,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.02363428331558852,
      "loss": 0.8377,
      "num_input_tokens_seen": 29172064,
      "step": 50260
    },
    {
      "epoch": 7.486595174262734,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.023632688976210226,
      "loss": 0.7607,
      "num_input_tokens_seen": 29174848,
      "step": 50265
    },
    {
      "epoch": 7.487339886803694,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02363109449099142,
      "loss": 0.8031,
      "num_input_tokens_seen": 29178176,
      "step": 50270
    },
    {
      "epoch": 7.488084599344653,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.023629499859959047,
      "loss": 0.8074,
      "num_input_tokens_seen": 29181184,
      "step": 50275
    },
    {
      "epoch": 7.488829311885612,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.023627905083140048,
      "loss": 0.8004,
      "num_input_tokens_seen": 29184256,
      "step": 50280
    },
    {
      "epoch": 7.489574024426571,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.023626310160561357,
      "loss": 0.8063,
      "num_input_tokens_seen": 29187296,
      "step": 50285
    },
    {
      "epoch": 7.490318736967531,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.023624715092249924,
      "loss": 0.7452,
      "num_input_tokens_seen": 29190368,
      "step": 50290
    },
    {
      "epoch": 7.49106344950849,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.023623119878232696,
      "loss": 0.754,
      "num_input_tokens_seen": 29193472,
      "step": 50295
    },
    {
      "epoch": 7.491808162049449,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.023621524518536625,
      "loss": 0.8023,
      "num_input_tokens_seen": 29196352,
      "step": 50300
    },
    {
      "epoch": 7.492552874590408,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.023619929013188653,
      "loss": 0.8283,
      "num_input_tokens_seen": 29199392,
      "step": 50305
    },
    {
      "epoch": 7.493297587131368,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02361833336221575,
      "loss": 0.7642,
      "num_input_tokens_seen": 29202272,
      "step": 50310
    },
    {
      "epoch": 7.494042299672326,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.023616737565644856,
      "loss": 0.7901,
      "num_input_tokens_seen": 29204960,
      "step": 50315
    },
    {
      "epoch": 7.494787012213286,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.023615141623502943,
      "loss": 0.7836,
      "num_input_tokens_seen": 29207808,
      "step": 50320
    },
    {
      "epoch": 7.495531724754245,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.023613545535816967,
      "loss": 0.7979,
      "num_input_tokens_seen": 29210592,
      "step": 50325
    },
    {
      "epoch": 7.496276437295204,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.023611949302613892,
      "loss": 0.8711,
      "num_input_tokens_seen": 29213536,
      "step": 50330
    },
    {
      "epoch": 7.497021149836163,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02361035292392069,
      "loss": 0.8736,
      "num_input_tokens_seen": 29216704,
      "step": 50335
    },
    {
      "epoch": 7.497765862377123,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.023608756399764324,
      "loss": 0.8244,
      "num_input_tokens_seen": 29219872,
      "step": 50340
    },
    {
      "epoch": 7.498510574918082,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.023607159730171766,
      "loss": 0.8111,
      "num_input_tokens_seen": 29222720,
      "step": 50345
    },
    {
      "epoch": 7.499255287459041,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02360556291517,
      "loss": 0.8182,
      "num_input_tokens_seen": 29225632,
      "step": 50350
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.023603965954785987,
      "loss": 0.7815,
      "num_input_tokens_seen": 29228704,
      "step": 50355
    },
    {
      "epoch": 7.500744712540959,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.023602368849046717,
      "loss": 0.7824,
      "num_input_tokens_seen": 29231712,
      "step": 50360
    },
    {
      "epoch": 7.501489425081918,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02360077159797917,
      "loss": 0.7964,
      "num_input_tokens_seen": 29234560,
      "step": 50365
    },
    {
      "epoch": 7.502234137622878,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.023599174201610326,
      "loss": 0.7995,
      "num_input_tokens_seen": 29237376,
      "step": 50370
    },
    {
      "epoch": 7.502978850163837,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.023597576659967173,
      "loss": 0.8172,
      "num_input_tokens_seen": 29240352,
      "step": 50375
    },
    {
      "epoch": 7.503723562704796,
      "grad_norm": 0.029296875,
      "learning_rate": 0.023595978973076703,
      "loss": 0.8127,
      "num_input_tokens_seen": 29243040,
      "step": 50380
    },
    {
      "epoch": 7.504468275245755,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.0235943811409659,
      "loss": 0.7895,
      "num_input_tokens_seen": 29246080,
      "step": 50385
    },
    {
      "epoch": 7.505212987786714,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.023592783163661767,
      "loss": 0.8412,
      "num_input_tokens_seen": 29248992,
      "step": 50390
    },
    {
      "epoch": 7.505957700327674,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.023591185041191294,
      "loss": 0.8129,
      "num_input_tokens_seen": 29251680,
      "step": 50395
    },
    {
      "epoch": 7.506702412868632,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.023589586773581483,
      "loss": 0.8159,
      "num_input_tokens_seen": 29254528,
      "step": 50400
    },
    {
      "epoch": 7.507447125409592,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.023587988360859336,
      "loss": 0.7986,
      "num_input_tokens_seen": 29257408,
      "step": 50405
    },
    {
      "epoch": 7.508191837950551,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02358638980305185,
      "loss": 0.8078,
      "num_input_tokens_seen": 29260192,
      "step": 50410
    },
    {
      "epoch": 7.50893655049151,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.023584791100186034,
      "loss": 0.8218,
      "num_input_tokens_seen": 29263168,
      "step": 50415
    },
    {
      "epoch": 7.509681263032469,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.023583192252288897,
      "loss": 0.7918,
      "num_input_tokens_seen": 29266208,
      "step": 50420
    },
    {
      "epoch": 7.510425975573429,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.023581593259387457,
      "loss": 0.8058,
      "num_input_tokens_seen": 29269376,
      "step": 50425
    },
    {
      "epoch": 7.511170688114388,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.023579994121508713,
      "loss": 0.7987,
      "num_input_tokens_seen": 29271968,
      "step": 50430
    },
    {
      "epoch": 7.511915400655347,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.02357839483867969,
      "loss": 0.8097,
      "num_input_tokens_seen": 29275264,
      "step": 50435
    },
    {
      "epoch": 7.512660113196306,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02357679541092741,
      "loss": 0.8293,
      "num_input_tokens_seen": 29278080,
      "step": 50440
    },
    {
      "epoch": 7.513404825737266,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.023575195838278887,
      "loss": 0.7953,
      "num_input_tokens_seen": 29280704,
      "step": 50445
    },
    {
      "epoch": 7.514149538278224,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.023573596120761146,
      "loss": 0.8099,
      "num_input_tokens_seen": 29283424,
      "step": 50450
    },
    {
      "epoch": 7.514894250819184,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02357199625840121,
      "loss": 0.8149,
      "num_input_tokens_seen": 29286432,
      "step": 50455
    },
    {
      "epoch": 7.515638963360143,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02357039625122611,
      "loss": 0.7936,
      "num_input_tokens_seen": 29289504,
      "step": 50460
    },
    {
      "epoch": 7.5163836759011025,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02356879609926287,
      "loss": 0.8024,
      "num_input_tokens_seen": 29292480,
      "step": 50465
    },
    {
      "epoch": 7.517128388442061,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.023567195802538538,
      "loss": 0.7986,
      "num_input_tokens_seen": 29295488,
      "step": 50470
    },
    {
      "epoch": 7.517873100983021,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.023565595361080136,
      "loss": 0.8009,
      "num_input_tokens_seen": 29298208,
      "step": 50475
    },
    {
      "epoch": 7.51861781352398,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02356399477491471,
      "loss": 0.8045,
      "num_input_tokens_seen": 29301024,
      "step": 50480
    },
    {
      "epoch": 7.519362526064939,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.023562394044069297,
      "loss": 0.7998,
      "num_input_tokens_seen": 29303840,
      "step": 50485
    },
    {
      "epoch": 7.520107238605898,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.02356079316857093,
      "loss": 0.8168,
      "num_input_tokens_seen": 29306784,
      "step": 50490
    },
    {
      "epoch": 7.520851951146858,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02355919214844667,
      "loss": 0.8075,
      "num_input_tokens_seen": 29309536,
      "step": 50495
    },
    {
      "epoch": 7.521596663687816,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02355759098372356,
      "loss": 0.8011,
      "num_input_tokens_seen": 29312896,
      "step": 50500
    },
    {
      "epoch": 7.522341376228776,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.023555989674428644,
      "loss": 0.7938,
      "num_input_tokens_seen": 29315840,
      "step": 50505
    },
    {
      "epoch": 7.523086088769735,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02355438822058898,
      "loss": 0.7943,
      "num_input_tokens_seen": 29318880,
      "step": 50510
    },
    {
      "epoch": 7.5238308013106945,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02355278662223162,
      "loss": 0.8115,
      "num_input_tokens_seen": 29321568,
      "step": 50515
    },
    {
      "epoch": 7.524575513851653,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02355118487938362,
      "loss": 0.8108,
      "num_input_tokens_seen": 29324544,
      "step": 50520
    },
    {
      "epoch": 7.525320226392612,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02354958299207205,
      "loss": 0.8103,
      "num_input_tokens_seen": 29327232,
      "step": 50525
    },
    {
      "epoch": 7.526064938933572,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02354798096032396,
      "loss": 0.804,
      "num_input_tokens_seen": 29330144,
      "step": 50530
    },
    {
      "epoch": 7.526809651474531,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.023546378784166416,
      "loss": 0.7936,
      "num_input_tokens_seen": 29333152,
      "step": 50535
    },
    {
      "epoch": 7.52755436401549,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02354477646362649,
      "loss": 0.8089,
      "num_input_tokens_seen": 29335872,
      "step": 50540
    },
    {
      "epoch": 7.528299076556449,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.023543173998731255,
      "loss": 0.8155,
      "num_input_tokens_seen": 29338720,
      "step": 50545
    },
    {
      "epoch": 7.5290437890974085,
      "grad_norm": 0.0123291015625,
      "learning_rate": 0.023541571389507777,
      "loss": 0.7904,
      "num_input_tokens_seen": 29341600,
      "step": 50550
    },
    {
      "epoch": 7.529788501638367,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02353996863598313,
      "loss": 0.8098,
      "num_input_tokens_seen": 29344608,
      "step": 50555
    },
    {
      "epoch": 7.530533214179327,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.023538365738184394,
      "loss": 0.8044,
      "num_input_tokens_seen": 29347680,
      "step": 50560
    },
    {
      "epoch": 7.531277926720286,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.023536762696138643,
      "loss": 0.821,
      "num_input_tokens_seen": 29350688,
      "step": 50565
    },
    {
      "epoch": 7.532022639261245,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02353515950987297,
      "loss": 0.8001,
      "num_input_tokens_seen": 29353632,
      "step": 50570
    },
    {
      "epoch": 7.532767351802204,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.023533556179414442,
      "loss": 0.7939,
      "num_input_tokens_seen": 29356448,
      "step": 50575
    },
    {
      "epoch": 7.533512064343164,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.023531952704790162,
      "loss": 0.8005,
      "num_input_tokens_seen": 29359488,
      "step": 50580
    },
    {
      "epoch": 7.534256776884122,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.023530349086027212,
      "loss": 0.7997,
      "num_input_tokens_seen": 29362176,
      "step": 50585
    },
    {
      "epoch": 7.535001489425082,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.023528745323152682,
      "loss": 0.8095,
      "num_input_tokens_seen": 29364704,
      "step": 50590
    },
    {
      "epoch": 7.535746201966041,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.023527141416193668,
      "loss": 0.7928,
      "num_input_tokens_seen": 29367328,
      "step": 50595
    },
    {
      "epoch": 7.5364909145070005,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02352553736517727,
      "loss": 0.815,
      "num_input_tokens_seen": 29370464,
      "step": 50600
    },
    {
      "epoch": 7.537235627047959,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.02352393317013058,
      "loss": 0.8061,
      "num_input_tokens_seen": 29373344,
      "step": 50605
    },
    {
      "epoch": 7.537980339588919,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.023522328831080706,
      "loss": 0.7883,
      "num_input_tokens_seen": 29376256,
      "step": 50610
    },
    {
      "epoch": 7.538725052129878,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.023520724348054745,
      "loss": 0.7964,
      "num_input_tokens_seen": 29378944,
      "step": 50615
    },
    {
      "epoch": 7.539469764670837,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.023519119721079805,
      "loss": 0.8106,
      "num_input_tokens_seen": 29381888,
      "step": 50620
    },
    {
      "epoch": 7.540214477211796,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.023517514950182995,
      "loss": 0.7911,
      "num_input_tokens_seen": 29384768,
      "step": 50625
    },
    {
      "epoch": 7.540959189752756,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02351591003539143,
      "loss": 0.8002,
      "num_input_tokens_seen": 29387648,
      "step": 50630
    },
    {
      "epoch": 7.5417039022937145,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.023514304976732216,
      "loss": 0.8055,
      "num_input_tokens_seen": 29390816,
      "step": 50635
    },
    {
      "epoch": 7.542448614834674,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.02351269977423248,
      "loss": 0.8074,
      "num_input_tokens_seen": 29393632,
      "step": 50640
    },
    {
      "epoch": 7.543193327375633,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.02351109442791933,
      "loss": 0.8038,
      "num_input_tokens_seen": 29396576,
      "step": 50645
    },
    {
      "epoch": 7.5439380399165925,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02350948893781989,
      "loss": 0.7955,
      "num_input_tokens_seen": 29399584,
      "step": 50650
    },
    {
      "epoch": 7.544682752457551,
      "grad_norm": 0.03125,
      "learning_rate": 0.023507883303961283,
      "loss": 0.7997,
      "num_input_tokens_seen": 29402336,
      "step": 50655
    },
    {
      "epoch": 7.545427464998511,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.023506277526370634,
      "loss": 0.7885,
      "num_input_tokens_seen": 29405216,
      "step": 50660
    },
    {
      "epoch": 7.54617217753947,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.023504671605075068,
      "loss": 0.7998,
      "num_input_tokens_seen": 29407968,
      "step": 50665
    },
    {
      "epoch": 7.546916890080429,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02350306554010172,
      "loss": 0.8044,
      "num_input_tokens_seen": 29410976,
      "step": 50670
    },
    {
      "epoch": 7.547661602621388,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02350145933147773,
      "loss": 0.8002,
      "num_input_tokens_seen": 29413568,
      "step": 50675
    },
    {
      "epoch": 7.548406315162348,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.023499852979230216,
      "loss": 0.7831,
      "num_input_tokens_seen": 29416224,
      "step": 50680
    },
    {
      "epoch": 7.5491510277033065,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02349824648338633,
      "loss": 0.7807,
      "num_input_tokens_seen": 29419328,
      "step": 50685
    },
    {
      "epoch": 7.549895740244265,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.023496639843973204,
      "loss": 0.7771,
      "num_input_tokens_seen": 29422240,
      "step": 50690
    },
    {
      "epoch": 7.550640452785225,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.023495033061017988,
      "loss": 0.8104,
      "num_input_tokens_seen": 29425056,
      "step": 50695
    },
    {
      "epoch": 7.5513851653261845,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.023493426134547815,
      "loss": 0.8029,
      "num_input_tokens_seen": 29427680,
      "step": 50700
    },
    {
      "epoch": 7.552129877867143,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02349181906458984,
      "loss": 0.8251,
      "num_input_tokens_seen": 29430304,
      "step": 50705
    },
    {
      "epoch": 7.552874590408102,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02349021185117122,
      "loss": 0.8137,
      "num_input_tokens_seen": 29433056,
      "step": 50710
    },
    {
      "epoch": 7.553619302949062,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.0234886044943191,
      "loss": 0.8137,
      "num_input_tokens_seen": 29435776,
      "step": 50715
    },
    {
      "epoch": 7.554364015490021,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.023486996994060625,
      "loss": 0.8094,
      "num_input_tokens_seen": 29438432,
      "step": 50720
    },
    {
      "epoch": 7.55510872803098,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02348538935042297,
      "loss": 0.7893,
      "num_input_tokens_seen": 29441568,
      "step": 50725
    },
    {
      "epoch": 7.555853440571939,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02348378156343328,
      "loss": 0.7926,
      "num_input_tokens_seen": 29444320,
      "step": 50730
    },
    {
      "epoch": 7.5565981531128985,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.023482173633118726,
      "loss": 0.8136,
      "num_input_tokens_seen": 29447104,
      "step": 50735
    },
    {
      "epoch": 7.557342865653857,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02348056555950647,
      "loss": 0.8052,
      "num_input_tokens_seen": 29450176,
      "step": 50740
    },
    {
      "epoch": 7.558087578194817,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02347895734262368,
      "loss": 0.7928,
      "num_input_tokens_seen": 29452992,
      "step": 50745
    },
    {
      "epoch": 7.558832290735776,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.023477348982497523,
      "loss": 0.7758,
      "num_input_tokens_seen": 29455680,
      "step": 50750
    },
    {
      "epoch": 7.559577003276735,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.02347574047915517,
      "loss": 0.8131,
      "num_input_tokens_seen": 29458400,
      "step": 50755
    },
    {
      "epoch": 7.560321715817694,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02347413183262379,
      "loss": 0.808,
      "num_input_tokens_seen": 29461600,
      "step": 50760
    },
    {
      "epoch": 7.561066428358654,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02347252304293057,
      "loss": 0.7849,
      "num_input_tokens_seen": 29464448,
      "step": 50765
    },
    {
      "epoch": 7.5618111408996125,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.02347091411010269,
      "loss": 0.8484,
      "num_input_tokens_seen": 29467424,
      "step": 50770
    },
    {
      "epoch": 7.562555853440572,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.023469305034167318,
      "loss": 0.7989,
      "num_input_tokens_seen": 29469952,
      "step": 50775
    },
    {
      "epoch": 7.563300565981531,
      "grad_norm": 0.021484375,
      "learning_rate": 0.023467695815151648,
      "loss": 0.7766,
      "num_input_tokens_seen": 29473056,
      "step": 50780
    },
    {
      "epoch": 7.5640452785224905,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.023466086453082863,
      "loss": 0.8153,
      "num_input_tokens_seen": 29475840,
      "step": 50785
    },
    {
      "epoch": 7.564789991063449,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02346447694798815,
      "loss": 0.7995,
      "num_input_tokens_seen": 29478784,
      "step": 50790
    },
    {
      "epoch": 7.565534703604409,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.023462867299894705,
      "loss": 0.8048,
      "num_input_tokens_seen": 29481792,
      "step": 50795
    },
    {
      "epoch": 7.566279416145368,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.023461257508829717,
      "loss": 0.8111,
      "num_input_tokens_seen": 29484608,
      "step": 50800
    },
    {
      "epoch": 7.567024128686327,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.023459647574820382,
      "loss": 0.8073,
      "num_input_tokens_seen": 29487456,
      "step": 50805
    },
    {
      "epoch": 7.567768841227286,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.023458037497893904,
      "loss": 0.7969,
      "num_input_tokens_seen": 29490176,
      "step": 50810
    },
    {
      "epoch": 7.568513553768246,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.023456427278077473,
      "loss": 0.8022,
      "num_input_tokens_seen": 29493536,
      "step": 50815
    },
    {
      "epoch": 7.5692582663092045,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.023454816915398302,
      "loss": 0.8005,
      "num_input_tokens_seen": 29496480,
      "step": 50820
    },
    {
      "epoch": 7.570002978850164,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02345320640988359,
      "loss": 0.8014,
      "num_input_tokens_seen": 29499136,
      "step": 50825
    },
    {
      "epoch": 7.570747691391123,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.023451595761560546,
      "loss": 0.7897,
      "num_input_tokens_seen": 29502208,
      "step": 50830
    },
    {
      "epoch": 7.571492403932083,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.023449984970456382,
      "loss": 0.7897,
      "num_input_tokens_seen": 29505216,
      "step": 50835
    },
    {
      "epoch": 7.572237116473041,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02344837403659831,
      "loss": 0.8152,
      "num_input_tokens_seen": 29508288,
      "step": 50840
    },
    {
      "epoch": 7.572981829014001,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.023446762960013547,
      "loss": 0.8006,
      "num_input_tokens_seen": 29511200,
      "step": 50845
    },
    {
      "epoch": 7.57372654155496,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02344515174072931,
      "loss": 0.7987,
      "num_input_tokens_seen": 29514112,
      "step": 50850
    },
    {
      "epoch": 7.5744712540959185,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.023443540378772814,
      "loss": 0.7891,
      "num_input_tokens_seen": 29517056,
      "step": 50855
    },
    {
      "epoch": 7.575215966636878,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.023441928874171288,
      "loss": 0.8188,
      "num_input_tokens_seen": 29519968,
      "step": 50860
    },
    {
      "epoch": 7.575960679177838,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02344031722695195,
      "loss": 0.7995,
      "num_input_tokens_seen": 29522784,
      "step": 50865
    },
    {
      "epoch": 7.5767053917187965,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02343870543714203,
      "loss": 0.8025,
      "num_input_tokens_seen": 29525440,
      "step": 50870
    },
    {
      "epoch": 7.577450104259755,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.023437093504768764,
      "loss": 0.8068,
      "num_input_tokens_seen": 29528352,
      "step": 50875
    },
    {
      "epoch": 7.578194816800715,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.023435481429859375,
      "loss": 0.8099,
      "num_input_tokens_seen": 29531392,
      "step": 50880
    },
    {
      "epoch": 7.578939529341675,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.023433869212441102,
      "loss": 0.7941,
      "num_input_tokens_seen": 29534240,
      "step": 50885
    },
    {
      "epoch": 7.579684241882633,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02343225685254118,
      "loss": 0.7888,
      "num_input_tokens_seen": 29537472,
      "step": 50890
    },
    {
      "epoch": 7.580428954423592,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02343064435018685,
      "loss": 0.8158,
      "num_input_tokens_seen": 29540224,
      "step": 50895
    },
    {
      "epoch": 7.581173666964552,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.023429031705405348,
      "loss": 0.7966,
      "num_input_tokens_seen": 29543200,
      "step": 50900
    },
    {
      "epoch": 7.5819183795055105,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.023427418918223926,
      "loss": 0.8065,
      "num_input_tokens_seen": 29545664,
      "step": 50905
    },
    {
      "epoch": 7.58266309204647,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.023425805988669824,
      "loss": 0.802,
      "num_input_tokens_seen": 29548576,
      "step": 50910
    },
    {
      "epoch": 7.583407804587429,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.023424192916770295,
      "loss": 0.7867,
      "num_input_tokens_seen": 29551648,
      "step": 50915
    },
    {
      "epoch": 7.584152517128389,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.023422579702552588,
      "loss": 0.8102,
      "num_input_tokens_seen": 29554656,
      "step": 50920
    },
    {
      "epoch": 7.584897229669347,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02342096634604396,
      "loss": 0.8283,
      "num_input_tokens_seen": 29557440,
      "step": 50925
    },
    {
      "epoch": 7.585641942210307,
      "grad_norm": 0.02734375,
      "learning_rate": 0.023419352847271663,
      "loss": 0.787,
      "num_input_tokens_seen": 29560544,
      "step": 50930
    },
    {
      "epoch": 7.586386654751266,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02341773920626295,
      "loss": 0.8114,
      "num_input_tokens_seen": 29563328,
      "step": 50935
    },
    {
      "epoch": 7.587131367292225,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.0234161254230451,
      "loss": 0.7933,
      "num_input_tokens_seen": 29566464,
      "step": 50940
    },
    {
      "epoch": 7.587876079833184,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.023414511497645358,
      "loss": 0.7975,
      "num_input_tokens_seen": 29569280,
      "step": 50945
    },
    {
      "epoch": 7.588620792374144,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.023412897430090996,
      "loss": 0.8073,
      "num_input_tokens_seen": 29572256,
      "step": 50950
    },
    {
      "epoch": 7.5893655049151025,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.023411283220409283,
      "loss": 0.7995,
      "num_input_tokens_seen": 29575200,
      "step": 50955
    },
    {
      "epoch": 7.590110217456062,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02340966886862749,
      "loss": 0.796,
      "num_input_tokens_seen": 29578176,
      "step": 50960
    },
    {
      "epoch": 7.590854929997021,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02340805437477289,
      "loss": 0.8095,
      "num_input_tokens_seen": 29580960,
      "step": 50965
    },
    {
      "epoch": 7.591599642537981,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.023406439738872756,
      "loss": 0.822,
      "num_input_tokens_seen": 29583776,
      "step": 50970
    },
    {
      "epoch": 7.592344355078939,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.023404824960954366,
      "loss": 0.8108,
      "num_input_tokens_seen": 29586720,
      "step": 50975
    },
    {
      "epoch": 7.593089067619899,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.023403210041044998,
      "loss": 0.8024,
      "num_input_tokens_seen": 29589600,
      "step": 50980
    },
    {
      "epoch": 7.593833780160858,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.023401594979171936,
      "loss": 0.776,
      "num_input_tokens_seen": 29592288,
      "step": 50985
    },
    {
      "epoch": 7.594578492701817,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.023399979775362473,
      "loss": 0.8334,
      "num_input_tokens_seen": 29595456,
      "step": 50990
    },
    {
      "epoch": 7.595323205242776,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.02339836442964388,
      "loss": 0.8055,
      "num_input_tokens_seen": 29598624,
      "step": 50995
    },
    {
      "epoch": 7.596067917783736,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.023396748942043463,
      "loss": 0.8157,
      "num_input_tokens_seen": 29601472,
      "step": 51000
    },
    {
      "epoch": 7.596812630324695,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.02339513331258851,
      "loss": 0.8135,
      "num_input_tokens_seen": 29604416,
      "step": 51005
    },
    {
      "epoch": 7.597557342865654,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.023393517541306304,
      "loss": 0.7849,
      "num_input_tokens_seen": 29607264,
      "step": 51010
    },
    {
      "epoch": 7.598302055406613,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02339190162822415,
      "loss": 0.8012,
      "num_input_tokens_seen": 29609952,
      "step": 51015
    },
    {
      "epoch": 7.599046767947573,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.02339028557336935,
      "loss": 0.7829,
      "num_input_tokens_seen": 29613152,
      "step": 51020
    },
    {
      "epoch": 7.599791480488531,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.0233886693767692,
      "loss": 0.7847,
      "num_input_tokens_seen": 29616160,
      "step": 51025
    },
    {
      "epoch": 7.600536193029491,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02338705303845101,
      "loss": 0.7941,
      "num_input_tokens_seen": 29618976,
      "step": 51030
    },
    {
      "epoch": 7.60128090557045,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.023385436558442083,
      "loss": 0.8035,
      "num_input_tokens_seen": 29621952,
      "step": 51035
    },
    {
      "epoch": 7.6020256181114085,
      "grad_norm": 0.01336669921875,
      "learning_rate": 0.023383819936769728,
      "loss": 0.8187,
      "num_input_tokens_seen": 29624736,
      "step": 51040
    },
    {
      "epoch": 7.602770330652368,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.023382203173461256,
      "loss": 0.7887,
      "num_input_tokens_seen": 29627840,
      "step": 51045
    },
    {
      "epoch": 7.603515043193328,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02338058626854398,
      "loss": 0.7834,
      "num_input_tokens_seen": 29630752,
      "step": 51050
    },
    {
      "epoch": 7.604259755734287,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02337896922204522,
      "loss": 0.7952,
      "num_input_tokens_seen": 29633600,
      "step": 51055
    },
    {
      "epoch": 7.605004468275245,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.023377352033992285,
      "loss": 0.8228,
      "num_input_tokens_seen": 29636640,
      "step": 51060
    },
    {
      "epoch": 7.605749180816205,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.023375734704412503,
      "loss": 0.8131,
      "num_input_tokens_seen": 29639904,
      "step": 51065
    },
    {
      "epoch": 7.606493893357164,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.023374117233333196,
      "loss": 0.7956,
      "num_input_tokens_seen": 29642592,
      "step": 51070
    },
    {
      "epoch": 7.607238605898123,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02337249962078169,
      "loss": 0.7928,
      "num_input_tokens_seen": 29645472,
      "step": 51075
    },
    {
      "epoch": 7.607983318439082,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.023370881866785313,
      "loss": 0.8199,
      "num_input_tokens_seen": 29648800,
      "step": 51080
    },
    {
      "epoch": 7.608728030980042,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.023369263971371398,
      "loss": 0.83,
      "num_input_tokens_seen": 29651712,
      "step": 51085
    },
    {
      "epoch": 7.609472743521001,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.02336764593456727,
      "loss": 0.8176,
      "num_input_tokens_seen": 29654624,
      "step": 51090
    },
    {
      "epoch": 7.61021745606196,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02336602775640027,
      "loss": 0.7925,
      "num_input_tokens_seen": 29657312,
      "step": 51095
    },
    {
      "epoch": 7.610962168602919,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.023364409436897732,
      "loss": 0.8043,
      "num_input_tokens_seen": 29660064,
      "step": 51100
    },
    {
      "epoch": 7.611706881143879,
      "grad_norm": 0.03125,
      "learning_rate": 0.023362790976087,
      "loss": 0.8018,
      "num_input_tokens_seen": 29663072,
      "step": 51105
    },
    {
      "epoch": 7.612451593684837,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.023361172373995414,
      "loss": 0.7951,
      "num_input_tokens_seen": 29666240,
      "step": 51110
    },
    {
      "epoch": 7.613196306225797,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.023359553630650315,
      "loss": 0.8142,
      "num_input_tokens_seen": 29669152,
      "step": 51115
    },
    {
      "epoch": 7.613941018766756,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.023357934746079057,
      "loss": 0.7972,
      "num_input_tokens_seen": 29671776,
      "step": 51120
    },
    {
      "epoch": 7.614685731307715,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02335631572030899,
      "loss": 0.829,
      "num_input_tokens_seen": 29674496,
      "step": 51125
    },
    {
      "epoch": 7.615430443848674,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.023354696553367457,
      "loss": 0.8019,
      "num_input_tokens_seen": 29676896,
      "step": 51130
    },
    {
      "epoch": 7.616175156389634,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.023353077245281816,
      "loss": 0.805,
      "num_input_tokens_seen": 29679712,
      "step": 51135
    },
    {
      "epoch": 7.616919868930593,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.023351457796079424,
      "loss": 0.8217,
      "num_input_tokens_seen": 29682496,
      "step": 51140
    },
    {
      "epoch": 7.617664581471552,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.023349838205787645,
      "loss": 0.8176,
      "num_input_tokens_seen": 29685472,
      "step": 51145
    },
    {
      "epoch": 7.618409294012511,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02334821847443383,
      "loss": 0.8075,
      "num_input_tokens_seen": 29688160,
      "step": 51150
    },
    {
      "epoch": 7.619154006553471,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.023346598602045354,
      "loss": 0.8055,
      "num_input_tokens_seen": 29690976,
      "step": 51155
    },
    {
      "epoch": 7.619898719094429,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.023344978588649572,
      "loss": 0.7955,
      "num_input_tokens_seen": 29693888,
      "step": 51160
    },
    {
      "epoch": 7.620643431635389,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02334335843427387,
      "loss": 0.8105,
      "num_input_tokens_seen": 29696320,
      "step": 51165
    },
    {
      "epoch": 7.621388144176348,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.023341738138945595,
      "loss": 0.7961,
      "num_input_tokens_seen": 29699232,
      "step": 51170
    },
    {
      "epoch": 7.6221328567173074,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02334011770269214,
      "loss": 0.813,
      "num_input_tokens_seen": 29702176,
      "step": 51175
    },
    {
      "epoch": 7.622877569258266,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.023338497125540866,
      "loss": 0.8282,
      "num_input_tokens_seen": 29704832,
      "step": 51180
    },
    {
      "epoch": 7.623622281799226,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.023336876407519164,
      "loss": 0.8036,
      "num_input_tokens_seen": 29707808,
      "step": 51185
    },
    {
      "epoch": 7.624366994340185,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.023335255548654406,
      "loss": 0.8197,
      "num_input_tokens_seen": 29710688,
      "step": 51190
    },
    {
      "epoch": 7.625111706881144,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.023333634548973973,
      "loss": 0.809,
      "num_input_tokens_seen": 29713472,
      "step": 51195
    },
    {
      "epoch": 7.625856419422103,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02333201340850526,
      "loss": 0.798,
      "num_input_tokens_seen": 29716480,
      "step": 51200
    },
    {
      "epoch": 7.626601131963062,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.023330392127275643,
      "loss": 0.8155,
      "num_input_tokens_seen": 29719232,
      "step": 51205
    },
    {
      "epoch": 7.627345844504021,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.023328770705312522,
      "loss": 0.8105,
      "num_input_tokens_seen": 29722016,
      "step": 51210
    },
    {
      "epoch": 7.628090557044981,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02332714914264329,
      "loss": 0.7998,
      "num_input_tokens_seen": 29724832,
      "step": 51215
    },
    {
      "epoch": 7.62883526958594,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.023325527439295328,
      "loss": 0.8061,
      "num_input_tokens_seen": 29727680,
      "step": 51220
    },
    {
      "epoch": 7.629579982126899,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.023323905595296044,
      "loss": 0.8067,
      "num_input_tokens_seen": 29730624,
      "step": 51225
    },
    {
      "epoch": 7.630324694667858,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.023322283610672834,
      "loss": 0.8055,
      "num_input_tokens_seen": 29733536,
      "step": 51230
    },
    {
      "epoch": 7.631069407208818,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.023320661485453104,
      "loss": 0.8099,
      "num_input_tokens_seen": 29736608,
      "step": 51235
    },
    {
      "epoch": 7.631814119749777,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.023319039219664252,
      "loss": 0.7925,
      "num_input_tokens_seen": 29739648,
      "step": 51240
    },
    {
      "epoch": 7.632558832290735,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02331741681333369,
      "loss": 0.8043,
      "num_input_tokens_seen": 29742528,
      "step": 51245
    },
    {
      "epoch": 7.633303544831695,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.02331579426648882,
      "loss": 0.7981,
      "num_input_tokens_seen": 29745504,
      "step": 51250
    },
    {
      "epoch": 7.634048257372654,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.023314171579157064,
      "loss": 0.7994,
      "num_input_tokens_seen": 29748384,
      "step": 51255
    },
    {
      "epoch": 7.6347929699136134,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.023312548751365826,
      "loss": 0.819,
      "num_input_tokens_seen": 29751328,
      "step": 51260
    },
    {
      "epoch": 7.635537682454572,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.023310925783142523,
      "loss": 0.8096,
      "num_input_tokens_seen": 29754176,
      "step": 51265
    },
    {
      "epoch": 7.636282394995532,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02330930267451458,
      "loss": 0.8095,
      "num_input_tokens_seen": 29757024,
      "step": 51270
    },
    {
      "epoch": 7.637027107536491,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02330767942550941,
      "loss": 0.7916,
      "num_input_tokens_seen": 29760128,
      "step": 51275
    },
    {
      "epoch": 7.63777182007745,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.02330605603615444,
      "loss": 0.8053,
      "num_input_tokens_seen": 29762752,
      "step": 51280
    },
    {
      "epoch": 7.638516532618409,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.0233044325064771,
      "loss": 0.7937,
      "num_input_tokens_seen": 29765984,
      "step": 51285
    },
    {
      "epoch": 7.639261245159369,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02330280883650481,
      "loss": 0.8,
      "num_input_tokens_seen": 29768768,
      "step": 51290
    },
    {
      "epoch": 7.640005957700327,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.023301185026265003,
      "loss": 0.7977,
      "num_input_tokens_seen": 29771520,
      "step": 51295
    },
    {
      "epoch": 7.640750670241287,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02329956107578511,
      "loss": 0.7901,
      "num_input_tokens_seen": 29774592,
      "step": 51300
    },
    {
      "epoch": 7.641495382782246,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.023297936985092572,
      "loss": 0.8041,
      "num_input_tokens_seen": 29777344,
      "step": 51305
    },
    {
      "epoch": 7.6422400953232055,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.02329631275421482,
      "loss": 0.8001,
      "num_input_tokens_seen": 29780256,
      "step": 51310
    },
    {
      "epoch": 7.642984807864164,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.023294688383179295,
      "loss": 0.804,
      "num_input_tokens_seen": 29783296,
      "step": 51315
    },
    {
      "epoch": 7.643729520405124,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.023293063872013443,
      "loss": 0.811,
      "num_input_tokens_seen": 29786240,
      "step": 51320
    },
    {
      "epoch": 7.644474232946083,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0232914392207447,
      "loss": 0.8008,
      "num_input_tokens_seen": 29789120,
      "step": 51325
    },
    {
      "epoch": 7.645218945487042,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.023289814429400525,
      "loss": 0.7979,
      "num_input_tokens_seen": 29791840,
      "step": 51330
    },
    {
      "epoch": 7.645963658028001,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.023288189498008358,
      "loss": 0.8065,
      "num_input_tokens_seen": 29794624,
      "step": 51335
    },
    {
      "epoch": 7.646708370568961,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.023286564426595654,
      "loss": 0.8076,
      "num_input_tokens_seen": 29797376,
      "step": 51340
    },
    {
      "epoch": 7.6474530831099194,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02328493921518986,
      "loss": 0.8059,
      "num_input_tokens_seen": 29800256,
      "step": 51345
    },
    {
      "epoch": 7.648197795650879,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.023283313863818444,
      "loss": 0.8065,
      "num_input_tokens_seen": 29803360,
      "step": 51350
    },
    {
      "epoch": 7.648942508191838,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.023281688372508862,
      "loss": 0.8093,
      "num_input_tokens_seen": 29806720,
      "step": 51355
    },
    {
      "epoch": 7.6496872207327975,
      "grad_norm": 0.0234375,
      "learning_rate": 0.023280062741288566,
      "loss": 0.7756,
      "num_input_tokens_seen": 29809408,
      "step": 51360
    },
    {
      "epoch": 7.650431933273756,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02327843697018503,
      "loss": 0.8124,
      "num_input_tokens_seen": 29812352,
      "step": 51365
    },
    {
      "epoch": 7.651176645814716,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.023276811059225713,
      "loss": 0.7867,
      "num_input_tokens_seen": 29815424,
      "step": 51370
    },
    {
      "epoch": 7.651921358355675,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.023275185008438084,
      "loss": 0.8,
      "num_input_tokens_seen": 29818016,
      "step": 51375
    },
    {
      "epoch": 7.652666070896634,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.023273558817849614,
      "loss": 0.8038,
      "num_input_tokens_seen": 29821024,
      "step": 51380
    },
    {
      "epoch": 7.653410783437593,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.02327193248748778,
      "loss": 0.8018,
      "num_input_tokens_seen": 29824128,
      "step": 51385
    },
    {
      "epoch": 7.654155495978552,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.023270306017380053,
      "loss": 0.8114,
      "num_input_tokens_seen": 29827008,
      "step": 51390
    },
    {
      "epoch": 7.6549002085195115,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.02326867940755391,
      "loss": 0.8065,
      "num_input_tokens_seen": 29829920,
      "step": 51395
    },
    {
      "epoch": 7.655644921060471,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.023267052658036837,
      "loss": 0.8106,
      "num_input_tokens_seen": 29832800,
      "step": 51400
    },
    {
      "epoch": 7.65638963360143,
      "grad_norm": 0.01361083984375,
      "learning_rate": 0.023265425768856308,
      "loss": 0.803,
      "num_input_tokens_seen": 29835584,
      "step": 51405
    },
    {
      "epoch": 7.657134346142389,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.02326379874003981,
      "loss": 0.8246,
      "num_input_tokens_seen": 29838784,
      "step": 51410
    },
    {
      "epoch": 7.657879058683348,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.023262171571614833,
      "loss": 0.8126,
      "num_input_tokens_seen": 29841760,
      "step": 51415
    },
    {
      "epoch": 7.658623771224307,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.023260544263608865,
      "loss": 0.8277,
      "num_input_tokens_seen": 29844608,
      "step": 51420
    },
    {
      "epoch": 7.659368483765267,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.023258916816049396,
      "loss": 0.7953,
      "num_input_tokens_seen": 29847488,
      "step": 51425
    },
    {
      "epoch": 7.6601131963062254,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.02325728922896392,
      "loss": 0.7896,
      "num_input_tokens_seen": 29850560,
      "step": 51430
    },
    {
      "epoch": 7.660857908847185,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02325566150237994,
      "loss": 0.8145,
      "num_input_tokens_seen": 29853632,
      "step": 51435
    },
    {
      "epoch": 7.661602621388144,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.023254033636324944,
      "loss": 0.8015,
      "num_input_tokens_seen": 29856512,
      "step": 51440
    },
    {
      "epoch": 7.6623473339291035,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02325240563082644,
      "loss": 0.7881,
      "num_input_tokens_seen": 29859424,
      "step": 51445
    },
    {
      "epoch": 7.663092046470062,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.023250777485911936,
      "loss": 0.7979,
      "num_input_tokens_seen": 29862176,
      "step": 51450
    },
    {
      "epoch": 7.663836759011022,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.023249149201608924,
      "loss": 0.8077,
      "num_input_tokens_seen": 29865120,
      "step": 51455
    },
    {
      "epoch": 7.664581471551981,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.023247520777944924,
      "loss": 0.7853,
      "num_input_tokens_seen": 29868096,
      "step": 51460
    },
    {
      "epoch": 7.66532618409294,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02324589221494744,
      "loss": 0.7948,
      "num_input_tokens_seen": 29870880,
      "step": 51465
    },
    {
      "epoch": 7.666070896633899,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02324426351264399,
      "loss": 0.7936,
      "num_input_tokens_seen": 29873632,
      "step": 51470
    },
    {
      "epoch": 7.666815609174859,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02324263467106209,
      "loss": 0.819,
      "num_input_tokens_seen": 29876448,
      "step": 51475
    },
    {
      "epoch": 7.6675603217158175,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02324100569022925,
      "loss": 0.8094,
      "num_input_tokens_seen": 29879424,
      "step": 51480
    },
    {
      "epoch": 7.668305034256777,
      "grad_norm": 0.029296875,
      "learning_rate": 0.023239376570172998,
      "loss": 0.8074,
      "num_input_tokens_seen": 29882336,
      "step": 51485
    },
    {
      "epoch": 7.669049746797736,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.023237747310920855,
      "loss": 0.7883,
      "num_input_tokens_seen": 29885088,
      "step": 51490
    },
    {
      "epoch": 7.6697944593386955,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02323611791250034,
      "loss": 0.7986,
      "num_input_tokens_seen": 29888032,
      "step": 51495
    },
    {
      "epoch": 7.670539171879654,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02323448837493898,
      "loss": 0.7997,
      "num_input_tokens_seen": 29890880,
      "step": 51500
    },
    {
      "epoch": 7.671283884420614,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.023232858698264313,
      "loss": 0.8057,
      "num_input_tokens_seen": 29893952,
      "step": 51505
    },
    {
      "epoch": 7.672028596961573,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.023231228882503865,
      "loss": 0.8056,
      "num_input_tokens_seen": 29896960,
      "step": 51510
    },
    {
      "epoch": 7.672773309502532,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.023229598927685172,
      "loss": 0.8059,
      "num_input_tokens_seen": 29900288,
      "step": 51515
    },
    {
      "epoch": 7.673518022043491,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02322796883383577,
      "loss": 0.8194,
      "num_input_tokens_seen": 29903520,
      "step": 51520
    },
    {
      "epoch": 7.674262734584451,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02322633860098319,
      "loss": 0.8254,
      "num_input_tokens_seen": 29906368,
      "step": 51525
    },
    {
      "epoch": 7.6750074471254095,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.023224708229154987,
      "loss": 0.7901,
      "num_input_tokens_seen": 29909376,
      "step": 51530
    },
    {
      "epoch": 7.675752159666369,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.0232230777183787,
      "loss": 0.8105,
      "num_input_tokens_seen": 29912544,
      "step": 51535
    },
    {
      "epoch": 7.676496872207328,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02322144706868187,
      "loss": 0.8023,
      "num_input_tokens_seen": 29915168,
      "step": 51540
    },
    {
      "epoch": 7.6772415847482876,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02321981628009204,
      "loss": 0.7946,
      "num_input_tokens_seen": 29918272,
      "step": 51545
    },
    {
      "epoch": 7.677986297289246,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.023218185352636776,
      "loss": 0.8306,
      "num_input_tokens_seen": 29921440,
      "step": 51550
    },
    {
      "epoch": 7.678731009830205,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.02321655428634362,
      "loss": 0.7955,
      "num_input_tokens_seen": 29924352,
      "step": 51555
    },
    {
      "epoch": 7.679475722371165,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02321492308124013,
      "loss": 0.8074,
      "num_input_tokens_seen": 29927488,
      "step": 51560
    },
    {
      "epoch": 7.680220434912124,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02321329173735387,
      "loss": 0.8061,
      "num_input_tokens_seen": 29930272,
      "step": 51565
    },
    {
      "epoch": 7.680965147453083,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02321166025471239,
      "loss": 0.8007,
      "num_input_tokens_seen": 29933152,
      "step": 51570
    },
    {
      "epoch": 7.681709859994042,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.023210028633343253,
      "loss": 0.7957,
      "num_input_tokens_seen": 29935936,
      "step": 51575
    },
    {
      "epoch": 7.6824545725350015,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02320839687327403,
      "loss": 0.7871,
      "num_input_tokens_seen": 29938688,
      "step": 51580
    },
    {
      "epoch": 7.683199285075961,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02320676497453228,
      "loss": 0.7906,
      "num_input_tokens_seen": 29941472,
      "step": 51585
    },
    {
      "epoch": 7.68394399761692,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02320513293714558,
      "loss": 0.8095,
      "num_input_tokens_seen": 29944224,
      "step": 51590
    },
    {
      "epoch": 7.684688710157879,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.023203500761141502,
      "loss": 0.8057,
      "num_input_tokens_seen": 29947296,
      "step": 51595
    },
    {
      "epoch": 7.685433422698838,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.023201868446547612,
      "loss": 0.8059,
      "num_input_tokens_seen": 29950336,
      "step": 51600
    },
    {
      "epoch": 7.686178135239797,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.02320023599339149,
      "loss": 0.8032,
      "num_input_tokens_seen": 29953184,
      "step": 51605
    },
    {
      "epoch": 7.686922847780757,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.023198603401700716,
      "loss": 0.8065,
      "num_input_tokens_seen": 29956000,
      "step": 51610
    },
    {
      "epoch": 7.6876675603217155,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02319697067150287,
      "loss": 0.7958,
      "num_input_tokens_seen": 29958944,
      "step": 51615
    },
    {
      "epoch": 7.688412272862675,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.02319533780282554,
      "loss": 0.817,
      "num_input_tokens_seen": 29962016,
      "step": 51620
    },
    {
      "epoch": 7.689156985403634,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.0231937047956963,
      "loss": 0.8029,
      "num_input_tokens_seen": 29964928,
      "step": 51625
    },
    {
      "epoch": 7.6899016979445936,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.023192071650142747,
      "loss": 0.8039,
      "num_input_tokens_seen": 29968000,
      "step": 51630
    },
    {
      "epoch": 7.690646410485552,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.023190438366192476,
      "loss": 0.791,
      "num_input_tokens_seen": 29970912,
      "step": 51635
    },
    {
      "epoch": 7.691391123026512,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.023188804943873067,
      "loss": 0.8054,
      "num_input_tokens_seen": 29973824,
      "step": 51640
    },
    {
      "epoch": 7.692135835567471,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.023187171383212118,
      "loss": 0.7913,
      "num_input_tokens_seen": 29976736,
      "step": 51645
    },
    {
      "epoch": 7.69288054810843,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02318553768423724,
      "loss": 0.8083,
      "num_input_tokens_seen": 29979712,
      "step": 51650
    },
    {
      "epoch": 7.693625260649389,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.023183903846976016,
      "loss": 0.798,
      "num_input_tokens_seen": 29982720,
      "step": 51655
    },
    {
      "epoch": 7.694369973190349,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02318226987145605,
      "loss": 0.8116,
      "num_input_tokens_seen": 29985632,
      "step": 51660
    },
    {
      "epoch": 7.6951146857313075,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02318063575770496,
      "loss": 0.7911,
      "num_input_tokens_seen": 29989120,
      "step": 51665
    },
    {
      "epoch": 7.695859398272267,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.023179001505750335,
      "loss": 0.8079,
      "num_input_tokens_seen": 29991968,
      "step": 51670
    },
    {
      "epoch": 7.696604110813226,
      "grad_norm": 0.033203125,
      "learning_rate": 0.023177367115619796,
      "loss": 0.8153,
      "num_input_tokens_seen": 29994720,
      "step": 51675
    },
    {
      "epoch": 7.697348823354186,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.023175732587340948,
      "loss": 0.8117,
      "num_input_tokens_seen": 29997728,
      "step": 51680
    },
    {
      "epoch": 7.698093535895144,
      "grad_norm": 0.021484375,
      "learning_rate": 0.023174097920941415,
      "loss": 0.8159,
      "num_input_tokens_seen": 30000480,
      "step": 51685
    },
    {
      "epoch": 7.698838248436104,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.023172463116448803,
      "loss": 0.8093,
      "num_input_tokens_seen": 30003648,
      "step": 51690
    },
    {
      "epoch": 7.699582960977063,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.023170828173890726,
      "loss": 0.811,
      "num_input_tokens_seen": 30006432,
      "step": 51695
    },
    {
      "epoch": 7.700327673518022,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.023169193093294814,
      "loss": 0.8112,
      "num_input_tokens_seen": 30009248,
      "step": 51700
    },
    {
      "epoch": 7.701072386058981,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02316755787468869,
      "loss": 0.7903,
      "num_input_tokens_seen": 30011968,
      "step": 51705
    },
    {
      "epoch": 7.701817098599941,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.023165922518099978,
      "loss": 0.8225,
      "num_input_tokens_seen": 30014816,
      "step": 51710
    },
    {
      "epoch": 7.7025618111408996,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.023164287023556298,
      "loss": 0.8021,
      "num_input_tokens_seen": 30017696,
      "step": 51715
    },
    {
      "epoch": 7.703306523681858,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.023162651391085292,
      "loss": 0.797,
      "num_input_tokens_seen": 30020608,
      "step": 51720
    },
    {
      "epoch": 7.704051236222818,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.023161015620714584,
      "loss": 0.8026,
      "num_input_tokens_seen": 30023744,
      "step": 51725
    },
    {
      "epoch": 7.704795948763778,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.02315937971247182,
      "loss": 0.8006,
      "num_input_tokens_seen": 30026880,
      "step": 51730
    },
    {
      "epoch": 7.705540661304736,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02315774366638462,
      "loss": 0.8118,
      "num_input_tokens_seen": 30029888,
      "step": 51735
    },
    {
      "epoch": 7.706285373845695,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.023156107482480634,
      "loss": 0.7998,
      "num_input_tokens_seen": 30033024,
      "step": 51740
    },
    {
      "epoch": 7.707030086386655,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.023154471160787497,
      "loss": 0.8213,
      "num_input_tokens_seen": 30035744,
      "step": 51745
    },
    {
      "epoch": 7.707774798927614,
      "grad_norm": 0.01953125,
      "learning_rate": 0.023152834701332864,
      "loss": 0.7949,
      "num_input_tokens_seen": 30038496,
      "step": 51750
    },
    {
      "epoch": 7.708519511468573,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02315119810414437,
      "loss": 0.7892,
      "num_input_tokens_seen": 30041376,
      "step": 51755
    },
    {
      "epoch": 7.709264224009532,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02314956136924967,
      "loss": 0.7993,
      "num_input_tokens_seen": 30044224,
      "step": 51760
    },
    {
      "epoch": 7.710008936550492,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.023147924496676422,
      "loss": 0.8025,
      "num_input_tokens_seen": 30046816,
      "step": 51765
    },
    {
      "epoch": 7.71075364909145,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02314628748645226,
      "loss": 0.7979,
      "num_input_tokens_seen": 30049600,
      "step": 51770
    },
    {
      "epoch": 7.71149836163241,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.023144650338604855,
      "loss": 0.7918,
      "num_input_tokens_seen": 30052736,
      "step": 51775
    },
    {
      "epoch": 7.712243074173369,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.02314301305316186,
      "loss": 0.7996,
      "num_input_tokens_seen": 30055520,
      "step": 51780
    },
    {
      "epoch": 7.712987786714328,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.023141375630150937,
      "loss": 0.8031,
      "num_input_tokens_seen": 30058528,
      "step": 51785
    },
    {
      "epoch": 7.713732499255287,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.023139738069599743,
      "loss": 0.8012,
      "num_input_tokens_seen": 30061312,
      "step": 51790
    },
    {
      "epoch": 7.714477211796247,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.023138100371535953,
      "loss": 0.8213,
      "num_input_tokens_seen": 30064320,
      "step": 51795
    },
    {
      "epoch": 7.7152219243372056,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02313646253598722,
      "loss": 0.792,
      "num_input_tokens_seen": 30067168,
      "step": 51800
    },
    {
      "epoch": 7.715966636878165,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02313482456298123,
      "loss": 0.8118,
      "num_input_tokens_seen": 30070144,
      "step": 51805
    },
    {
      "epoch": 7.716711349419124,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02313318645254564,
      "loss": 0.7691,
      "num_input_tokens_seen": 30073088,
      "step": 51810
    },
    {
      "epoch": 7.717456061960084,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.023131548204708138,
      "loss": 0.8079,
      "num_input_tokens_seen": 30075872,
      "step": 51815
    },
    {
      "epoch": 7.718200774501042,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02312990981949639,
      "loss": 0.8064,
      "num_input_tokens_seen": 30078816,
      "step": 51820
    },
    {
      "epoch": 7.718945487042002,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.023128271296938077,
      "loss": 0.7826,
      "num_input_tokens_seen": 30081472,
      "step": 51825
    },
    {
      "epoch": 7.719690199582961,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.023126632637060888,
      "loss": 0.7916,
      "num_input_tokens_seen": 30084352,
      "step": 51830
    },
    {
      "epoch": 7.72043491212392,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.023124993839892492,
      "loss": 0.7978,
      "num_input_tokens_seen": 30087296,
      "step": 51835
    },
    {
      "epoch": 7.721179624664879,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02312335490546059,
      "loss": 0.8062,
      "num_input_tokens_seen": 30090304,
      "step": 51840
    },
    {
      "epoch": 7.721924337205839,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02312171583379286,
      "loss": 0.8035,
      "num_input_tokens_seen": 30092928,
      "step": 51845
    },
    {
      "epoch": 7.722669049746798,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.023120076624916996,
      "loss": 0.8111,
      "num_input_tokens_seen": 30096032,
      "step": 51850
    },
    {
      "epoch": 7.723413762287757,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02311843727886069,
      "loss": 0.8334,
      "num_input_tokens_seen": 30098592,
      "step": 51855
    },
    {
      "epoch": 7.724158474828716,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.023116797795651636,
      "loss": 0.8009,
      "num_input_tokens_seen": 30101632,
      "step": 51860
    },
    {
      "epoch": 7.724903187369676,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.023115158175317535,
      "loss": 0.8014,
      "num_input_tokens_seen": 30104608,
      "step": 51865
    },
    {
      "epoch": 7.725647899910634,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.023113518417886083,
      "loss": 0.8006,
      "num_input_tokens_seen": 30107200,
      "step": 51870
    },
    {
      "epoch": 7.726392612451594,
      "grad_norm": 0.017578125,
      "learning_rate": 0.023111878523384984,
      "loss": 0.7984,
      "num_input_tokens_seen": 30110112,
      "step": 51875
    },
    {
      "epoch": 7.727137324992553,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02311023849184194,
      "loss": 0.8035,
      "num_input_tokens_seen": 30112928,
      "step": 51880
    },
    {
      "epoch": 7.727882037533512,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.023108598323284664,
      "loss": 0.7826,
      "num_input_tokens_seen": 30116096,
      "step": 51885
    },
    {
      "epoch": 7.728626750074471,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.023106958017740858,
      "loss": 0.8138,
      "num_input_tokens_seen": 30119552,
      "step": 51890
    },
    {
      "epoch": 7.729371462615431,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.023105317575238232,
      "loss": 0.7994,
      "num_input_tokens_seen": 30122336,
      "step": 51895
    },
    {
      "epoch": 7.73011617515639,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02310367699580451,
      "loss": 0.8094,
      "num_input_tokens_seen": 30125536,
      "step": 51900
    },
    {
      "epoch": 7.730860887697348,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.023102036279467397,
      "loss": 0.8004,
      "num_input_tokens_seen": 30128352,
      "step": 51905
    },
    {
      "epoch": 7.731605600238308,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02310039542625462,
      "loss": 0.7869,
      "num_input_tokens_seen": 30131072,
      "step": 51910
    },
    {
      "epoch": 7.732350312779268,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02309875443619389,
      "loss": 0.7908,
      "num_input_tokens_seen": 30133856,
      "step": 51915
    },
    {
      "epoch": 7.733095025320226,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02309711330931294,
      "loss": 0.8134,
      "num_input_tokens_seen": 30136864,
      "step": 51920
    },
    {
      "epoch": 7.733839737861185,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.023095472045639484,
      "loss": 0.8066,
      "num_input_tokens_seen": 30139808,
      "step": 51925
    },
    {
      "epoch": 7.734584450402145,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.02309383064520126,
      "loss": 0.8019,
      "num_input_tokens_seen": 30142624,
      "step": 51930
    },
    {
      "epoch": 7.735329162943104,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.023092189108025993,
      "loss": 0.8028,
      "num_input_tokens_seen": 30145472,
      "step": 51935
    },
    {
      "epoch": 7.736073875484063,
      "grad_norm": 0.037109375,
      "learning_rate": 0.023090547434141413,
      "loss": 0.817,
      "num_input_tokens_seen": 30148032,
      "step": 51940
    },
    {
      "epoch": 7.736818588025022,
      "grad_norm": 0.01544189453125,
      "learning_rate": 0.02308890562357526,
      "loss": 0.7976,
      "num_input_tokens_seen": 30150848,
      "step": 51945
    },
    {
      "epoch": 7.737563300565982,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02308726367635527,
      "loss": 0.8063,
      "num_input_tokens_seen": 30153728,
      "step": 51950
    },
    {
      "epoch": 7.73830801310694,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.023085621592509174,
      "loss": 0.7879,
      "num_input_tokens_seen": 30156512,
      "step": 51955
    },
    {
      "epoch": 7.7390527256479,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.02308397937206472,
      "loss": 0.8068,
      "num_input_tokens_seen": 30159360,
      "step": 51960
    },
    {
      "epoch": 7.739797438188859,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.023082337015049654,
      "loss": 0.8175,
      "num_input_tokens_seen": 30162208,
      "step": 51965
    },
    {
      "epoch": 7.740542150729818,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.023080694521491713,
      "loss": 0.8123,
      "num_input_tokens_seen": 30165216,
      "step": 51970
    },
    {
      "epoch": 7.741286863270777,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.023079051891418657,
      "loss": 0.7945,
      "num_input_tokens_seen": 30168256,
      "step": 51975
    },
    {
      "epoch": 7.742031575811737,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.023077409124858235,
      "loss": 0.8003,
      "num_input_tokens_seen": 30171104,
      "step": 51980
    },
    {
      "epoch": 7.742776288352696,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.023075766221838188,
      "loss": 0.8023,
      "num_input_tokens_seen": 30174112,
      "step": 51985
    },
    {
      "epoch": 7.743521000893655,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.023074123182386275,
      "loss": 0.7979,
      "num_input_tokens_seen": 30176928,
      "step": 51990
    },
    {
      "epoch": 7.744265713434614,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02307248000653026,
      "loss": 0.8126,
      "num_input_tokens_seen": 30179872,
      "step": 51995
    },
    {
      "epoch": 7.745010425975574,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.0230708366942979,
      "loss": 0.8123,
      "num_input_tokens_seen": 30182912,
      "step": 52000
    },
    {
      "epoch": 7.745755138516532,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.023069193245716962,
      "loss": 0.8017,
      "num_input_tokens_seen": 30186272,
      "step": 52005
    },
    {
      "epoch": 7.746499851057492,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.0230675496608152,
      "loss": 0.8245,
      "num_input_tokens_seen": 30189152,
      "step": 52010
    },
    {
      "epoch": 7.747244563598451,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02306590593962039,
      "loss": 0.7928,
      "num_input_tokens_seen": 30192128,
      "step": 52015
    },
    {
      "epoch": 7.7479892761394105,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02306426208216029,
      "loss": 0.7995,
      "num_input_tokens_seen": 30194848,
      "step": 52020
    },
    {
      "epoch": 7.748733988680369,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02306261808846268,
      "loss": 0.7929,
      "num_input_tokens_seen": 30197760,
      "step": 52025
    },
    {
      "epoch": 7.749478701221329,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.023060973958555338,
      "loss": 0.7959,
      "num_input_tokens_seen": 30200704,
      "step": 52030
    },
    {
      "epoch": 7.750223413762288,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02305932969246603,
      "loss": 0.8069,
      "num_input_tokens_seen": 30203904,
      "step": 52035
    },
    {
      "epoch": 7.750968126303247,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02305768529022254,
      "loss": 0.7856,
      "num_input_tokens_seen": 30207104,
      "step": 52040
    },
    {
      "epoch": 7.751712838844206,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.023056040751852643,
      "loss": 0.7975,
      "num_input_tokens_seen": 30209888,
      "step": 52045
    },
    {
      "epoch": 7.752457551385166,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.023054396077384125,
      "loss": 0.8193,
      "num_input_tokens_seen": 30212672,
      "step": 52050
    },
    {
      "epoch": 7.753202263926124,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.023052751266844775,
      "loss": 0.8013,
      "num_input_tokens_seen": 30215616,
      "step": 52055
    },
    {
      "epoch": 7.753946976467084,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.023051106320262374,
      "loss": 0.7935,
      "num_input_tokens_seen": 30219456,
      "step": 52060
    },
    {
      "epoch": 7.754691689008043,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.023049461237664715,
      "loss": 0.7978,
      "num_input_tokens_seen": 30222656,
      "step": 52065
    },
    {
      "epoch": 7.755436401549002,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.02304781601907959,
      "loss": 0.8234,
      "num_input_tokens_seen": 30225504,
      "step": 52070
    },
    {
      "epoch": 7.756181114089961,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.023046170664534796,
      "loss": 0.8154,
      "num_input_tokens_seen": 30228256,
      "step": 52075
    },
    {
      "epoch": 7.756925826630921,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.023044525174058122,
      "loss": 0.7985,
      "num_input_tokens_seen": 30231200,
      "step": 52080
    },
    {
      "epoch": 7.75767053917188,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02304287954767737,
      "loss": 0.7943,
      "num_input_tokens_seen": 30234272,
      "step": 52085
    },
    {
      "epoch": 7.758415251712838,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.023041233785420345,
      "loss": 0.8041,
      "num_input_tokens_seen": 30236992,
      "step": 52090
    },
    {
      "epoch": 7.759159964253798,
      "grad_norm": 0.46875,
      "learning_rate": 0.02303958788731485,
      "loss": 0.7793,
      "num_input_tokens_seen": 30240064,
      "step": 52095
    },
    {
      "epoch": 7.759904676794758,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.02303794185338869,
      "loss": 0.7804,
      "num_input_tokens_seen": 30243424,
      "step": 52100
    },
    {
      "epoch": 7.7606493893357165,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02303629568366966,
      "loss": 0.7959,
      "num_input_tokens_seen": 30246048,
      "step": 52105
    },
    {
      "epoch": 7.761394101876675,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.023034649378185598,
      "loss": 0.8089,
      "num_input_tokens_seen": 30248832,
      "step": 52110
    },
    {
      "epoch": 7.762138814417635,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02303300293696429,
      "loss": 0.8183,
      "num_input_tokens_seen": 30251552,
      "step": 52115
    },
    {
      "epoch": 7.762883526958594,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.02303135636003356,
      "loss": 0.8257,
      "num_input_tokens_seen": 30254304,
      "step": 52120
    },
    {
      "epoch": 7.763628239499553,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.023029709647421234,
      "loss": 0.8107,
      "num_input_tokens_seen": 30257312,
      "step": 52125
    },
    {
      "epoch": 7.764372952040512,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.023028062799155122,
      "loss": 0.8115,
      "num_input_tokens_seen": 30260160,
      "step": 52130
    },
    {
      "epoch": 7.765117664581472,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.023026415815263053,
      "loss": 0.8286,
      "num_input_tokens_seen": 30263200,
      "step": 52135
    },
    {
      "epoch": 7.76586237712243,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.023024768695772843,
      "loss": 0.7944,
      "num_input_tokens_seen": 30266048,
      "step": 52140
    },
    {
      "epoch": 7.76660708966339,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02302312144071232,
      "loss": 0.8002,
      "num_input_tokens_seen": 30269024,
      "step": 52145
    },
    {
      "epoch": 7.767351802204349,
      "grad_norm": 0.03515625,
      "learning_rate": 0.023021474050109316,
      "loss": 0.8043,
      "num_input_tokens_seen": 30271904,
      "step": 52150
    },
    {
      "epoch": 7.7680965147453085,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02301982652399166,
      "loss": 0.7745,
      "num_input_tokens_seen": 30274752,
      "step": 52155
    },
    {
      "epoch": 7.768841227286267,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.023018178862387186,
      "loss": 0.8124,
      "num_input_tokens_seen": 30277696,
      "step": 52160
    },
    {
      "epoch": 7.769585939827227,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02301653106532373,
      "loss": 0.8295,
      "num_input_tokens_seen": 30280832,
      "step": 52165
    },
    {
      "epoch": 7.770330652368186,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02301488313282913,
      "loss": 0.7899,
      "num_input_tokens_seen": 30283872,
      "step": 52170
    },
    {
      "epoch": 7.771075364909145,
      "grad_norm": 0.01171875,
      "learning_rate": 0.023013235064931224,
      "loss": 0.8286,
      "num_input_tokens_seen": 30286944,
      "step": 52175
    },
    {
      "epoch": 7.771820077450104,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02301158686165786,
      "loss": 0.8054,
      "num_input_tokens_seen": 30289856,
      "step": 52180
    },
    {
      "epoch": 7.772564789991064,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02300993852303687,
      "loss": 0.8166,
      "num_input_tokens_seen": 30292736,
      "step": 52185
    },
    {
      "epoch": 7.7733095025320225,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.023008290049096117,
      "loss": 0.8119,
      "num_input_tokens_seen": 30295520,
      "step": 52190
    },
    {
      "epoch": 7.774054215072982,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.023006641439863442,
      "loss": 0.7912,
      "num_input_tokens_seen": 30298976,
      "step": 52195
    },
    {
      "epoch": 7.774798927613941,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.023004992695366693,
      "loss": 0.8113,
      "num_input_tokens_seen": 30301792,
      "step": 52200
    },
    {
      "epoch": 7.7755436401549005,
      "grad_norm": 0.0135498046875,
      "learning_rate": 0.023003343815633733,
      "loss": 0.802,
      "num_input_tokens_seen": 30304416,
      "step": 52205
    },
    {
      "epoch": 7.776288352695859,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02300169480069241,
      "loss": 0.8023,
      "num_input_tokens_seen": 30307552,
      "step": 52210
    },
    {
      "epoch": 7.777033065236819,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02300004565057059,
      "loss": 0.8089,
      "num_input_tokens_seen": 30310560,
      "step": 52215
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.022998396365296127,
      "loss": 0.823,
      "num_input_tokens_seen": 30313472,
      "step": 52220
    },
    {
      "epoch": 7.778522490318737,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.022996746944896885,
      "loss": 0.8019,
      "num_input_tokens_seen": 30316480,
      "step": 52225
    },
    {
      "epoch": 7.779267202859696,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.022995097389400736,
      "loss": 0.8095,
      "num_input_tokens_seen": 30319776,
      "step": 52230
    },
    {
      "epoch": 7.780011915400656,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.022993447698835535,
      "loss": 0.8113,
      "num_input_tokens_seen": 30322752,
      "step": 52235
    },
    {
      "epoch": 7.7807566279416145,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.022991797873229164,
      "loss": 0.7991,
      "num_input_tokens_seen": 30325472,
      "step": 52240
    },
    {
      "epoch": 7.781501340482574,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.022990147912609494,
      "loss": 0.8045,
      "num_input_tokens_seen": 30328320,
      "step": 52245
    },
    {
      "epoch": 7.782246053023533,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.022988497817004388,
      "loss": 0.8085,
      "num_input_tokens_seen": 30331168,
      "step": 52250
    },
    {
      "epoch": 7.782990765564492,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.022986847586441737,
      "loss": 0.8017,
      "num_input_tokens_seen": 30333856,
      "step": 52255
    },
    {
      "epoch": 7.783735478105451,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02298519722094941,
      "loss": 0.7811,
      "num_input_tokens_seen": 30336640,
      "step": 52260
    },
    {
      "epoch": 7.784480190646411,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.022983546720555293,
      "loss": 0.784,
      "num_input_tokens_seen": 30339968,
      "step": 52265
    },
    {
      "epoch": 7.78522490318737,
      "grad_norm": 0.01953125,
      "learning_rate": 0.022981896085287265,
      "loss": 0.8017,
      "num_input_tokens_seen": 30342688,
      "step": 52270
    },
    {
      "epoch": 7.7859696157283285,
      "grad_norm": 0.01544189453125,
      "learning_rate": 0.02298024531517322,
      "loss": 0.7751,
      "num_input_tokens_seen": 30345856,
      "step": 52275
    },
    {
      "epoch": 7.786714328269288,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.022978594410241037,
      "loss": 0.7995,
      "num_input_tokens_seen": 30348896,
      "step": 52280
    },
    {
      "epoch": 7.787459040810247,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.022976943370518615,
      "loss": 0.7966,
      "num_input_tokens_seen": 30351872,
      "step": 52285
    },
    {
      "epoch": 7.7882037533512065,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.022975292196033836,
      "loss": 0.7827,
      "num_input_tokens_seen": 30354784,
      "step": 52290
    },
    {
      "epoch": 7.788948465892165,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.022973640886814608,
      "loss": 0.7974,
      "num_input_tokens_seen": 30357600,
      "step": 52295
    },
    {
      "epoch": 7.789693178433125,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.022971989442888818,
      "loss": 0.8035,
      "num_input_tokens_seen": 30360384,
      "step": 52300
    },
    {
      "epoch": 7.790437890974084,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.022970337864284367,
      "loss": 0.7926,
      "num_input_tokens_seen": 30363008,
      "step": 52305
    },
    {
      "epoch": 7.791182603515043,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.022968686151029163,
      "loss": 0.7975,
      "num_input_tokens_seen": 30366016,
      "step": 52310
    },
    {
      "epoch": 7.791927316056002,
      "grad_norm": 0.021484375,
      "learning_rate": 0.022967034303151102,
      "loss": 0.787,
      "num_input_tokens_seen": 30368928,
      "step": 52315
    },
    {
      "epoch": 7.792672028596962,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.022965382320678095,
      "loss": 0.7823,
      "num_input_tokens_seen": 30372224,
      "step": 52320
    },
    {
      "epoch": 7.7934167411379205,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.022963730203638046,
      "loss": 0.8224,
      "num_input_tokens_seen": 30374880,
      "step": 52325
    },
    {
      "epoch": 7.79416145367888,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.022962077952058874,
      "loss": 0.7983,
      "num_input_tokens_seen": 30377824,
      "step": 52330
    },
    {
      "epoch": 7.794906166219839,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.022960425565968487,
      "loss": 0.7831,
      "num_input_tokens_seen": 30380896,
      "step": 52335
    },
    {
      "epoch": 7.7956508787607985,
      "grad_norm": 0.0234375,
      "learning_rate": 0.0229587730453948,
      "loss": 0.8135,
      "num_input_tokens_seen": 30383616,
      "step": 52340
    },
    {
      "epoch": 7.796395591301757,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02295712039036573,
      "loss": 0.8284,
      "num_input_tokens_seen": 30386624,
      "step": 52345
    },
    {
      "epoch": 7.797140303842717,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.022955467600909194,
      "loss": 0.7862,
      "num_input_tokens_seen": 30389728,
      "step": 52350
    },
    {
      "epoch": 7.797885016383676,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02295381467705313,
      "loss": 0.8092,
      "num_input_tokens_seen": 30392608,
      "step": 52355
    },
    {
      "epoch": 7.798629728924635,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.022952161618825438,
      "loss": 0.8059,
      "num_input_tokens_seen": 30395712,
      "step": 52360
    },
    {
      "epoch": 7.799374441465594,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.022950508426254065,
      "loss": 0.784,
      "num_input_tokens_seen": 30398560,
      "step": 52365
    },
    {
      "epoch": 7.800119154006554,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.022948855099366932,
      "loss": 0.7691,
      "num_input_tokens_seen": 30401280,
      "step": 52370
    },
    {
      "epoch": 7.8008638665475125,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02294720163819197,
      "loss": 0.7903,
      "num_input_tokens_seen": 30403840,
      "step": 52375
    },
    {
      "epoch": 7.801608579088472,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.022945548042757117,
      "loss": 0.7863,
      "num_input_tokens_seen": 30406848,
      "step": 52380
    },
    {
      "epoch": 7.802353291629431,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.0229438943130903,
      "loss": 0.7978,
      "num_input_tokens_seen": 30409600,
      "step": 52385
    },
    {
      "epoch": 7.803098004170391,
      "grad_norm": 0.017578125,
      "learning_rate": 0.022942240449219466,
      "loss": 0.8065,
      "num_input_tokens_seen": 30412416,
      "step": 52390
    },
    {
      "epoch": 7.803842716711349,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02294058645117255,
      "loss": 0.7912,
      "num_input_tokens_seen": 30415200,
      "step": 52395
    },
    {
      "epoch": 7.804587429252309,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.022938932318977492,
      "loss": 0.8125,
      "num_input_tokens_seen": 30418080,
      "step": 52400
    },
    {
      "epoch": 7.805332141793268,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.022937278052662248,
      "loss": 0.8,
      "num_input_tokens_seen": 30420896,
      "step": 52405
    },
    {
      "epoch": 7.806076854334227,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.022935623652254756,
      "loss": 0.8135,
      "num_input_tokens_seen": 30423904,
      "step": 52410
    },
    {
      "epoch": 7.806821566875186,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02293396911778297,
      "loss": 0.8067,
      "num_input_tokens_seen": 30426720,
      "step": 52415
    },
    {
      "epoch": 7.807566279416145,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.02293231444927484,
      "loss": 0.785,
      "num_input_tokens_seen": 30429504,
      "step": 52420
    },
    {
      "epoch": 7.8083109919571045,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.022930659646758316,
      "loss": 0.796,
      "num_input_tokens_seen": 30432640,
      "step": 52425
    },
    {
      "epoch": 7.809055704498064,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.022929004710261356,
      "loss": 0.7857,
      "num_input_tokens_seen": 30435520,
      "step": 52430
    },
    {
      "epoch": 7.809800417039023,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.022927349639811922,
      "loss": 0.803,
      "num_input_tokens_seen": 30438272,
      "step": 52435
    },
    {
      "epoch": 7.810545129579982,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.022925694435437977,
      "loss": 0.8008,
      "num_input_tokens_seen": 30441312,
      "step": 52440
    },
    {
      "epoch": 7.811289842120941,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.022924039097167476,
      "loss": 0.8128,
      "num_input_tokens_seen": 30444384,
      "step": 52445
    },
    {
      "epoch": 7.812034554661901,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.022922383625028384,
      "loss": 0.7844,
      "num_input_tokens_seen": 30447296,
      "step": 52450
    },
    {
      "epoch": 7.81277926720286,
      "grad_norm": 0.02734375,
      "learning_rate": 0.022920728019048677,
      "loss": 0.7839,
      "num_input_tokens_seen": 30450624,
      "step": 52455
    },
    {
      "epoch": 7.8135239797438185,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02291907227925632,
      "loss": 0.7739,
      "num_input_tokens_seen": 30453248,
      "step": 52460
    },
    {
      "epoch": 7.814268692284778,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02291741640567929,
      "loss": 0.7574,
      "num_input_tokens_seen": 30456096,
      "step": 52465
    },
    {
      "epoch": 7.815013404825737,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02291576039834555,
      "loss": 0.8414,
      "num_input_tokens_seen": 30459008,
      "step": 52470
    },
    {
      "epoch": 7.815758117366697,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02291410425728309,
      "loss": 0.7804,
      "num_input_tokens_seen": 30461824,
      "step": 52475
    },
    {
      "epoch": 7.816502829907655,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.022912447982519873,
      "loss": 0.7818,
      "num_input_tokens_seen": 30464704,
      "step": 52480
    },
    {
      "epoch": 7.817247542448615,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.022910791574083895,
      "loss": 0.8036,
      "num_input_tokens_seen": 30467584,
      "step": 52485
    },
    {
      "epoch": 7.817992254989574,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.022909135032003133,
      "loss": 0.8203,
      "num_input_tokens_seen": 30470560,
      "step": 52490
    },
    {
      "epoch": 7.818736967530533,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.022907478356305574,
      "loss": 0.775,
      "num_input_tokens_seen": 30473568,
      "step": 52495
    },
    {
      "epoch": 7.819481680071492,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.0229058215470192,
      "loss": 0.7684,
      "num_input_tokens_seen": 30476384,
      "step": 52500
    },
    {
      "epoch": 7.820226392612452,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.022904164604172016,
      "loss": 0.827,
      "num_input_tokens_seen": 30479360,
      "step": 52505
    },
    {
      "epoch": 7.8209711051534105,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.022902507527791997,
      "loss": 0.8111,
      "num_input_tokens_seen": 30482400,
      "step": 52510
    },
    {
      "epoch": 7.82171581769437,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02290085031790715,
      "loss": 0.7772,
      "num_input_tokens_seen": 30485344,
      "step": 52515
    },
    {
      "epoch": 7.822460530235329,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.022899192974545457,
      "loss": 0.8095,
      "num_input_tokens_seen": 30488384,
      "step": 52520
    },
    {
      "epoch": 7.823205242776289,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.022897535497734933,
      "loss": 0.7898,
      "num_input_tokens_seen": 30491424,
      "step": 52525
    },
    {
      "epoch": 7.823949955317247,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.022895877887503573,
      "loss": 0.7552,
      "num_input_tokens_seen": 30494304,
      "step": 52530
    },
    {
      "epoch": 7.824694667858207,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.022894220143879384,
      "loss": 0.7861,
      "num_input_tokens_seen": 30497312,
      "step": 52535
    },
    {
      "epoch": 7.825439380399166,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02289256226689037,
      "loss": 0.7711,
      "num_input_tokens_seen": 30500160,
      "step": 52540
    },
    {
      "epoch": 7.826184092940125,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.022890904256564537,
      "loss": 0.8249,
      "num_input_tokens_seen": 30502944,
      "step": 52545
    },
    {
      "epoch": 7.826928805481084,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.022889246112929893,
      "loss": 0.8088,
      "num_input_tokens_seen": 30505824,
      "step": 52550
    },
    {
      "epoch": 7.827673518022044,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.022887587836014456,
      "loss": 0.7998,
      "num_input_tokens_seen": 30508832,
      "step": 52555
    },
    {
      "epoch": 7.828418230563003,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.022885929425846236,
      "loss": 0.8014,
      "num_input_tokens_seen": 30511776,
      "step": 52560
    },
    {
      "epoch": 7.829162943103962,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.022884270882453255,
      "loss": 0.7989,
      "num_input_tokens_seen": 30514592,
      "step": 52565
    },
    {
      "epoch": 7.829907655644921,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.022882612205863535,
      "loss": 0.7747,
      "num_input_tokens_seen": 30517504,
      "step": 52570
    },
    {
      "epoch": 7.830652368185881,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02288095339610509,
      "loss": 0.82,
      "num_input_tokens_seen": 30520096,
      "step": 52575
    },
    {
      "epoch": 7.831397080726839,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02287929445320595,
      "loss": 0.7811,
      "num_input_tokens_seen": 30523264,
      "step": 52580
    },
    {
      "epoch": 7.832141793267798,
      "grad_norm": 0.017578125,
      "learning_rate": 0.022877635377194137,
      "loss": 0.7777,
      "num_input_tokens_seen": 30526048,
      "step": 52585
    },
    {
      "epoch": 7.832886505808758,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.022875976168097674,
      "loss": 0.8235,
      "num_input_tokens_seen": 30529088,
      "step": 52590
    },
    {
      "epoch": 7.833631218349717,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.022874316825944603,
      "loss": 0.7733,
      "num_input_tokens_seen": 30531936,
      "step": 52595
    },
    {
      "epoch": 7.834375930890676,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.022872657350762955,
      "loss": 0.8009,
      "num_input_tokens_seen": 30534912,
      "step": 52600
    },
    {
      "epoch": 7.835120643431635,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02287099774258076,
      "loss": 0.8553,
      "num_input_tokens_seen": 30538944,
      "step": 52605
    },
    {
      "epoch": 7.835865355972595,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02286933800142606,
      "loss": 0.7957,
      "num_input_tokens_seen": 30541888,
      "step": 52610
    },
    {
      "epoch": 7.836610068513554,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02286767812732689,
      "loss": 0.8074,
      "num_input_tokens_seen": 30544576,
      "step": 52615
    },
    {
      "epoch": 7.837354781054513,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.022866018120311297,
      "loss": 0.7749,
      "num_input_tokens_seen": 30547360,
      "step": 52620
    },
    {
      "epoch": 7.838099493595472,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.022864357980407323,
      "loss": 0.8331,
      "num_input_tokens_seen": 30550816,
      "step": 52625
    },
    {
      "epoch": 7.838844206136431,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.022862697707643008,
      "loss": 0.8259,
      "num_input_tokens_seen": 30553536,
      "step": 52630
    },
    {
      "epoch": 7.83958891867739,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02286103730204641,
      "loss": 0.8143,
      "num_input_tokens_seen": 30556032,
      "step": 52635
    },
    {
      "epoch": 7.84033363121835,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.02285937676364558,
      "loss": 0.8142,
      "num_input_tokens_seen": 30558784,
      "step": 52640
    },
    {
      "epoch": 7.841078343759309,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02285771609246856,
      "loss": 0.8204,
      "num_input_tokens_seen": 30561920,
      "step": 52645
    },
    {
      "epoch": 7.841823056300268,
      "grad_norm": 0.01953125,
      "learning_rate": 0.02285605528854342,
      "loss": 0.8011,
      "num_input_tokens_seen": 30564800,
      "step": 52650
    },
    {
      "epoch": 7.842567768841227,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.022854394351898205,
      "loss": 0.796,
      "num_input_tokens_seen": 30567968,
      "step": 52655
    },
    {
      "epoch": 7.843312481382187,
      "grad_norm": 0.029296875,
      "learning_rate": 0.022852733282560985,
      "loss": 0.7698,
      "num_input_tokens_seen": 30570592,
      "step": 52660
    },
    {
      "epoch": 7.844057193923145,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.022851072080559816,
      "loss": 0.812,
      "num_input_tokens_seen": 30573440,
      "step": 52665
    },
    {
      "epoch": 7.844801906464105,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02284941074592276,
      "loss": 0.7763,
      "num_input_tokens_seen": 30576128,
      "step": 52670
    },
    {
      "epoch": 7.845546619005064,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.022847749278677894,
      "loss": 0.8267,
      "num_input_tokens_seen": 30579072,
      "step": 52675
    },
    {
      "epoch": 7.846291331546023,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.022846087678853277,
      "loss": 0.816,
      "num_input_tokens_seen": 30582624,
      "step": 52680
    },
    {
      "epoch": 7.847036044086982,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.022844425946476978,
      "loss": 0.7793,
      "num_input_tokens_seen": 30585344,
      "step": 52685
    },
    {
      "epoch": 7.847780756627942,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02284276408157708,
      "loss": 0.7982,
      "num_input_tokens_seen": 30587968,
      "step": 52690
    },
    {
      "epoch": 7.848525469168901,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.022841102084181653,
      "loss": 0.8185,
      "num_input_tokens_seen": 30590976,
      "step": 52695
    },
    {
      "epoch": 7.84927018170986,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.022839439954318776,
      "loss": 0.8189,
      "num_input_tokens_seen": 30593888,
      "step": 52700
    },
    {
      "epoch": 7.850014894250819,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.022837777692016534,
      "loss": 0.7797,
      "num_input_tokens_seen": 30596480,
      "step": 52705
    },
    {
      "epoch": 7.850759606791779,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.022836115297303,
      "loss": 0.8072,
      "num_input_tokens_seen": 30599744,
      "step": 52710
    },
    {
      "epoch": 7.851504319332737,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.022834452770206258,
      "loss": 0.7833,
      "num_input_tokens_seen": 30602656,
      "step": 52715
    },
    {
      "epoch": 7.852249031873697,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.022832790110754397,
      "loss": 0.7872,
      "num_input_tokens_seen": 30605280,
      "step": 52720
    },
    {
      "epoch": 7.852993744414656,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.022831127318975516,
      "loss": 0.7995,
      "num_input_tokens_seen": 30608608,
      "step": 52725
    },
    {
      "epoch": 7.8537384569556155,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02282946439489769,
      "loss": 0.8197,
      "num_input_tokens_seen": 30612096,
      "step": 52730
    },
    {
      "epoch": 7.854483169496574,
      "grad_norm": 0.0125732421875,
      "learning_rate": 0.022827801338549034,
      "loss": 0.7893,
      "num_input_tokens_seen": 30614848,
      "step": 52735
    },
    {
      "epoch": 7.855227882037534,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.022826138149957622,
      "loss": 0.7906,
      "num_input_tokens_seen": 30617632,
      "step": 52740
    },
    {
      "epoch": 7.855972594578493,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.022824474829151557,
      "loss": 0.8177,
      "num_input_tokens_seen": 30620480,
      "step": 52745
    },
    {
      "epoch": 7.856717307119452,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.022822811376158943,
      "loss": 0.7845,
      "num_input_tokens_seen": 30623392,
      "step": 52750
    },
    {
      "epoch": 7.857462019660411,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.022821147791007884,
      "loss": 0.7869,
      "num_input_tokens_seen": 30626144,
      "step": 52755
    },
    {
      "epoch": 7.858206732201371,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02281948407372648,
      "loss": 0.8352,
      "num_input_tokens_seen": 30628928,
      "step": 52760
    },
    {
      "epoch": 7.858951444742329,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.022817820224342842,
      "loss": 0.8156,
      "num_input_tokens_seen": 30632000,
      "step": 52765
    },
    {
      "epoch": 7.859696157283288,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02281615624288508,
      "loss": 0.8319,
      "num_input_tokens_seen": 30634944,
      "step": 52770
    },
    {
      "epoch": 7.860440869824248,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.022814492129381296,
      "loss": 0.8269,
      "num_input_tokens_seen": 30637472,
      "step": 52775
    },
    {
      "epoch": 7.8611855823652075,
      "grad_norm": 0.03125,
      "learning_rate": 0.022812827883859618,
      "loss": 0.7861,
      "num_input_tokens_seen": 30640512,
      "step": 52780
    },
    {
      "epoch": 7.861930294906166,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.022811163506348146,
      "loss": 0.7887,
      "num_input_tokens_seen": 30643264,
      "step": 52785
    },
    {
      "epoch": 7.862675007447125,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.022809498996875007,
      "loss": 0.802,
      "num_input_tokens_seen": 30646080,
      "step": 52790
    },
    {
      "epoch": 7.863419719988085,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02280783435546832,
      "loss": 0.7781,
      "num_input_tokens_seen": 30648928,
      "step": 52795
    },
    {
      "epoch": 7.864164432529043,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02280616958215621,
      "loss": 0.8111,
      "num_input_tokens_seen": 30652000,
      "step": 52800
    },
    {
      "epoch": 7.864909145070003,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.022804504676966795,
      "loss": 0.7946,
      "num_input_tokens_seen": 30654752,
      "step": 52805
    },
    {
      "epoch": 7.865653857610962,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.022802839639928206,
      "loss": 0.7985,
      "num_input_tokens_seen": 30657568,
      "step": 52810
    },
    {
      "epoch": 7.8663985701519215,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.022801174471068573,
      "loss": 0.7982,
      "num_input_tokens_seen": 30660288,
      "step": 52815
    },
    {
      "epoch": 7.86714328269288,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02279950917041603,
      "loss": 0.7971,
      "num_input_tokens_seen": 30663296,
      "step": 52820
    },
    {
      "epoch": 7.86788799523384,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.022797843737998702,
      "loss": 0.8458,
      "num_input_tokens_seen": 30666144,
      "step": 52825
    },
    {
      "epoch": 7.868632707774799,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.022796178173844724,
      "loss": 0.7943,
      "num_input_tokens_seen": 30669120,
      "step": 52830
    },
    {
      "epoch": 7.869377420315758,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.022794512477982243,
      "loss": 0.8023,
      "num_input_tokens_seen": 30672608,
      "step": 52835
    },
    {
      "epoch": 7.870122132856717,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.022792846650439402,
      "loss": 0.7754,
      "num_input_tokens_seen": 30675360,
      "step": 52840
    },
    {
      "epoch": 7.870866845397677,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.022791180691244334,
      "loss": 0.8006,
      "num_input_tokens_seen": 30678304,
      "step": 52845
    },
    {
      "epoch": 7.871611557938635,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.022789514600425186,
      "loss": 0.8079,
      "num_input_tokens_seen": 30681440,
      "step": 52850
    },
    {
      "epoch": 7.872356270479595,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.022787848378010105,
      "loss": 0.8006,
      "num_input_tokens_seen": 30684480,
      "step": 52855
    },
    {
      "epoch": 7.873100983020554,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02278618202402724,
      "loss": 0.8077,
      "num_input_tokens_seen": 30687456,
      "step": 52860
    },
    {
      "epoch": 7.8738456955615135,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.022784515538504746,
      "loss": 0.7944,
      "num_input_tokens_seen": 30690560,
      "step": 52865
    },
    {
      "epoch": 7.874590408102472,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.022782848921470776,
      "loss": 0.8273,
      "num_input_tokens_seen": 30693408,
      "step": 52870
    },
    {
      "epoch": 7.875335120643432,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02278118217295348,
      "loss": 0.8082,
      "num_input_tokens_seen": 30696352,
      "step": 52875
    },
    {
      "epoch": 7.876079833184391,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.022779515292981018,
      "loss": 0.8284,
      "num_input_tokens_seen": 30699456,
      "step": 52880
    },
    {
      "epoch": 7.87682454572535,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02277784828158155,
      "loss": 0.7799,
      "num_input_tokens_seen": 30702368,
      "step": 52885
    },
    {
      "epoch": 7.877569258266309,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02277618113878324,
      "loss": 0.786,
      "num_input_tokens_seen": 30705024,
      "step": 52890
    },
    {
      "epoch": 7.878313970807269,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.022774513864614262,
      "loss": 0.7987,
      "num_input_tokens_seen": 30707936,
      "step": 52895
    },
    {
      "epoch": 7.8790586833482275,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.02277284645910277,
      "loss": 0.814,
      "num_input_tokens_seen": 30710848,
      "step": 52900
    },
    {
      "epoch": 7.879803395889187,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02277117892227694,
      "loss": 0.8148,
      "num_input_tokens_seen": 30713664,
      "step": 52905
    },
    {
      "epoch": 7.880548108430146,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02276951125416493,
      "loss": 0.7962,
      "num_input_tokens_seen": 30716512,
      "step": 52910
    },
    {
      "epoch": 7.8812928209711055,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02276784345479493,
      "loss": 0.797,
      "num_input_tokens_seen": 30719520,
      "step": 52915
    },
    {
      "epoch": 7.882037533512064,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.022766175524195107,
      "loss": 0.819,
      "num_input_tokens_seen": 30722752,
      "step": 52920
    },
    {
      "epoch": 7.882782246053024,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02276450746239365,
      "loss": 0.8159,
      "num_input_tokens_seen": 30725824,
      "step": 52925
    },
    {
      "epoch": 7.883526958593983,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.022762839269418723,
      "loss": 0.7855,
      "num_input_tokens_seen": 30728672,
      "step": 52930
    },
    {
      "epoch": 7.884271671134941,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02276117094529852,
      "loss": 0.8175,
      "num_input_tokens_seen": 30731776,
      "step": 52935
    },
    {
      "epoch": 7.885016383675901,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.022759502490061225,
      "loss": 0.8074,
      "num_input_tokens_seen": 30734880,
      "step": 52940
    },
    {
      "epoch": 7.885761096216861,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.022757833903735016,
      "loss": 0.8163,
      "num_input_tokens_seen": 30737632,
      "step": 52945
    },
    {
      "epoch": 7.8865058087578195,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.022756165186348088,
      "loss": 0.7968,
      "num_input_tokens_seen": 30740256,
      "step": 52950
    },
    {
      "epoch": 7.887250521298778,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.022754496337928636,
      "loss": 0.8206,
      "num_input_tokens_seen": 30743008,
      "step": 52955
    },
    {
      "epoch": 7.887995233839738,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.022752827358504844,
      "loss": 0.8044,
      "num_input_tokens_seen": 30745664,
      "step": 52960
    },
    {
      "epoch": 7.8887399463806975,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.022751158248104916,
      "loss": 0.811,
      "num_input_tokens_seen": 30748416,
      "step": 52965
    },
    {
      "epoch": 7.889484658921656,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.022749489006757048,
      "loss": 0.8084,
      "num_input_tokens_seen": 30751776,
      "step": 52970
    },
    {
      "epoch": 7.890229371462615,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02274781963448944,
      "loss": 0.8034,
      "num_input_tokens_seen": 30754720,
      "step": 52975
    },
    {
      "epoch": 7.890974084003575,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.022746150131330294,
      "loss": 0.8073,
      "num_input_tokens_seen": 30757504,
      "step": 52980
    },
    {
      "epoch": 7.8917187965445335,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.022744480497307814,
      "loss": 0.8119,
      "num_input_tokens_seen": 30760416,
      "step": 52985
    },
    {
      "epoch": 7.892463509085493,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.022742810732450207,
      "loss": 0.8078,
      "num_input_tokens_seen": 30763264,
      "step": 52990
    },
    {
      "epoch": 7.893208221626452,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02274114083678568,
      "loss": 0.8156,
      "num_input_tokens_seen": 30765952,
      "step": 52995
    },
    {
      "epoch": 7.8939529341674115,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.022739470810342447,
      "loss": 0.8156,
      "num_input_tokens_seen": 30768864,
      "step": 53000
    },
    {
      "epoch": 7.89469764670837,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.022737800653148725,
      "loss": 0.795,
      "num_input_tokens_seen": 30771680,
      "step": 53005
    },
    {
      "epoch": 7.89544235924933,
      "grad_norm": 0.015625,
      "learning_rate": 0.022736130365232723,
      "loss": 0.8082,
      "num_input_tokens_seen": 30774720,
      "step": 53010
    },
    {
      "epoch": 7.896187071790289,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.02273445994662266,
      "loss": 0.8006,
      "num_input_tokens_seen": 30777696,
      "step": 53015
    },
    {
      "epoch": 7.896931784331248,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.022732789397346756,
      "loss": 0.7889,
      "num_input_tokens_seen": 30780448,
      "step": 53020
    },
    {
      "epoch": 7.897676496872207,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02273111871743324,
      "loss": 0.8206,
      "num_input_tokens_seen": 30783104,
      "step": 53025
    },
    {
      "epoch": 7.898421209413167,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.022729447906910324,
      "loss": 0.807,
      "num_input_tokens_seen": 30786176,
      "step": 53030
    },
    {
      "epoch": 7.8991659219541255,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.022727776965806246,
      "loss": 0.7968,
      "num_input_tokens_seen": 30788992,
      "step": 53035
    },
    {
      "epoch": 7.899910634495085,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02272610589414923,
      "loss": 0.8034,
      "num_input_tokens_seen": 30792000,
      "step": 53040
    },
    {
      "epoch": 7.900655347036044,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.022724434691967502,
      "loss": 0.818,
      "num_input_tokens_seen": 30794880,
      "step": 53045
    },
    {
      "epoch": 7.9014000595770035,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.022722763359289307,
      "loss": 0.7987,
      "num_input_tokens_seen": 30797856,
      "step": 53050
    },
    {
      "epoch": 7.902144772117962,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.02272109189614287,
      "loss": 0.7899,
      "num_input_tokens_seen": 30800608,
      "step": 53055
    },
    {
      "epoch": 7.902889484658922,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.022719420302556435,
      "loss": 0.8121,
      "num_input_tokens_seen": 30803616,
      "step": 53060
    },
    {
      "epoch": 7.903634197199881,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.022717748578558242,
      "loss": 0.7982,
      "num_input_tokens_seen": 30806528,
      "step": 53065
    },
    {
      "epoch": 7.90437890974084,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.022716076724176526,
      "loss": 0.8096,
      "num_input_tokens_seen": 30809344,
      "step": 53070
    },
    {
      "epoch": 7.905123622281799,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.022714404739439538,
      "loss": 0.811,
      "num_input_tokens_seen": 30812224,
      "step": 53075
    },
    {
      "epoch": 7.905868334822759,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.022712732624375523,
      "loss": 0.822,
      "num_input_tokens_seen": 30815296,
      "step": 53080
    },
    {
      "epoch": 7.9066130473637175,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02271106037901273,
      "loss": 0.7898,
      "num_input_tokens_seen": 30818240,
      "step": 53085
    },
    {
      "epoch": 7.907357759904677,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.022709388003379405,
      "loss": 0.8233,
      "num_input_tokens_seen": 30821216,
      "step": 53090
    },
    {
      "epoch": 7.908102472445636,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02270771549750381,
      "loss": 0.7794,
      "num_input_tokens_seen": 30823872,
      "step": 53095
    },
    {
      "epoch": 7.908847184986596,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.022706042861414193,
      "loss": 0.7972,
      "num_input_tokens_seen": 30826720,
      "step": 53100
    },
    {
      "epoch": 7.909591897527554,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02270437009513882,
      "loss": 0.7943,
      "num_input_tokens_seen": 30829536,
      "step": 53105
    },
    {
      "epoch": 7.910336610068514,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.022702697198705935,
      "loss": 0.8157,
      "num_input_tokens_seen": 30832352,
      "step": 53110
    },
    {
      "epoch": 7.911081322609473,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02270102417214381,
      "loss": 0.7909,
      "num_input_tokens_seen": 30835296,
      "step": 53115
    },
    {
      "epoch": 7.9118260351504315,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.022699351015480713,
      "loss": 0.7952,
      "num_input_tokens_seen": 30838240,
      "step": 53120
    },
    {
      "epoch": 7.912570747691391,
      "grad_norm": 0.021484375,
      "learning_rate": 0.0226976777287449,
      "loss": 0.808,
      "num_input_tokens_seen": 30841120,
      "step": 53125
    },
    {
      "epoch": 7.913315460232351,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.022696004311964654,
      "loss": 0.8162,
      "num_input_tokens_seen": 30843904,
      "step": 53130
    },
    {
      "epoch": 7.9140601727733095,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.022694330765168233,
      "loss": 0.812,
      "num_input_tokens_seen": 30846816,
      "step": 53135
    },
    {
      "epoch": 7.914804885314268,
      "grad_norm": 0.0234375,
      "learning_rate": 0.022692657088383916,
      "loss": 0.8062,
      "num_input_tokens_seen": 30849696,
      "step": 53140
    },
    {
      "epoch": 7.915549597855228,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.022690983281639977,
      "loss": 0.8184,
      "num_input_tokens_seen": 30852480,
      "step": 53145
    },
    {
      "epoch": 7.916294310396187,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.02268930934496469,
      "loss": 0.8043,
      "num_input_tokens_seen": 30855264,
      "step": 53150
    },
    {
      "epoch": 7.917039022937146,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.022687635278386332,
      "loss": 0.7912,
      "num_input_tokens_seen": 30857888,
      "step": 53155
    },
    {
      "epoch": 7.917783735478105,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.022685961081933194,
      "loss": 0.818,
      "num_input_tokens_seen": 30860832,
      "step": 53160
    },
    {
      "epoch": 7.918528448019065,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02268428675563355,
      "loss": 0.813,
      "num_input_tokens_seen": 30863584,
      "step": 53165
    },
    {
      "epoch": 7.9192731605600235,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.022682612299515702,
      "loss": 0.7896,
      "num_input_tokens_seen": 30866720,
      "step": 53170
    },
    {
      "epoch": 7.920017873100983,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.022680937713607918,
      "loss": 0.8246,
      "num_input_tokens_seen": 30869600,
      "step": 53175
    },
    {
      "epoch": 7.920762585641942,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.0226792629979385,
      "loss": 0.8027,
      "num_input_tokens_seen": 30872480,
      "step": 53180
    },
    {
      "epoch": 7.921507298182902,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.02267758815253574,
      "loss": 0.7984,
      "num_input_tokens_seen": 30875168,
      "step": 53185
    },
    {
      "epoch": 7.92225201072386,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.022675913177427936,
      "loss": 0.8036,
      "num_input_tokens_seen": 30878176,
      "step": 53190
    },
    {
      "epoch": 7.92299672326482,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.022674238072643373,
      "loss": 0.8013,
      "num_input_tokens_seen": 30881408,
      "step": 53195
    },
    {
      "epoch": 7.923741435805779,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02267256283821036,
      "loss": 0.8295,
      "num_input_tokens_seen": 30884224,
      "step": 53200
    },
    {
      "epoch": 7.924486148346738,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.022670887474157197,
      "loss": 0.7992,
      "num_input_tokens_seen": 30887072,
      "step": 53205
    },
    {
      "epoch": 7.925230860887697,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02266921198051218,
      "loss": 0.7947,
      "num_input_tokens_seen": 30889824,
      "step": 53210
    },
    {
      "epoch": 7.925975573428657,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.022667536357303626,
      "loss": 0.8045,
      "num_input_tokens_seen": 30892736,
      "step": 53215
    },
    {
      "epoch": 7.9267202859696155,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.022665860604559836,
      "loss": 0.8094,
      "num_input_tokens_seen": 30895744,
      "step": 53220
    },
    {
      "epoch": 7.927464998510575,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.02266418472230913,
      "loss": 0.8011,
      "num_input_tokens_seen": 30898528,
      "step": 53225
    },
    {
      "epoch": 7.928209711051534,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.022662508710579808,
      "loss": 0.8144,
      "num_input_tokens_seen": 30901440,
      "step": 53230
    },
    {
      "epoch": 7.928954423592494,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02266083256940019,
      "loss": 0.7955,
      "num_input_tokens_seen": 30904384,
      "step": 53235
    },
    {
      "epoch": 7.929699136133452,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02265915629879859,
      "loss": 0.815,
      "num_input_tokens_seen": 30907264,
      "step": 53240
    },
    {
      "epoch": 7.930443848674412,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.022657479898803325,
      "loss": 0.8016,
      "num_input_tokens_seen": 30910304,
      "step": 53245
    },
    {
      "epoch": 7.931188561215371,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.022655803369442725,
      "loss": 0.8143,
      "num_input_tokens_seen": 30913120,
      "step": 53250
    },
    {
      "epoch": 7.93193327375633,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.022654126710745105,
      "loss": 0.7973,
      "num_input_tokens_seen": 30916096,
      "step": 53255
    },
    {
      "epoch": 7.932677986297289,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.022652449922738793,
      "loss": 0.7959,
      "num_input_tokens_seen": 30919200,
      "step": 53260
    },
    {
      "epoch": 7.933422698838249,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.022650773005452124,
      "loss": 0.8072,
      "num_input_tokens_seen": 30922048,
      "step": 53265
    },
    {
      "epoch": 7.934167411379208,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.02264909595891342,
      "loss": 0.7975,
      "num_input_tokens_seen": 30925120,
      "step": 53270
    },
    {
      "epoch": 7.934912123920167,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.022647418783151012,
      "loss": 0.7979,
      "num_input_tokens_seen": 30927872,
      "step": 53275
    },
    {
      "epoch": 7.935656836461126,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.022645741478193235,
      "loss": 0.8026,
      "num_input_tokens_seen": 30930784,
      "step": 53280
    },
    {
      "epoch": 7.936401549002085,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.022644064044068424,
      "loss": 0.7988,
      "num_input_tokens_seen": 30933760,
      "step": 53285
    },
    {
      "epoch": 7.937146261543044,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.022642386480804924,
      "loss": 0.7847,
      "num_input_tokens_seen": 30937088,
      "step": 53290
    },
    {
      "epoch": 7.937890974084004,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.022640708788431066,
      "loss": 0.8011,
      "num_input_tokens_seen": 30939744,
      "step": 53295
    },
    {
      "epoch": 7.938635686624963,
      "grad_norm": 0.0137939453125,
      "learning_rate": 0.022639030966975206,
      "loss": 0.8083,
      "num_input_tokens_seen": 30942528,
      "step": 53300
    },
    {
      "epoch": 7.9393803991659215,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.022637353016465682,
      "loss": 0.8,
      "num_input_tokens_seen": 30945664,
      "step": 53305
    },
    {
      "epoch": 7.940125111706881,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.022635674936930836,
      "loss": 0.8055,
      "num_input_tokens_seen": 30948416,
      "step": 53310
    },
    {
      "epoch": 7.940869824247841,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02263399672839902,
      "loss": 0.8107,
      "num_input_tokens_seen": 30951264,
      "step": 53315
    },
    {
      "epoch": 7.9416145367888,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.022632318390898593,
      "loss": 0.7808,
      "num_input_tokens_seen": 30954176,
      "step": 53320
    },
    {
      "epoch": 7.942359249329758,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.0226306399244579,
      "loss": 0.8031,
      "num_input_tokens_seen": 30957088,
      "step": 53325
    },
    {
      "epoch": 7.943103961870718,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.022628961329105305,
      "loss": 0.8165,
      "num_input_tokens_seen": 30959936,
      "step": 53330
    },
    {
      "epoch": 7.943848674411677,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.022627282604869164,
      "loss": 0.7896,
      "num_input_tokens_seen": 30962752,
      "step": 53335
    },
    {
      "epoch": 7.944593386952636,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.022625603751777832,
      "loss": 0.7935,
      "num_input_tokens_seen": 30965920,
      "step": 53340
    },
    {
      "epoch": 7.945338099493595,
      "grad_norm": 0.01190185546875,
      "learning_rate": 0.02262392476985968,
      "loss": 0.8063,
      "num_input_tokens_seen": 30968704,
      "step": 53345
    },
    {
      "epoch": 7.946082812034555,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02262224565914306,
      "loss": 0.7989,
      "num_input_tokens_seen": 30971424,
      "step": 53350
    },
    {
      "epoch": 7.946827524575514,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02262056641965635,
      "loss": 0.8104,
      "num_input_tokens_seen": 30974176,
      "step": 53355
    },
    {
      "epoch": 7.947572237116473,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.022618887051427915,
      "loss": 0.7959,
      "num_input_tokens_seen": 30977120,
      "step": 53360
    },
    {
      "epoch": 7.948316949657432,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02261720755448613,
      "loss": 0.809,
      "num_input_tokens_seen": 30979936,
      "step": 53365
    },
    {
      "epoch": 7.949061662198392,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02261552792885936,
      "loss": 0.8244,
      "num_input_tokens_seen": 30982944,
      "step": 53370
    },
    {
      "epoch": 7.94980637473935,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.022613848174575988,
      "loss": 0.7859,
      "num_input_tokens_seen": 30985888,
      "step": 53375
    },
    {
      "epoch": 7.95055108728031,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02261216829166439,
      "loss": 0.809,
      "num_input_tokens_seen": 30989056,
      "step": 53380
    },
    {
      "epoch": 7.951295799821269,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.022610488280152944,
      "loss": 0.803,
      "num_input_tokens_seen": 30991968,
      "step": 53385
    },
    {
      "epoch": 7.952040512362228,
      "grad_norm": 0.01239013671875,
      "learning_rate": 0.022608808140070033,
      "loss": 0.7974,
      "num_input_tokens_seen": 30994816,
      "step": 53390
    },
    {
      "epoch": 7.952785224903187,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02260712787144404,
      "loss": 0.8171,
      "num_input_tokens_seen": 30997760,
      "step": 53395
    },
    {
      "epoch": 7.953529937444147,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.022605447474303358,
      "loss": 0.7938,
      "num_input_tokens_seen": 31000640,
      "step": 53400
    },
    {
      "epoch": 7.954274649985106,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.02260376694867637,
      "loss": 0.794,
      "num_input_tokens_seen": 31003808,
      "step": 53405
    },
    {
      "epoch": 7.955019362526065,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.022602086294591466,
      "loss": 0.8027,
      "num_input_tokens_seen": 31006720,
      "step": 53410
    },
    {
      "epoch": 7.955764075067024,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.022600405512077038,
      "loss": 0.8203,
      "num_input_tokens_seen": 31009568,
      "step": 53415
    },
    {
      "epoch": 7.956508787607984,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02259872460116149,
      "loss": 0.7768,
      "num_input_tokens_seen": 31012288,
      "step": 53420
    },
    {
      "epoch": 7.957253500148942,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.022597043561873212,
      "loss": 0.8024,
      "num_input_tokens_seen": 31015296,
      "step": 53425
    },
    {
      "epoch": 7.957998212689902,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.0225953623942406,
      "loss": 0.8109,
      "num_input_tokens_seen": 31018048,
      "step": 53430
    },
    {
      "epoch": 7.958742925230861,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.022593681098292063,
      "loss": 0.819,
      "num_input_tokens_seen": 31020960,
      "step": 53435
    },
    {
      "epoch": 7.9594876377718204,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.022591999674055997,
      "loss": 0.7901,
      "num_input_tokens_seen": 31024064,
      "step": 53440
    },
    {
      "epoch": 7.960232350312779,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.022590318121560815,
      "loss": 0.7992,
      "num_input_tokens_seen": 31026688,
      "step": 53445
    },
    {
      "epoch": 7.960977062853738,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.022588636440834923,
      "loss": 0.7853,
      "num_input_tokens_seen": 31029536,
      "step": 53450
    },
    {
      "epoch": 7.961721775394698,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.022586954631906735,
      "loss": 0.8076,
      "num_input_tokens_seen": 31032160,
      "step": 53455
    },
    {
      "epoch": 7.962466487935657,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.022585272694804655,
      "loss": 0.7885,
      "num_input_tokens_seen": 31035200,
      "step": 53460
    },
    {
      "epoch": 7.963211200476616,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.022583590629557105,
      "loss": 0.7942,
      "num_input_tokens_seen": 31038240,
      "step": 53465
    },
    {
      "epoch": 7.963955913017575,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.022581908436192497,
      "loss": 0.8236,
      "num_input_tokens_seen": 31041504,
      "step": 53470
    },
    {
      "epoch": 7.964700625558534,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02258022611473925,
      "loss": 0.778,
      "num_input_tokens_seen": 31044448,
      "step": 53475
    },
    {
      "epoch": 7.965445338099494,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02257854366522579,
      "loss": 0.8001,
      "num_input_tokens_seen": 31047904,
      "step": 53480
    },
    {
      "epoch": 7.966190050640453,
      "grad_norm": 0.02734375,
      "learning_rate": 0.022576861087680537,
      "loss": 0.8239,
      "num_input_tokens_seen": 31050784,
      "step": 53485
    },
    {
      "epoch": 7.966934763181412,
      "grad_norm": 0.01129150390625,
      "learning_rate": 0.022575178382131918,
      "loss": 0.8022,
      "num_input_tokens_seen": 31053536,
      "step": 53490
    },
    {
      "epoch": 7.967679475722371,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.022573495548608352,
      "loss": 0.7992,
      "num_input_tokens_seen": 31056672,
      "step": 53495
    },
    {
      "epoch": 7.96842418826333,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.022571812587138286,
      "loss": 0.7722,
      "num_input_tokens_seen": 31059808,
      "step": 53500
    },
    {
      "epoch": 7.96916890080429,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02257012949775014,
      "loss": 0.7794,
      "num_input_tokens_seen": 31062656,
      "step": 53505
    },
    {
      "epoch": 7.969913613345248,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.022568446280472346,
      "loss": 0.8215,
      "num_input_tokens_seen": 31065312,
      "step": 53510
    },
    {
      "epoch": 7.970658325886208,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.022566762935333346,
      "loss": 0.7942,
      "num_input_tokens_seen": 31068736,
      "step": 53515
    },
    {
      "epoch": 7.971403038427167,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.022565079462361574,
      "loss": 0.7997,
      "num_input_tokens_seen": 31071456,
      "step": 53520
    },
    {
      "epoch": 7.9721477509681264,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.022563395861585477,
      "loss": 0.7738,
      "num_input_tokens_seen": 31074080,
      "step": 53525
    },
    {
      "epoch": 7.972892463509085,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.022561712133033495,
      "loss": 0.8083,
      "num_input_tokens_seen": 31076960,
      "step": 53530
    },
    {
      "epoch": 7.973637176050045,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.022560028276734066,
      "loss": 0.834,
      "num_input_tokens_seen": 31080000,
      "step": 53535
    },
    {
      "epoch": 7.974381888591004,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.02255834429271565,
      "loss": 0.7977,
      "num_input_tokens_seen": 31083104,
      "step": 53540
    },
    {
      "epoch": 7.975126601131963,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.022556660181006688,
      "loss": 0.8106,
      "num_input_tokens_seen": 31085888,
      "step": 53545
    },
    {
      "epoch": 7.975871313672922,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.022554975941635625,
      "loss": 0.7837,
      "num_input_tokens_seen": 31088576,
      "step": 53550
    },
    {
      "epoch": 7.976616026213882,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02255329157463093,
      "loss": 0.7958,
      "num_input_tokens_seen": 31091424,
      "step": 53555
    },
    {
      "epoch": 7.97736073875484,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.022551607080021047,
      "loss": 0.8287,
      "num_input_tokens_seen": 31094432,
      "step": 53560
    },
    {
      "epoch": 7.9781054512958,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.022549922457834437,
      "loss": 0.7935,
      "num_input_tokens_seen": 31097280,
      "step": 53565
    },
    {
      "epoch": 7.978850163836759,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.022548237708099564,
      "loss": 0.7999,
      "num_input_tokens_seen": 31099904,
      "step": 53570
    },
    {
      "epoch": 7.9795948763777185,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.022546552830844883,
      "loss": 0.8044,
      "num_input_tokens_seen": 31102976,
      "step": 53575
    },
    {
      "epoch": 7.980339588918677,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.022544867826098863,
      "loss": 0.8233,
      "num_input_tokens_seen": 31105632,
      "step": 53580
    },
    {
      "epoch": 7.981084301459637,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02254318269388997,
      "loss": 0.788,
      "num_input_tokens_seen": 31108480,
      "step": 53585
    },
    {
      "epoch": 7.981829014000596,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.022541497434246674,
      "loss": 0.7997,
      "num_input_tokens_seen": 31111328,
      "step": 53590
    },
    {
      "epoch": 7.982573726541555,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02253981204719744,
      "loss": 0.7988,
      "num_input_tokens_seen": 31113888,
      "step": 53595
    },
    {
      "epoch": 7.983318439082514,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.02253812653277075,
      "loss": 0.7774,
      "num_input_tokens_seen": 31117056,
      "step": 53600
    },
    {
      "epoch": 7.984063151623474,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.022536440890995067,
      "loss": 0.8037,
      "num_input_tokens_seen": 31119968,
      "step": 53605
    },
    {
      "epoch": 7.9848078641644324,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02253475512189888,
      "loss": 0.8146,
      "num_input_tokens_seen": 31122816,
      "step": 53610
    },
    {
      "epoch": 7.985552576705392,
      "grad_norm": 0.033203125,
      "learning_rate": 0.022533069225510658,
      "loss": 0.8037,
      "num_input_tokens_seen": 31125728,
      "step": 53615
    },
    {
      "epoch": 7.986297289246351,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.022531383201858886,
      "loss": 0.7921,
      "num_input_tokens_seen": 31128704,
      "step": 53620
    },
    {
      "epoch": 7.9870420017873105,
      "grad_norm": 0.01251220703125,
      "learning_rate": 0.022529697050972058,
      "loss": 0.777,
      "num_input_tokens_seen": 31131424,
      "step": 53625
    },
    {
      "epoch": 7.987786714328269,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.022528010772878644,
      "loss": 0.7994,
      "num_input_tokens_seen": 31134176,
      "step": 53630
    },
    {
      "epoch": 7.988531426869228,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.022526324367607146,
      "loss": 0.8114,
      "num_input_tokens_seen": 31136960,
      "step": 53635
    },
    {
      "epoch": 7.989276139410188,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.022524637835186045,
      "loss": 0.7852,
      "num_input_tokens_seen": 31139968,
      "step": 53640
    },
    {
      "epoch": 7.990020851951147,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.022522951175643832,
      "loss": 0.8006,
      "num_input_tokens_seen": 31142784,
      "step": 53645
    },
    {
      "epoch": 7.990765564492106,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.022521264389009005,
      "loss": 0.7964,
      "num_input_tokens_seen": 31145568,
      "step": 53650
    },
    {
      "epoch": 7.991510277033065,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02251957747531006,
      "loss": 0.8141,
      "num_input_tokens_seen": 31148608,
      "step": 53655
    },
    {
      "epoch": 7.9922549895740245,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.022517890434575496,
      "loss": 0.7894,
      "num_input_tokens_seen": 31151424,
      "step": 53660
    },
    {
      "epoch": 7.992999702114983,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02251620326683382,
      "loss": 0.7948,
      "num_input_tokens_seen": 31154368,
      "step": 53665
    },
    {
      "epoch": 7.993744414655943,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.02251451597211353,
      "loss": 0.8293,
      "num_input_tokens_seen": 31157440,
      "step": 53670
    },
    {
      "epoch": 7.994489127196902,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.022512828550443127,
      "loss": 0.792,
      "num_input_tokens_seen": 31160384,
      "step": 53675
    },
    {
      "epoch": 7.995233839737861,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02251114100185111,
      "loss": 0.8049,
      "num_input_tokens_seen": 31163008,
      "step": 53680
    },
    {
      "epoch": 7.99597855227882,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.022509453326366013,
      "loss": 0.8149,
      "num_input_tokens_seen": 31166048,
      "step": 53685
    },
    {
      "epoch": 7.99672326481978,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.02250776552401633,
      "loss": 0.8177,
      "num_input_tokens_seen": 31169152,
      "step": 53690
    },
    {
      "epoch": 7.9974679773607384,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.022506077594830583,
      "loss": 0.7925,
      "num_input_tokens_seen": 31171936,
      "step": 53695
    },
    {
      "epoch": 7.998212689901698,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02250438953883728,
      "loss": 0.8188,
      "num_input_tokens_seen": 31174496,
      "step": 53700
    },
    {
      "epoch": 7.998957402442657,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.022502701356064945,
      "loss": 0.8215,
      "num_input_tokens_seen": 31177248,
      "step": 53705
    },
    {
      "epoch": 7.9997021149836165,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.022501013046542093,
      "loss": 0.7873,
      "num_input_tokens_seen": 31180224,
      "step": 53710
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.8020382523536682,
      "eval_runtime": 70.7731,
      "eval_samples_per_second": 42.163,
      "eval_steps_per_second": 10.541,
      "num_input_tokens_seen": 31180904,
      "step": 53712
    },
    {
      "epoch": 8.000446827524575,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.022499324610297248,
      "loss": 0.804,
      "num_input_tokens_seen": 31182472,
      "step": 53715
    },
    {
      "epoch": 8.001191540065534,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02249763604735894,
      "loss": 0.8004,
      "num_input_tokens_seen": 31185448,
      "step": 53720
    },
    {
      "epoch": 8.001936252606495,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02249594735775569,
      "loss": 0.7829,
      "num_input_tokens_seen": 31188168,
      "step": 53725
    },
    {
      "epoch": 8.002680965147453,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02249425854151603,
      "loss": 0.7923,
      "num_input_tokens_seen": 31190984,
      "step": 53730
    },
    {
      "epoch": 8.003425677688412,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.022492569598668487,
      "loss": 0.7797,
      "num_input_tokens_seen": 31193960,
      "step": 53735
    },
    {
      "epoch": 8.00417039022937,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02249088052924159,
      "loss": 0.8292,
      "num_input_tokens_seen": 31197192,
      "step": 53740
    },
    {
      "epoch": 8.004915102770331,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.022489191333263887,
      "loss": 0.7918,
      "num_input_tokens_seen": 31199944,
      "step": 53745
    },
    {
      "epoch": 8.00565981531129,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.022487502010763907,
      "loss": 0.8241,
      "num_input_tokens_seen": 31202888,
      "step": 53750
    },
    {
      "epoch": 8.006404527852249,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.022485812561770187,
      "loss": 0.8301,
      "num_input_tokens_seen": 31205960,
      "step": 53755
    },
    {
      "epoch": 8.007149240393208,
      "grad_norm": 0.0113525390625,
      "learning_rate": 0.02248412298631127,
      "loss": 0.7769,
      "num_input_tokens_seen": 31208968,
      "step": 53760
    },
    {
      "epoch": 8.007893952934168,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02248243328441571,
      "loss": 0.8259,
      "num_input_tokens_seen": 31211880,
      "step": 53765
    },
    {
      "epoch": 8.008638665475127,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02248074345611204,
      "loss": 0.8146,
      "num_input_tokens_seen": 31214760,
      "step": 53770
    },
    {
      "epoch": 8.009383378016086,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.022479053501428814,
      "loss": 0.8088,
      "num_input_tokens_seen": 31217704,
      "step": 53775
    },
    {
      "epoch": 8.010128090557044,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02247736342039458,
      "loss": 0.7821,
      "num_input_tokens_seen": 31220520,
      "step": 53780
    },
    {
      "epoch": 8.010872803098005,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.022475673213037888,
      "loss": 0.7828,
      "num_input_tokens_seen": 31223432,
      "step": 53785
    },
    {
      "epoch": 8.011617515638964,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.0224739828793873,
      "loss": 0.8364,
      "num_input_tokens_seen": 31226152,
      "step": 53790
    },
    {
      "epoch": 8.012362228179922,
      "grad_norm": 0.021484375,
      "learning_rate": 0.022472292419471362,
      "loss": 0.8121,
      "num_input_tokens_seen": 31228904,
      "step": 53795
    },
    {
      "epoch": 8.013106940720881,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.022470601833318643,
      "loss": 0.8066,
      "num_input_tokens_seen": 31231816,
      "step": 53800
    },
    {
      "epoch": 8.013851653261842,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.022468911120957696,
      "loss": 0.7875,
      "num_input_tokens_seen": 31234952,
      "step": 53805
    },
    {
      "epoch": 8.0145963658028,
      "grad_norm": 0.014892578125,
      "learning_rate": 0.022467220282417086,
      "loss": 0.8233,
      "num_input_tokens_seen": 31238088,
      "step": 53810
    },
    {
      "epoch": 8.01534107834376,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02246552931772538,
      "loss": 0.7931,
      "num_input_tokens_seen": 31240936,
      "step": 53815
    },
    {
      "epoch": 8.016085790884718,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.022463838226911143,
      "loss": 0.8051,
      "num_input_tokens_seen": 31243848,
      "step": 53820
    },
    {
      "epoch": 8.016830503425677,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.022462147010002945,
      "loss": 0.7965,
      "num_input_tokens_seen": 31246920,
      "step": 53825
    },
    {
      "epoch": 8.017575215966637,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02246045566702936,
      "loss": 0.7851,
      "num_input_tokens_seen": 31249768,
      "step": 53830
    },
    {
      "epoch": 8.018319928507596,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.022458764198018955,
      "loss": 0.8002,
      "num_input_tokens_seen": 31252328,
      "step": 53835
    },
    {
      "epoch": 8.019064641048555,
      "grad_norm": 0.0234375,
      "learning_rate": 0.022457072603000314,
      "loss": 0.7882,
      "num_input_tokens_seen": 31255432,
      "step": 53840
    },
    {
      "epoch": 8.019809353589514,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02245538088200201,
      "loss": 0.8293,
      "num_input_tokens_seen": 31258504,
      "step": 53845
    },
    {
      "epoch": 8.020554066130474,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02245368903505262,
      "loss": 0.7924,
      "num_input_tokens_seen": 31261480,
      "step": 53850
    },
    {
      "epoch": 8.021298778671433,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.022451997062180727,
      "loss": 0.825,
      "num_input_tokens_seen": 31264328,
      "step": 53855
    },
    {
      "epoch": 8.022043491212392,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.022450304963414925,
      "loss": 0.8087,
      "num_input_tokens_seen": 31267368,
      "step": 53860
    },
    {
      "epoch": 8.02278820375335,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.022448612738783787,
      "loss": 0.7949,
      "num_input_tokens_seen": 31270344,
      "step": 53865
    },
    {
      "epoch": 8.023532916294311,
      "grad_norm": 0.013427734375,
      "learning_rate": 0.022446920388315913,
      "loss": 0.7906,
      "num_input_tokens_seen": 31273192,
      "step": 53870
    },
    {
      "epoch": 8.02427762883527,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.022445227912039883,
      "loss": 0.818,
      "num_input_tokens_seen": 31275720,
      "step": 53875
    },
    {
      "epoch": 8.025022341376228,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.022443535309984292,
      "loss": 0.8266,
      "num_input_tokens_seen": 31278600,
      "step": 53880
    },
    {
      "epoch": 8.025767053917187,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02244184258217774,
      "loss": 0.8403,
      "num_input_tokens_seen": 31281288,
      "step": 53885
    },
    {
      "epoch": 8.026511766458148,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02244014972864882,
      "loss": 0.8197,
      "num_input_tokens_seen": 31284488,
      "step": 53890
    },
    {
      "epoch": 8.027256478999107,
      "grad_norm": 0.041015625,
      "learning_rate": 0.022438456749426138,
      "loss": 0.8115,
      "num_input_tokens_seen": 31287624,
      "step": 53895
    },
    {
      "epoch": 8.028001191540065,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.022436763644538282,
      "loss": 0.7939,
      "num_input_tokens_seen": 31290760,
      "step": 53900
    },
    {
      "epoch": 8.028745904081024,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.022435070414013866,
      "loss": 0.8076,
      "num_input_tokens_seen": 31293768,
      "step": 53905
    },
    {
      "epoch": 8.029490616621985,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.0224333770578815,
      "loss": 0.8106,
      "num_input_tokens_seen": 31296680,
      "step": 53910
    },
    {
      "epoch": 8.030235329162943,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02243168357616977,
      "loss": 0.8269,
      "num_input_tokens_seen": 31299368,
      "step": 53915
    },
    {
      "epoch": 8.030980041703902,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.022429989968907302,
      "loss": 0.7927,
      "num_input_tokens_seen": 31302632,
      "step": 53920
    },
    {
      "epoch": 8.03172475424486,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.02242829623612271,
      "loss": 0.8106,
      "num_input_tokens_seen": 31305608,
      "step": 53925
    },
    {
      "epoch": 8.032469466785821,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.022426602377844603,
      "loss": 0.807,
      "num_input_tokens_seen": 31308360,
      "step": 53930
    },
    {
      "epoch": 8.03321417932678,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.022424908394101595,
      "loss": 0.7911,
      "num_input_tokens_seen": 31311144,
      "step": 53935
    },
    {
      "epoch": 8.033958891867739,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.0224232142849223,
      "loss": 0.8078,
      "num_input_tokens_seen": 31313960,
      "step": 53940
    },
    {
      "epoch": 8.034703604408698,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02242152005033535,
      "loss": 0.8005,
      "num_input_tokens_seen": 31317160,
      "step": 53945
    },
    {
      "epoch": 8.035448316949658,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.022419825690369367,
      "loss": 0.8128,
      "num_input_tokens_seen": 31320424,
      "step": 53950
    },
    {
      "epoch": 8.036193029490617,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02241813120505296,
      "loss": 0.816,
      "num_input_tokens_seen": 31323112,
      "step": 53955
    },
    {
      "epoch": 8.036937742031576,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02241643659441477,
      "loss": 0.7907,
      "num_input_tokens_seen": 31325864,
      "step": 53960
    },
    {
      "epoch": 8.037682454572534,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02241474185848342,
      "loss": 0.805,
      "num_input_tokens_seen": 31329032,
      "step": 53965
    },
    {
      "epoch": 8.038427167113495,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.022413046997287547,
      "loss": 0.8096,
      "num_input_tokens_seen": 31332296,
      "step": 53970
    },
    {
      "epoch": 8.039171879654454,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.022411352010855777,
      "loss": 0.8151,
      "num_input_tokens_seen": 31335272,
      "step": 53975
    },
    {
      "epoch": 8.039916592195413,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.022409656899216746,
      "loss": 0.8094,
      "num_input_tokens_seen": 31338184,
      "step": 53980
    },
    {
      "epoch": 8.040661304736371,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.022407961662399088,
      "loss": 0.808,
      "num_input_tokens_seen": 31341064,
      "step": 53985
    },
    {
      "epoch": 8.041406017277332,
      "grad_norm": 0.029296875,
      "learning_rate": 0.022406266300431455,
      "loss": 0.812,
      "num_input_tokens_seen": 31344104,
      "step": 53990
    },
    {
      "epoch": 8.04215072981829,
      "grad_norm": 0.041015625,
      "learning_rate": 0.022404570813342475,
      "loss": 0.8031,
      "num_input_tokens_seen": 31347112,
      "step": 53995
    },
    {
      "epoch": 8.04289544235925,
      "grad_norm": 0.044921875,
      "learning_rate": 0.022402875201160802,
      "loss": 0.8078,
      "num_input_tokens_seen": 31350088,
      "step": 54000
    },
    {
      "epoch": 8.043640154900208,
      "grad_norm": 0.03125,
      "learning_rate": 0.022401179463915077,
      "loss": 0.798,
      "num_input_tokens_seen": 31352872,
      "step": 54005
    },
    {
      "epoch": 8.044384867441167,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.022399483601633942,
      "loss": 0.8036,
      "num_input_tokens_seen": 31355720,
      "step": 54010
    },
    {
      "epoch": 8.045129579982127,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02239778761434605,
      "loss": 0.8098,
      "num_input_tokens_seen": 31358600,
      "step": 54015
    },
    {
      "epoch": 8.045874292523086,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.022396091502080058,
      "loss": 0.7994,
      "num_input_tokens_seen": 31361480,
      "step": 54020
    },
    {
      "epoch": 8.046619005064045,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.02239439526486462,
      "loss": 0.8116,
      "num_input_tokens_seen": 31364648,
      "step": 54025
    },
    {
      "epoch": 8.047363717605004,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.022392698902728388,
      "loss": 0.8047,
      "num_input_tokens_seen": 31367272,
      "step": 54030
    },
    {
      "epoch": 8.048108430145964,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02239100241570002,
      "loss": 0.8024,
      "num_input_tokens_seen": 31370472,
      "step": 54035
    },
    {
      "epoch": 8.048853142686923,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02238930580380818,
      "loss": 0.8101,
      "num_input_tokens_seen": 31373608,
      "step": 54040
    },
    {
      "epoch": 8.049597855227882,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.022387609067081527,
      "loss": 0.8121,
      "num_input_tokens_seen": 31376552,
      "step": 54045
    },
    {
      "epoch": 8.05034256776884,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.022385912205548725,
      "loss": 0.7968,
      "num_input_tokens_seen": 31379304,
      "step": 54050
    },
    {
      "epoch": 8.051087280309801,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.022384215219238446,
      "loss": 0.8131,
      "num_input_tokens_seen": 31382152,
      "step": 54055
    },
    {
      "epoch": 8.05183199285076,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.02238251810817936,
      "loss": 0.7913,
      "num_input_tokens_seen": 31385224,
      "step": 54060
    },
    {
      "epoch": 8.052576705391719,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.022380820872400127,
      "loss": 0.8063,
      "num_input_tokens_seen": 31388168,
      "step": 54065
    },
    {
      "epoch": 8.053321417932677,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02237912351192943,
      "loss": 0.7928,
      "num_input_tokens_seen": 31390888,
      "step": 54070
    },
    {
      "epoch": 8.054066130473638,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02237742602679594,
      "loss": 0.8025,
      "num_input_tokens_seen": 31393640,
      "step": 54075
    },
    {
      "epoch": 8.054810843014597,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.022375728417028333,
      "loss": 0.7929,
      "num_input_tokens_seen": 31396840,
      "step": 54080
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.022374030682655296,
      "loss": 0.7971,
      "num_input_tokens_seen": 31399688,
      "step": 54085
    },
    {
      "epoch": 8.056300268096514,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.0223723328237055,
      "loss": 0.8073,
      "num_input_tokens_seen": 31402376,
      "step": 54090
    },
    {
      "epoch": 8.057044980637475,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02237063484020764,
      "loss": 0.7998,
      "num_input_tokens_seen": 31405640,
      "step": 54095
    },
    {
      "epoch": 8.057789693178433,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.02236893673219039,
      "loss": 0.796,
      "num_input_tokens_seen": 31408712,
      "step": 54100
    },
    {
      "epoch": 8.058534405719392,
      "grad_norm": 0.0133056640625,
      "learning_rate": 0.022367238499682442,
      "loss": 0.7983,
      "num_input_tokens_seen": 31411496,
      "step": 54105
    },
    {
      "epoch": 8.059279118260351,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.022365540142712494,
      "loss": 0.798,
      "num_input_tokens_seen": 31414440,
      "step": 54110
    },
    {
      "epoch": 8.060023830801311,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02236384166130923,
      "loss": 0.7928,
      "num_input_tokens_seen": 31417224,
      "step": 54115
    },
    {
      "epoch": 8.06076854334227,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02236214305550134,
      "loss": 0.8182,
      "num_input_tokens_seen": 31420168,
      "step": 54120
    },
    {
      "epoch": 8.061513255883229,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.02236044432531753,
      "loss": 0.8093,
      "num_input_tokens_seen": 31423112,
      "step": 54125
    },
    {
      "epoch": 8.062257968424188,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.022358745470786496,
      "loss": 0.8014,
      "num_input_tokens_seen": 31426408,
      "step": 54130
    },
    {
      "epoch": 8.063002680965148,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02235704649193693,
      "loss": 0.8011,
      "num_input_tokens_seen": 31429384,
      "step": 54135
    },
    {
      "epoch": 8.063747393506107,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.022355347388797544,
      "loss": 0.7948,
      "num_input_tokens_seen": 31432328,
      "step": 54140
    },
    {
      "epoch": 8.064492106047066,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02235364816139704,
      "loss": 0.8358,
      "num_input_tokens_seen": 31435400,
      "step": 54145
    },
    {
      "epoch": 8.065236818588025,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.022351948809764124,
      "loss": 0.8067,
      "num_input_tokens_seen": 31438216,
      "step": 54150
    },
    {
      "epoch": 8.065981531128985,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.022350249333927506,
      "loss": 0.7856,
      "num_input_tokens_seen": 31441128,
      "step": 54155
    },
    {
      "epoch": 8.066726243669944,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.022348549733915892,
      "loss": 0.7974,
      "num_input_tokens_seen": 31444136,
      "step": 54160
    },
    {
      "epoch": 8.067470956210903,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.022346850009758005,
      "loss": 0.8158,
      "num_input_tokens_seen": 31447432,
      "step": 54165
    },
    {
      "epoch": 8.068215668751861,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02234515016148255,
      "loss": 0.7832,
      "num_input_tokens_seen": 31450440,
      "step": 54170
    },
    {
      "epoch": 8.06896038129282,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.022343450189118246,
      "loss": 0.7759,
      "num_input_tokens_seen": 31453288,
      "step": 54175
    },
    {
      "epoch": 8.06970509383378,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.022341750092693822,
      "loss": 0.8169,
      "num_input_tokens_seen": 31456136,
      "step": 54180
    },
    {
      "epoch": 8.07044980637474,
      "grad_norm": 0.014404296875,
      "learning_rate": 0.022340049872237987,
      "loss": 0.8222,
      "num_input_tokens_seen": 31459048,
      "step": 54185
    },
    {
      "epoch": 8.071194518915698,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.02233834952777947,
      "loss": 0.817,
      "num_input_tokens_seen": 31461928,
      "step": 54190
    },
    {
      "epoch": 8.071939231456657,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.022336649059346996,
      "loss": 0.8173,
      "num_input_tokens_seen": 31464872,
      "step": 54195
    },
    {
      "epoch": 8.072683943997617,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.022334948466969293,
      "loss": 0.7753,
      "num_input_tokens_seen": 31467464,
      "step": 54200
    },
    {
      "epoch": 8.073428656538576,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.022333247750675092,
      "loss": 0.7913,
      "num_input_tokens_seen": 31470344,
      "step": 54205
    },
    {
      "epoch": 8.074173369079535,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.022331546910493123,
      "loss": 0.8002,
      "num_input_tokens_seen": 31473352,
      "step": 54210
    },
    {
      "epoch": 8.074918081620494,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02232984594645212,
      "loss": 0.8202,
      "num_input_tokens_seen": 31476328,
      "step": 54215
    },
    {
      "epoch": 8.075662794161454,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.022328144858580815,
      "loss": 0.7896,
      "num_input_tokens_seen": 31479656,
      "step": 54220
    },
    {
      "epoch": 8.076407506702413,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02232644364690796,
      "loss": 0.8159,
      "num_input_tokens_seen": 31482536,
      "step": 54225
    },
    {
      "epoch": 8.077152219243372,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.022324742311462278,
      "loss": 0.8148,
      "num_input_tokens_seen": 31485480,
      "step": 54230
    },
    {
      "epoch": 8.07789693178433,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.022323040852272522,
      "loss": 0.8129,
      "num_input_tokens_seen": 31488392,
      "step": 54235
    },
    {
      "epoch": 8.078641644325291,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.022321339269367434,
      "loss": 0.8021,
      "num_input_tokens_seen": 31491304,
      "step": 54240
    },
    {
      "epoch": 8.07938635686625,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02231963756277576,
      "loss": 0.7996,
      "num_input_tokens_seen": 31494280,
      "step": 54245
    },
    {
      "epoch": 8.080131069407209,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.022317935732526245,
      "loss": 0.7853,
      "num_input_tokens_seen": 31497128,
      "step": 54250
    },
    {
      "epoch": 8.080875781948167,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.022316233778647648,
      "loss": 0.8027,
      "num_input_tokens_seen": 31499848,
      "step": 54255
    },
    {
      "epoch": 8.081620494489128,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.022314531701168715,
      "loss": 0.855,
      "num_input_tokens_seen": 31502760,
      "step": 54260
    },
    {
      "epoch": 8.082365207030087,
      "grad_norm": 0.021484375,
      "learning_rate": 0.022312829500118207,
      "loss": 0.7939,
      "num_input_tokens_seen": 31505480,
      "step": 54265
    },
    {
      "epoch": 8.083109919571045,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.022311127175524872,
      "loss": 0.817,
      "num_input_tokens_seen": 31508104,
      "step": 54270
    },
    {
      "epoch": 8.083854632112004,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.022309424727417475,
      "loss": 0.7977,
      "num_input_tokens_seen": 31511272,
      "step": 54275
    },
    {
      "epoch": 8.084599344652965,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02230772215582478,
      "loss": 0.8093,
      "num_input_tokens_seen": 31514216,
      "step": 54280
    },
    {
      "epoch": 8.085344057193923,
      "grad_norm": 0.01544189453125,
      "learning_rate": 0.022306019460775543,
      "loss": 0.8148,
      "num_input_tokens_seen": 31517096,
      "step": 54285
    },
    {
      "epoch": 8.086088769734882,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02230431664229853,
      "loss": 0.7919,
      "num_input_tokens_seen": 31519912,
      "step": 54290
    },
    {
      "epoch": 8.086833482275841,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.022302613700422518,
      "loss": 0.7994,
      "num_input_tokens_seen": 31522824,
      "step": 54295
    },
    {
      "epoch": 8.087578194816802,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.022300910635176266,
      "loss": 0.8156,
      "num_input_tokens_seen": 31525960,
      "step": 54300
    },
    {
      "epoch": 8.08832290735776,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02229920744658855,
      "loss": 0.8071,
      "num_input_tokens_seen": 31528936,
      "step": 54305
    },
    {
      "epoch": 8.089067619898719,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02229750413468814,
      "loss": 0.8145,
      "num_input_tokens_seen": 31531656,
      "step": 54310
    },
    {
      "epoch": 8.089812332439678,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.022295800699503814,
      "loss": 0.7879,
      "num_input_tokens_seen": 31534440,
      "step": 54315
    },
    {
      "epoch": 8.090557044980638,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.022294097141064353,
      "loss": 0.8027,
      "num_input_tokens_seen": 31537128,
      "step": 54320
    },
    {
      "epoch": 8.091301757521597,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.022292393459398534,
      "loss": 0.791,
      "num_input_tokens_seen": 31539848,
      "step": 54325
    },
    {
      "epoch": 8.092046470062556,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.022290689654535144,
      "loss": 0.797,
      "num_input_tokens_seen": 31542792,
      "step": 54330
    },
    {
      "epoch": 8.092791182603515,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.022288985726502956,
      "loss": 0.8037,
      "num_input_tokens_seen": 31545640,
      "step": 54335
    },
    {
      "epoch": 8.093535895144473,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.022287281675330763,
      "loss": 0.8013,
      "num_input_tokens_seen": 31548392,
      "step": 54340
    },
    {
      "epoch": 8.094280607685434,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.022285577501047354,
      "loss": 0.8077,
      "num_input_tokens_seen": 31551240,
      "step": 54345
    },
    {
      "epoch": 8.095025320226393,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.022283873203681512,
      "loss": 0.8082,
      "num_input_tokens_seen": 31554472,
      "step": 54350
    },
    {
      "epoch": 8.095770032767351,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.022282168783262044,
      "loss": 0.8071,
      "num_input_tokens_seen": 31557224,
      "step": 54355
    },
    {
      "epoch": 8.09651474530831,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02228046423981773,
      "loss": 0.8058,
      "num_input_tokens_seen": 31560168,
      "step": 54360
    },
    {
      "epoch": 8.09725945784927,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.022278759573377376,
      "loss": 0.8055,
      "num_input_tokens_seen": 31562920,
      "step": 54365
    },
    {
      "epoch": 8.09800417039023,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.022277054783969776,
      "loss": 0.7852,
      "num_input_tokens_seen": 31566056,
      "step": 54370
    },
    {
      "epoch": 8.098748882931188,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02227534987162373,
      "loss": 0.7831,
      "num_input_tokens_seen": 31568744,
      "step": 54375
    },
    {
      "epoch": 8.099493595472147,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02227364483636804,
      "loss": 0.7902,
      "num_input_tokens_seen": 31571688,
      "step": 54380
    },
    {
      "epoch": 8.100238308013108,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.022271939678231517,
      "loss": 0.825,
      "num_input_tokens_seen": 31574632,
      "step": 54385
    },
    {
      "epoch": 8.100983020554066,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.022270234397242964,
      "loss": 0.7977,
      "num_input_tokens_seen": 31577672,
      "step": 54390
    },
    {
      "epoch": 8.101727733095025,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.022268528993431187,
      "loss": 0.8093,
      "num_input_tokens_seen": 31580616,
      "step": 54395
    },
    {
      "epoch": 8.102472445635984,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.022266823466825004,
      "loss": 0.8093,
      "num_input_tokens_seen": 31583368,
      "step": 54400
    },
    {
      "epoch": 8.103217158176944,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02226511781745322,
      "loss": 0.7852,
      "num_input_tokens_seen": 31586280,
      "step": 54405
    },
    {
      "epoch": 8.103961870717903,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.022263412045344655,
      "loss": 0.7831,
      "num_input_tokens_seen": 31589032,
      "step": 54410
    },
    {
      "epoch": 8.104706583258862,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02226170615052813,
      "loss": 0.8008,
      "num_input_tokens_seen": 31591848,
      "step": 54415
    },
    {
      "epoch": 8.10545129579982,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.022260000133032455,
      "loss": 0.7819,
      "num_input_tokens_seen": 31594888,
      "step": 54420
    },
    {
      "epoch": 8.106196008340781,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.022258293992886462,
      "loss": 0.7859,
      "num_input_tokens_seen": 31598216,
      "step": 54425
    },
    {
      "epoch": 8.10694072088174,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.022256587730118967,
      "loss": 0.7876,
      "num_input_tokens_seen": 31601128,
      "step": 54430
    },
    {
      "epoch": 8.107685433422699,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.022254881344758795,
      "loss": 0.8056,
      "num_input_tokens_seen": 31603912,
      "step": 54435
    },
    {
      "epoch": 8.108430145963657,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.022253174836834776,
      "loss": 0.7892,
      "num_input_tokens_seen": 31606856,
      "step": 54440
    },
    {
      "epoch": 8.109174858504618,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02225146820637574,
      "loss": 0.7823,
      "num_input_tokens_seen": 31609544,
      "step": 54445
    },
    {
      "epoch": 8.109919571045577,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.022249761453410517,
      "loss": 0.7655,
      "num_input_tokens_seen": 31612520,
      "step": 54450
    },
    {
      "epoch": 8.110664283586535,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.022248054577967948,
      "loss": 0.7802,
      "num_input_tokens_seen": 31615368,
      "step": 54455
    },
    {
      "epoch": 8.111408996127494,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.022246347580076865,
      "loss": 0.8261,
      "num_input_tokens_seen": 31618184,
      "step": 54460
    },
    {
      "epoch": 8.112153708668455,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.0222446404597661,
      "loss": 0.8033,
      "num_input_tokens_seen": 31620968,
      "step": 54465
    },
    {
      "epoch": 8.112898421209414,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.022242933217064494,
      "loss": 0.8017,
      "num_input_tokens_seen": 31623976,
      "step": 54470
    },
    {
      "epoch": 8.113643133750372,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.022241225852000904,
      "loss": 0.7892,
      "num_input_tokens_seen": 31626920,
      "step": 54475
    },
    {
      "epoch": 8.114387846291331,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.022239518364604152,
      "loss": 0.8022,
      "num_input_tokens_seen": 31629832,
      "step": 54480
    },
    {
      "epoch": 8.115132558832292,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.0222378107549031,
      "loss": 0.7679,
      "num_input_tokens_seen": 31632776,
      "step": 54485
    },
    {
      "epoch": 8.11587727137325,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.022236103022926588,
      "loss": 0.7816,
      "num_input_tokens_seen": 31635656,
      "step": 54490
    },
    {
      "epoch": 8.116621983914209,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.022234395168703475,
      "loss": 0.8069,
      "num_input_tokens_seen": 31638504,
      "step": 54495
    },
    {
      "epoch": 8.117366696455168,
      "grad_norm": 0.0234375,
      "learning_rate": 0.022232687192262603,
      "loss": 0.7815,
      "num_input_tokens_seen": 31641544,
      "step": 54500
    },
    {
      "epoch": 8.118111408996128,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.022230979093632836,
      "loss": 0.7814,
      "num_input_tokens_seen": 31644360,
      "step": 54505
    },
    {
      "epoch": 8.118856121537087,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.022229270872843022,
      "loss": 0.798,
      "num_input_tokens_seen": 31647272,
      "step": 54510
    },
    {
      "epoch": 8.119600834078046,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02222756252992203,
      "loss": 0.8257,
      "num_input_tokens_seen": 31650440,
      "step": 54515
    },
    {
      "epoch": 8.120345546619005,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.022225854064898708,
      "loss": 0.7714,
      "num_input_tokens_seen": 31653576,
      "step": 54520
    },
    {
      "epoch": 8.121090259159963,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02222414547780193,
      "loss": 0.827,
      "num_input_tokens_seen": 31656488,
      "step": 54525
    },
    {
      "epoch": 8.121834971700924,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.022222436768660557,
      "loss": 0.8096,
      "num_input_tokens_seen": 31659112,
      "step": 54530
    },
    {
      "epoch": 8.122579684241883,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.022220727937503456,
      "loss": 0.7991,
      "num_input_tokens_seen": 31662056,
      "step": 54535
    },
    {
      "epoch": 8.123324396782841,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02221901898435949,
      "loss": 0.8252,
      "num_input_tokens_seen": 31664936,
      "step": 54540
    },
    {
      "epoch": 8.1240691093238,
      "grad_norm": 0.017578125,
      "learning_rate": 0.022217309909257542,
      "loss": 0.8215,
      "num_input_tokens_seen": 31668232,
      "step": 54545
    },
    {
      "epoch": 8.12481382186476,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02221560071222647,
      "loss": 0.8229,
      "num_input_tokens_seen": 31671144,
      "step": 54550
    },
    {
      "epoch": 8.12555853440572,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.022213891393295167,
      "loss": 0.7762,
      "num_input_tokens_seen": 31673992,
      "step": 54555
    },
    {
      "epoch": 8.126303246946678,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.022212181952492496,
      "loss": 0.8055,
      "num_input_tokens_seen": 31676936,
      "step": 54560
    },
    {
      "epoch": 8.127047959487637,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.022210472389847344,
      "loss": 0.7883,
      "num_input_tokens_seen": 31680008,
      "step": 54565
    },
    {
      "epoch": 8.127792672028598,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.022208762705388588,
      "loss": 0.8019,
      "num_input_tokens_seen": 31682760,
      "step": 54570
    },
    {
      "epoch": 8.128537384569556,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.02220705289914511,
      "loss": 0.802,
      "num_input_tokens_seen": 31685672,
      "step": 54575
    },
    {
      "epoch": 8.129282097110515,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0222053429711458,
      "loss": 0.7629,
      "num_input_tokens_seen": 31688392,
      "step": 54580
    },
    {
      "epoch": 8.130026809651474,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.022203632921419542,
      "loss": 0.7764,
      "num_input_tokens_seen": 31691528,
      "step": 54585
    },
    {
      "epoch": 8.130771522192434,
      "grad_norm": 0.0142822265625,
      "learning_rate": 0.02220192274999523,
      "loss": 0.7744,
      "num_input_tokens_seen": 31694568,
      "step": 54590
    },
    {
      "epoch": 8.131516234733393,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.022200212456901756,
      "loss": 0.7984,
      "num_input_tokens_seen": 31698152,
      "step": 54595
    },
    {
      "epoch": 8.132260947274352,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.022198502042168004,
      "loss": 0.8199,
      "num_input_tokens_seen": 31700968,
      "step": 54600
    },
    {
      "epoch": 8.13300565981531,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.02219679150582288,
      "loss": 0.8046,
      "num_input_tokens_seen": 31704136,
      "step": 54605
    },
    {
      "epoch": 8.133750372356271,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.022195080847895274,
      "loss": 0.7749,
      "num_input_tokens_seen": 31706952,
      "step": 54610
    },
    {
      "epoch": 8.13449508489723,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.02219337006841409,
      "loss": 0.8061,
      "num_input_tokens_seen": 31709864,
      "step": 54615
    },
    {
      "epoch": 8.135239797438189,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.022191659167408232,
      "loss": 0.8398,
      "num_input_tokens_seen": 31713128,
      "step": 54620
    },
    {
      "epoch": 8.135984509979147,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.022189948144906605,
      "loss": 0.7973,
      "num_input_tokens_seen": 31716008,
      "step": 54625
    },
    {
      "epoch": 8.136729222520108,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.022188237000938107,
      "loss": 0.7849,
      "num_input_tokens_seen": 31719048,
      "step": 54630
    },
    {
      "epoch": 8.137473935061067,
      "grad_norm": 0.01287841796875,
      "learning_rate": 0.022186525735531655,
      "loss": 0.7814,
      "num_input_tokens_seen": 31721608,
      "step": 54635
    },
    {
      "epoch": 8.138218647602026,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02218481434871615,
      "loss": 0.7923,
      "num_input_tokens_seen": 31724520,
      "step": 54640
    },
    {
      "epoch": 8.138963360142984,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02218310284052051,
      "loss": 0.8073,
      "num_input_tokens_seen": 31727240,
      "step": 54645
    },
    {
      "epoch": 8.139708072683945,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.02218139121097365,
      "loss": 0.7825,
      "num_input_tokens_seen": 31730120,
      "step": 54650
    },
    {
      "epoch": 8.140452785224904,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.022179679460104483,
      "loss": 0.814,
      "num_input_tokens_seen": 31732872,
      "step": 54655
    },
    {
      "epoch": 8.141197497765862,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02217796758794193,
      "loss": 0.815,
      "num_input_tokens_seen": 31735976,
      "step": 54660
    },
    {
      "epoch": 8.141942210306821,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02217625559451491,
      "loss": 0.7858,
      "num_input_tokens_seen": 31738984,
      "step": 54665
    },
    {
      "epoch": 8.142686922847782,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.02217454347985234,
      "loss": 0.7793,
      "num_input_tokens_seen": 31741736,
      "step": 54670
    },
    {
      "epoch": 8.14343163538874,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.022172831243983154,
      "loss": 0.8039,
      "num_input_tokens_seen": 31745064,
      "step": 54675
    },
    {
      "epoch": 8.1441763479297,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.022171118886936275,
      "loss": 0.8294,
      "num_input_tokens_seen": 31748072,
      "step": 54680
    },
    {
      "epoch": 8.144921060470658,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02216940640874063,
      "loss": 0.8156,
      "num_input_tokens_seen": 31750888,
      "step": 54685
    },
    {
      "epoch": 8.145665773011617,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.022167693809425147,
      "loss": 0.7717,
      "num_input_tokens_seen": 31753896,
      "step": 54690
    },
    {
      "epoch": 8.146410485552577,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02216598108901877,
      "loss": 0.8,
      "num_input_tokens_seen": 31756584,
      "step": 54695
    },
    {
      "epoch": 8.147155198093536,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02216426824755042,
      "loss": 0.7702,
      "num_input_tokens_seen": 31759336,
      "step": 54700
    },
    {
      "epoch": 8.147899910634495,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.022162555285049037,
      "loss": 0.8465,
      "num_input_tokens_seen": 31762152,
      "step": 54705
    },
    {
      "epoch": 8.148644623175453,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.022160842201543567,
      "loss": 0.8065,
      "num_input_tokens_seen": 31765160,
      "step": 54710
    },
    {
      "epoch": 8.149389335716414,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02215912899706295,
      "loss": 0.8141,
      "num_input_tokens_seen": 31767976,
      "step": 54715
    },
    {
      "epoch": 8.150134048257373,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.022157415671636115,
      "loss": 0.7971,
      "num_input_tokens_seen": 31770856,
      "step": 54720
    },
    {
      "epoch": 8.150878760798332,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02215570222529202,
      "loss": 0.78,
      "num_input_tokens_seen": 31773864,
      "step": 54725
    },
    {
      "epoch": 8.15162347333929,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.02215398865805961,
      "loss": 0.7864,
      "num_input_tokens_seen": 31776776,
      "step": 54730
    },
    {
      "epoch": 8.15236818588025,
      "grad_norm": 0.01519775390625,
      "learning_rate": 0.02215227496996783,
      "loss": 0.8164,
      "num_input_tokens_seen": 31779496,
      "step": 54735
    },
    {
      "epoch": 8.15311289842121,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.022150561161045638,
      "loss": 0.7839,
      "num_input_tokens_seen": 31782536,
      "step": 54740
    },
    {
      "epoch": 8.153857610962168,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.02214884723132198,
      "loss": 0.8211,
      "num_input_tokens_seen": 31785448,
      "step": 54745
    },
    {
      "epoch": 8.154602323503127,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.022147133180825812,
      "loss": 0.8044,
      "num_input_tokens_seen": 31788232,
      "step": 54750
    },
    {
      "epoch": 8.155347036044088,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.0221454190095861,
      "loss": 0.8193,
      "num_input_tokens_seen": 31791240,
      "step": 54755
    },
    {
      "epoch": 8.156091748585046,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02214370471763179,
      "loss": 0.8227,
      "num_input_tokens_seen": 31794088,
      "step": 54760
    },
    {
      "epoch": 8.156836461126005,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02214199030499185,
      "loss": 0.8052,
      "num_input_tokens_seen": 31797128,
      "step": 54765
    },
    {
      "epoch": 8.157581173666964,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.022140275771695243,
      "loss": 0.8289,
      "num_input_tokens_seen": 31799912,
      "step": 54770
    },
    {
      "epoch": 8.158325886207924,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.022138561117770936,
      "loss": 0.7847,
      "num_input_tokens_seen": 31803016,
      "step": 54775
    },
    {
      "epoch": 8.159070598748883,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02213684634324789,
      "loss": 0.8013,
      "num_input_tokens_seen": 31805832,
      "step": 54780
    },
    {
      "epoch": 8.159815311289842,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.02213513144815508,
      "loss": 0.8114,
      "num_input_tokens_seen": 31808776,
      "step": 54785
    },
    {
      "epoch": 8.1605600238308,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02213341643252148,
      "loss": 0.8045,
      "num_input_tokens_seen": 31812200,
      "step": 54790
    },
    {
      "epoch": 8.161304736371761,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02213170129637606,
      "loss": 0.7847,
      "num_input_tokens_seen": 31814952,
      "step": 54795
    },
    {
      "epoch": 8.16204944891272,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02212998603974779,
      "loss": 0.7987,
      "num_input_tokens_seen": 31817864,
      "step": 54800
    },
    {
      "epoch": 8.162794161453679,
      "grad_norm": 0.046875,
      "learning_rate": 0.022128270662665653,
      "loss": 0.8261,
      "num_input_tokens_seen": 31820872,
      "step": 54805
    },
    {
      "epoch": 8.163538873994638,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02212655516515863,
      "loss": 0.8085,
      "num_input_tokens_seen": 31823656,
      "step": 54810
    },
    {
      "epoch": 8.164283586535598,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.022124839547255696,
      "loss": 0.8053,
      "num_input_tokens_seen": 31826760,
      "step": 54815
    },
    {
      "epoch": 8.165028299076557,
      "grad_norm": 0.021484375,
      "learning_rate": 0.022123123808985845,
      "loss": 0.7833,
      "num_input_tokens_seen": 31829768,
      "step": 54820
    },
    {
      "epoch": 8.165773011617516,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.022121407950378053,
      "loss": 0.8148,
      "num_input_tokens_seen": 31832776,
      "step": 54825
    },
    {
      "epoch": 8.166517724158474,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02211969197146131,
      "loss": 0.8037,
      "num_input_tokens_seen": 31835720,
      "step": 54830
    },
    {
      "epoch": 8.167262436699435,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.022117975872264613,
      "loss": 0.799,
      "num_input_tokens_seen": 31838408,
      "step": 54835
    },
    {
      "epoch": 8.168007149240394,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02211625965281695,
      "loss": 0.7958,
      "num_input_tokens_seen": 31841416,
      "step": 54840
    },
    {
      "epoch": 8.168751861781352,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.022114543313147307,
      "loss": 0.7874,
      "num_input_tokens_seen": 31844232,
      "step": 54845
    },
    {
      "epoch": 8.169496574322311,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.022112826853284685,
      "loss": 0.8192,
      "num_input_tokens_seen": 31847208,
      "step": 54850
    },
    {
      "epoch": 8.17024128686327,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.022111110273258083,
      "loss": 0.7894,
      "num_input_tokens_seen": 31849832,
      "step": 54855
    },
    {
      "epoch": 8.17098599940423,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.022109393573096502,
      "loss": 0.8086,
      "num_input_tokens_seen": 31852776,
      "step": 54860
    },
    {
      "epoch": 8.17173071194519,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.02210767675282894,
      "loss": 0.8064,
      "num_input_tokens_seen": 31856008,
      "step": 54865
    },
    {
      "epoch": 8.172475424486148,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.022105959812484408,
      "loss": 0.8096,
      "num_input_tokens_seen": 31858888,
      "step": 54870
    },
    {
      "epoch": 8.173220137027107,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.022104242752091906,
      "loss": 0.7896,
      "num_input_tokens_seen": 31861672,
      "step": 54875
    },
    {
      "epoch": 8.173964849568067,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02210252557168044,
      "loss": 0.7913,
      "num_input_tokens_seen": 31864744,
      "step": 54880
    },
    {
      "epoch": 8.174709562109026,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.022100808271279023,
      "loss": 0.7934,
      "num_input_tokens_seen": 31867912,
      "step": 54885
    },
    {
      "epoch": 8.175454274649985,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02209909085091667,
      "loss": 0.7963,
      "num_input_tokens_seen": 31870856,
      "step": 54890
    },
    {
      "epoch": 8.176198987190944,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.022097373310622392,
      "loss": 0.8175,
      "num_input_tokens_seen": 31873832,
      "step": 54895
    },
    {
      "epoch": 8.176943699731904,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.022095655650425202,
      "loss": 0.7914,
      "num_input_tokens_seen": 31876424,
      "step": 54900
    },
    {
      "epoch": 8.177688412272863,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.022093937870354123,
      "loss": 0.8069,
      "num_input_tokens_seen": 31879368,
      "step": 54905
    },
    {
      "epoch": 8.178433124813822,
      "grad_norm": 0.025390625,
      "learning_rate": 0.022092219970438174,
      "loss": 0.8183,
      "num_input_tokens_seen": 31882312,
      "step": 54910
    },
    {
      "epoch": 8.17917783735478,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.022090501950706377,
      "loss": 0.8044,
      "num_input_tokens_seen": 31885224,
      "step": 54915
    },
    {
      "epoch": 8.17992254989574,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02208878381118776,
      "loss": 0.7738,
      "num_input_tokens_seen": 31888040,
      "step": 54920
    },
    {
      "epoch": 8.1806672624367,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.022087065551911342,
      "loss": 0.8079,
      "num_input_tokens_seen": 31891144,
      "step": 54925
    },
    {
      "epoch": 8.181411974977658,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.022085347172906156,
      "loss": 0.8234,
      "num_input_tokens_seen": 31894216,
      "step": 54930
    },
    {
      "epoch": 8.182156687518617,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.022083628674201227,
      "loss": 0.8037,
      "num_input_tokens_seen": 31897192,
      "step": 54935
    },
    {
      "epoch": 8.182901400059578,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02208191005582559,
      "loss": 0.8161,
      "num_input_tokens_seen": 31900072,
      "step": 54940
    },
    {
      "epoch": 8.183646112600536,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.022080191317808282,
      "loss": 0.7914,
      "num_input_tokens_seen": 31903112,
      "step": 54945
    },
    {
      "epoch": 8.184390825141495,
      "grad_norm": 0.02734375,
      "learning_rate": 0.022078472460178338,
      "loss": 0.7799,
      "num_input_tokens_seen": 31905864,
      "step": 54950
    },
    {
      "epoch": 8.185135537682454,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.0220767534829648,
      "loss": 0.8059,
      "num_input_tokens_seen": 31909096,
      "step": 54955
    },
    {
      "epoch": 8.185880250223414,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.022075034386196703,
      "loss": 0.8035,
      "num_input_tokens_seen": 31911848,
      "step": 54960
    },
    {
      "epoch": 8.186624962764373,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.022073315169903088,
      "loss": 0.7823,
      "num_input_tokens_seen": 31914536,
      "step": 54965
    },
    {
      "epoch": 8.187369675305332,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.022071595834113,
      "loss": 0.786,
      "num_input_tokens_seen": 31917448,
      "step": 54970
    },
    {
      "epoch": 8.18811438784629,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02206987637885549,
      "loss": 0.8182,
      "num_input_tokens_seen": 31920456,
      "step": 54975
    },
    {
      "epoch": 8.188859100387251,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.022068156804159605,
      "loss": 0.8094,
      "num_input_tokens_seen": 31923368,
      "step": 54980
    },
    {
      "epoch": 8.18960381292821,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.022066437110054397,
      "loss": 0.8023,
      "num_input_tokens_seen": 31926376,
      "step": 54985
    },
    {
      "epoch": 8.190348525469169,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.022064717296568916,
      "loss": 0.7883,
      "num_input_tokens_seen": 31929160,
      "step": 54990
    },
    {
      "epoch": 8.191093238010128,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02206299736373221,
      "loss": 0.7918,
      "num_input_tokens_seen": 31932040,
      "step": 54995
    },
    {
      "epoch": 8.191837950551088,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.022061277311573348,
      "loss": 0.7946,
      "num_input_tokens_seen": 31935432,
      "step": 55000
    },
    {
      "epoch": 8.192582663092047,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.022059557140121384,
      "loss": 0.8045,
      "num_input_tokens_seen": 31938120,
      "step": 55005
    },
    {
      "epoch": 8.193327375633006,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02205783684940537,
      "loss": 0.7982,
      "num_input_tokens_seen": 31940968,
      "step": 55010
    },
    {
      "epoch": 8.194072088173964,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.02205611643945438,
      "loss": 0.7895,
      "num_input_tokens_seen": 31943720,
      "step": 55015
    },
    {
      "epoch": 8.194816800714925,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02205439591029747,
      "loss": 0.7944,
      "num_input_tokens_seen": 31946664,
      "step": 55020
    },
    {
      "epoch": 8.195561513255884,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.022052675261963715,
      "loss": 0.7958,
      "num_input_tokens_seen": 31949544,
      "step": 55025
    },
    {
      "epoch": 8.196306225796842,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.022050954494482182,
      "loss": 0.8142,
      "num_input_tokens_seen": 31952392,
      "step": 55030
    },
    {
      "epoch": 8.197050938337801,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.022049233607881934,
      "loss": 0.7951,
      "num_input_tokens_seen": 31955304,
      "step": 55035
    },
    {
      "epoch": 8.19779565087876,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.022047512602192055,
      "loss": 0.7986,
      "num_input_tokens_seen": 31958056,
      "step": 55040
    },
    {
      "epoch": 8.19854036341972,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02204579147744161,
      "loss": 0.7952,
      "num_input_tokens_seen": 31961096,
      "step": 55045
    },
    {
      "epoch": 8.19928507596068,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.022044070233659674,
      "loss": 0.8096,
      "num_input_tokens_seen": 31963976,
      "step": 55050
    },
    {
      "epoch": 8.200029788501638,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.022042348870875333,
      "loss": 0.7956,
      "num_input_tokens_seen": 31966504,
      "step": 55055
    },
    {
      "epoch": 8.200774501042597,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.022040627389117672,
      "loss": 0.8189,
      "num_input_tokens_seen": 31969256,
      "step": 55060
    },
    {
      "epoch": 8.201519213583557,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.02203890578841576,
      "loss": 0.8234,
      "num_input_tokens_seen": 31972232,
      "step": 55065
    },
    {
      "epoch": 8.202263926124516,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02203718406879869,
      "loss": 0.7787,
      "num_input_tokens_seen": 31975176,
      "step": 55070
    },
    {
      "epoch": 8.203008638665475,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02203546223029555,
      "loss": 0.787,
      "num_input_tokens_seen": 31977992,
      "step": 55075
    },
    {
      "epoch": 8.203753351206434,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.022033740272935424,
      "loss": 0.7734,
      "num_input_tokens_seen": 31980680,
      "step": 55080
    },
    {
      "epoch": 8.204498063747394,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.022032018196747407,
      "loss": 0.7997,
      "num_input_tokens_seen": 31983432,
      "step": 55085
    },
    {
      "epoch": 8.205242776288353,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02203029600176059,
      "loss": 0.7758,
      "num_input_tokens_seen": 31986376,
      "step": 55090
    },
    {
      "epoch": 8.205987488829312,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.022028573688004063,
      "loss": 0.8015,
      "num_input_tokens_seen": 31989032,
      "step": 55095
    },
    {
      "epoch": 8.20673220137027,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.022026851255506928,
      "loss": 0.8171,
      "num_input_tokens_seen": 31991784,
      "step": 55100
    },
    {
      "epoch": 8.207476913911231,
      "grad_norm": 0.03125,
      "learning_rate": 0.022025128704298284,
      "loss": 0.8061,
      "num_input_tokens_seen": 31994504,
      "step": 55105
    },
    {
      "epoch": 8.20822162645219,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.022023406034407233,
      "loss": 0.8103,
      "num_input_tokens_seen": 31997576,
      "step": 55110
    },
    {
      "epoch": 8.208966338993148,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02202168324586287,
      "loss": 0.7928,
      "num_input_tokens_seen": 32000360,
      "step": 55115
    },
    {
      "epoch": 8.209711051534107,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02201996033869431,
      "loss": 0.789,
      "num_input_tokens_seen": 32003080,
      "step": 55120
    },
    {
      "epoch": 8.210455764075068,
      "grad_norm": 0.037109375,
      "learning_rate": 0.022018237312930657,
      "loss": 0.8063,
      "num_input_tokens_seen": 32006248,
      "step": 55125
    },
    {
      "epoch": 8.211200476616026,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.022016514168601013,
      "loss": 0.804,
      "num_input_tokens_seen": 32009192,
      "step": 55130
    },
    {
      "epoch": 8.211945189156985,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02201479090573449,
      "loss": 0.7917,
      "num_input_tokens_seen": 32011848,
      "step": 55135
    },
    {
      "epoch": 8.212689901697944,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.02201306752436021,
      "loss": 0.8032,
      "num_input_tokens_seen": 32015016,
      "step": 55140
    },
    {
      "epoch": 8.213434614238905,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.022011344024507277,
      "loss": 0.8176,
      "num_input_tokens_seen": 32018344,
      "step": 55145
    },
    {
      "epoch": 8.214179326779863,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.022009620406204817,
      "loss": 0.7956,
      "num_input_tokens_seen": 32021160,
      "step": 55150
    },
    {
      "epoch": 8.214924039320822,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.022007896669481945,
      "loss": 0.808,
      "num_input_tokens_seen": 32024008,
      "step": 55155
    },
    {
      "epoch": 8.21566875186178,
      "grad_norm": 0.029296875,
      "learning_rate": 0.022006172814367782,
      "loss": 0.8349,
      "num_input_tokens_seen": 32026760,
      "step": 55160
    },
    {
      "epoch": 8.216413464402741,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.022004448840891446,
      "loss": 0.7903,
      "num_input_tokens_seen": 32029928,
      "step": 55165
    },
    {
      "epoch": 8.2171581769437,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02200272474908207,
      "loss": 0.8146,
      "num_input_tokens_seen": 32032680,
      "step": 55170
    },
    {
      "epoch": 8.217902889484659,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.022001000538968772,
      "loss": 0.8051,
      "num_input_tokens_seen": 32035592,
      "step": 55175
    },
    {
      "epoch": 8.218647602025618,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.021999276210580686,
      "loss": 0.8062,
      "num_input_tokens_seen": 32038696,
      "step": 55180
    },
    {
      "epoch": 8.219392314566578,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.021997551763946945,
      "loss": 0.794,
      "num_input_tokens_seen": 32041608,
      "step": 55185
    },
    {
      "epoch": 8.220137027107537,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.02199582719909668,
      "loss": 0.8007,
      "num_input_tokens_seen": 32044616,
      "step": 55190
    },
    {
      "epoch": 8.220881739648496,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.021994102516059016,
      "loss": 0.8017,
      "num_input_tokens_seen": 32047240,
      "step": 55195
    },
    {
      "epoch": 8.221626452189454,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02199237771486311,
      "loss": 0.7942,
      "num_input_tokens_seen": 32050184,
      "step": 55200
    },
    {
      "epoch": 8.222371164730415,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.021990652795538083,
      "loss": 0.8055,
      "num_input_tokens_seen": 32052776,
      "step": 55205
    },
    {
      "epoch": 8.223115877271374,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.021988927758113085,
      "loss": 0.8042,
      "num_input_tokens_seen": 32055496,
      "step": 55210
    },
    {
      "epoch": 8.223860589812332,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.021987202602617252,
      "loss": 0.8037,
      "num_input_tokens_seen": 32058472,
      "step": 55215
    },
    {
      "epoch": 8.224605302353291,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.021985477329079732,
      "loss": 0.8083,
      "num_input_tokens_seen": 32061256,
      "step": 55220
    },
    {
      "epoch": 8.22535001489425,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.021983751937529678,
      "loss": 0.8114,
      "num_input_tokens_seen": 32064040,
      "step": 55225
    },
    {
      "epoch": 8.22609472743521,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.021982026427996227,
      "loss": 0.7915,
      "num_input_tokens_seen": 32066792,
      "step": 55230
    },
    {
      "epoch": 8.22683943997617,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02198030080050854,
      "loss": 0.7935,
      "num_input_tokens_seen": 32069832,
      "step": 55235
    },
    {
      "epoch": 8.227584152517128,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.021978575055095765,
      "loss": 0.7868,
      "num_input_tokens_seen": 32072552,
      "step": 55240
    },
    {
      "epoch": 8.228328865058087,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02197684919178705,
      "loss": 0.8066,
      "num_input_tokens_seen": 32075304,
      "step": 55245
    },
    {
      "epoch": 8.229073577599047,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.021975123210611566,
      "loss": 0.8175,
      "num_input_tokens_seen": 32078216,
      "step": 55250
    },
    {
      "epoch": 8.229818290140006,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02197339711159847,
      "loss": 0.8117,
      "num_input_tokens_seen": 32081256,
      "step": 55255
    },
    {
      "epoch": 8.230563002680965,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.021971670894776908,
      "loss": 0.8026,
      "num_input_tokens_seen": 32084488,
      "step": 55260
    },
    {
      "epoch": 8.231307715221924,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.021969944560176053,
      "loss": 0.7904,
      "num_input_tokens_seen": 32087368,
      "step": 55265
    },
    {
      "epoch": 8.232052427762884,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02196821810782507,
      "loss": 0.7964,
      "num_input_tokens_seen": 32090440,
      "step": 55270
    },
    {
      "epoch": 8.232797140303843,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02196649153775312,
      "loss": 0.8049,
      "num_input_tokens_seen": 32093160,
      "step": 55275
    },
    {
      "epoch": 8.233541852844802,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.021964764849989386,
      "loss": 0.7996,
      "num_input_tokens_seen": 32095752,
      "step": 55280
    },
    {
      "epoch": 8.23428656538576,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02196303804456302,
      "loss": 0.7736,
      "num_input_tokens_seen": 32098728,
      "step": 55285
    },
    {
      "epoch": 8.235031277926721,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.021961311121503208,
      "loss": 0.797,
      "num_input_tokens_seen": 32101352,
      "step": 55290
    },
    {
      "epoch": 8.23577599046768,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.021959584080839117,
      "loss": 0.8465,
      "num_input_tokens_seen": 32104200,
      "step": 55295
    },
    {
      "epoch": 8.236520703008638,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.02195785692259993,
      "loss": 0.7996,
      "num_input_tokens_seen": 32107048,
      "step": 55300
    },
    {
      "epoch": 8.237265415549597,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.021956129646814815,
      "loss": 0.7912,
      "num_input_tokens_seen": 32109928,
      "step": 55305
    },
    {
      "epoch": 8.238010128090558,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.021954402253512962,
      "loss": 0.7846,
      "num_input_tokens_seen": 32112808,
      "step": 55310
    },
    {
      "epoch": 8.238754840631517,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.021952674742723555,
      "loss": 0.8039,
      "num_input_tokens_seen": 32115432,
      "step": 55315
    },
    {
      "epoch": 8.239499553172475,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.021950947114475775,
      "loss": 0.8107,
      "num_input_tokens_seen": 32118376,
      "step": 55320
    },
    {
      "epoch": 8.240244265713434,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.021949219368798806,
      "loss": 0.7975,
      "num_input_tokens_seen": 32121032,
      "step": 55325
    },
    {
      "epoch": 8.240988978254395,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.021947491505721835,
      "loss": 0.82,
      "num_input_tokens_seen": 32123880,
      "step": 55330
    },
    {
      "epoch": 8.241733690795353,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02194576352527406,
      "loss": 0.784,
      "num_input_tokens_seen": 32126536,
      "step": 55335
    },
    {
      "epoch": 8.242478403336312,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.021944035427484666,
      "loss": 0.7914,
      "num_input_tokens_seen": 32129576,
      "step": 55340
    },
    {
      "epoch": 8.24322311587727,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.021942307212382852,
      "loss": 0.8265,
      "num_input_tokens_seen": 32132296,
      "step": 55345
    },
    {
      "epoch": 8.243967828418231,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.021940578879997813,
      "loss": 0.8073,
      "num_input_tokens_seen": 32135144,
      "step": 55350
    },
    {
      "epoch": 8.24471254095919,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.021938850430358755,
      "loss": 0.8085,
      "num_input_tokens_seen": 32138280,
      "step": 55355
    },
    {
      "epoch": 8.245457253500149,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.021937121863494866,
      "loss": 0.7998,
      "num_input_tokens_seen": 32141160,
      "step": 55360
    },
    {
      "epoch": 8.246201966041108,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.021935393179435353,
      "loss": 0.7977,
      "num_input_tokens_seen": 32144232,
      "step": 55365
    },
    {
      "epoch": 8.246946678582066,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02193366437820942,
      "loss": 0.8012,
      "num_input_tokens_seen": 32147272,
      "step": 55370
    },
    {
      "epoch": 8.247691391123027,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.02193193545984628,
      "loss": 0.7939,
      "num_input_tokens_seen": 32149992,
      "step": 55375
    },
    {
      "epoch": 8.248436103663986,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.021930206424375133,
      "loss": 0.7906,
      "num_input_tokens_seen": 32153032,
      "step": 55380
    },
    {
      "epoch": 8.249180816204944,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.02192847727182519,
      "loss": 0.8085,
      "num_input_tokens_seen": 32155688,
      "step": 55385
    },
    {
      "epoch": 8.249925528745903,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.021926748002225667,
      "loss": 0.7938,
      "num_input_tokens_seen": 32158504,
      "step": 55390
    },
    {
      "epoch": 8.250670241286864,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.021925018615605776,
      "loss": 0.7769,
      "num_input_tokens_seen": 32161320,
      "step": 55395
    },
    {
      "epoch": 8.251414953827823,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.021923289111994735,
      "loss": 0.8223,
      "num_input_tokens_seen": 32164200,
      "step": 55400
    },
    {
      "epoch": 8.252159666368781,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.021921559491421758,
      "loss": 0.7996,
      "num_input_tokens_seen": 32167176,
      "step": 55405
    },
    {
      "epoch": 8.25290437890974,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02191982975391607,
      "loss": 0.767,
      "num_input_tokens_seen": 32169896,
      "step": 55410
    },
    {
      "epoch": 8.2536490914507,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02191809989950689,
      "loss": 0.7971,
      "num_input_tokens_seen": 32172680,
      "step": 55415
    },
    {
      "epoch": 8.25439380399166,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.021916369928223448,
      "loss": 0.811,
      "num_input_tokens_seen": 32175816,
      "step": 55420
    },
    {
      "epoch": 8.255138516532618,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.021914639840094962,
      "loss": 0.8008,
      "num_input_tokens_seen": 32178600,
      "step": 55425
    },
    {
      "epoch": 8.255883229073577,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02191290963515066,
      "loss": 0.7921,
      "num_input_tokens_seen": 32181448,
      "step": 55430
    },
    {
      "epoch": 8.256627941614537,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02191117931341978,
      "loss": 0.7917,
      "num_input_tokens_seen": 32184104,
      "step": 55435
    },
    {
      "epoch": 8.257372654155496,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02190944887493155,
      "loss": 0.8285,
      "num_input_tokens_seen": 32187112,
      "step": 55440
    },
    {
      "epoch": 8.258117366696455,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.021907718319715205,
      "loss": 0.8237,
      "num_input_tokens_seen": 32190184,
      "step": 55445
    },
    {
      "epoch": 8.258862079237414,
      "grad_norm": 0.03515625,
      "learning_rate": 0.021905987647799975,
      "loss": 0.8108,
      "num_input_tokens_seen": 32193128,
      "step": 55450
    },
    {
      "epoch": 8.259606791778374,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.02190425685921511,
      "loss": 0.8005,
      "num_input_tokens_seen": 32196104,
      "step": 55455
    },
    {
      "epoch": 8.260351504319333,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.021902525953989834,
      "loss": 0.8234,
      "num_input_tokens_seen": 32198888,
      "step": 55460
    },
    {
      "epoch": 8.261096216860292,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.0219007949321534,
      "loss": 0.8099,
      "num_input_tokens_seen": 32201544,
      "step": 55465
    },
    {
      "epoch": 8.26184092940125,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.021899063793735053,
      "loss": 0.8058,
      "num_input_tokens_seen": 32204552,
      "step": 55470
    },
    {
      "epoch": 8.262585641942211,
      "grad_norm": 0.041015625,
      "learning_rate": 0.021897332538764032,
      "loss": 0.7906,
      "num_input_tokens_seen": 32207336,
      "step": 55475
    },
    {
      "epoch": 8.26333035448317,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.021895601167269588,
      "loss": 0.8063,
      "num_input_tokens_seen": 32210184,
      "step": 55480
    },
    {
      "epoch": 8.264075067024129,
      "grad_norm": 0.0546875,
      "learning_rate": 0.021893869679280968,
      "loss": 0.812,
      "num_input_tokens_seen": 32213064,
      "step": 55485
    },
    {
      "epoch": 8.264819779565087,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.021892138074827434,
      "loss": 0.805,
      "num_input_tokens_seen": 32216008,
      "step": 55490
    },
    {
      "epoch": 8.265564492106048,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.021890406353938227,
      "loss": 0.8065,
      "num_input_tokens_seen": 32219176,
      "step": 55495
    },
    {
      "epoch": 8.266309204647007,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.021888674516642607,
      "loss": 0.7987,
      "num_input_tokens_seen": 32222664,
      "step": 55500
    },
    {
      "epoch": 8.267053917187965,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.021886942562969833,
      "loss": 0.7905,
      "num_input_tokens_seen": 32225416,
      "step": 55505
    },
    {
      "epoch": 8.267798629728924,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.021885210492949163,
      "loss": 0.8066,
      "num_input_tokens_seen": 32228584,
      "step": 55510
    },
    {
      "epoch": 8.268543342269885,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.021883478306609864,
      "loss": 0.7969,
      "num_input_tokens_seen": 32231368,
      "step": 55515
    },
    {
      "epoch": 8.269288054810843,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.021881746003981194,
      "loss": 0.8152,
      "num_input_tokens_seen": 32234344,
      "step": 55520
    },
    {
      "epoch": 8.270032767351802,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02188001358509242,
      "loss": 0.8037,
      "num_input_tokens_seen": 32237416,
      "step": 55525
    },
    {
      "epoch": 8.270777479892761,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.021878281049972807,
      "loss": 0.7941,
      "num_input_tokens_seen": 32240584,
      "step": 55530
    },
    {
      "epoch": 8.271522192433721,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.021876548398651625,
      "loss": 0.7888,
      "num_input_tokens_seen": 32243528,
      "step": 55535
    },
    {
      "epoch": 8.27226690497468,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.02187481563115815,
      "loss": 0.8049,
      "num_input_tokens_seen": 32246472,
      "step": 55540
    },
    {
      "epoch": 8.273011617515639,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.021873082747521647,
      "loss": 0.801,
      "num_input_tokens_seen": 32249576,
      "step": 55545
    },
    {
      "epoch": 8.273756330056598,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.0218713497477714,
      "loss": 0.8226,
      "num_input_tokens_seen": 32252104,
      "step": 55550
    },
    {
      "epoch": 8.274501042597556,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.021869616631936685,
      "loss": 0.798,
      "num_input_tokens_seen": 32255144,
      "step": 55555
    },
    {
      "epoch": 8.275245755138517,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02186788340004677,
      "loss": 0.8211,
      "num_input_tokens_seen": 32257832,
      "step": 55560
    },
    {
      "epoch": 8.275990467679476,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.021866150052130957,
      "loss": 0.7979,
      "num_input_tokens_seen": 32260616,
      "step": 55565
    },
    {
      "epoch": 8.276735180220435,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02186441658821851,
      "loss": 0.7854,
      "num_input_tokens_seen": 32263240,
      "step": 55570
    },
    {
      "epoch": 8.277479892761393,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.021862683008338723,
      "loss": 0.8018,
      "num_input_tokens_seen": 32266408,
      "step": 55575
    },
    {
      "epoch": 8.278224605302354,
      "grad_norm": 0.02734375,
      "learning_rate": 0.021860949312520882,
      "loss": 0.8028,
      "num_input_tokens_seen": 32269384,
      "step": 55580
    },
    {
      "epoch": 8.278969317843313,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.02185921550079428,
      "loss": 0.8294,
      "num_input_tokens_seen": 32272392,
      "step": 55585
    },
    {
      "epoch": 8.279714030384271,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.021857481573188196,
      "loss": 0.8346,
      "num_input_tokens_seen": 32275176,
      "step": 55590
    },
    {
      "epoch": 8.28045874292523,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02185574752973193,
      "loss": 0.8182,
      "num_input_tokens_seen": 32278376,
      "step": 55595
    },
    {
      "epoch": 8.28120345546619,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02185401337045478,
      "loss": 0.8,
      "num_input_tokens_seen": 32281096,
      "step": 55600
    },
    {
      "epoch": 8.28194816800715,
      "grad_norm": 0.0130615234375,
      "learning_rate": 0.021852279095386045,
      "loss": 0.8089,
      "num_input_tokens_seen": 32283848,
      "step": 55605
    },
    {
      "epoch": 8.282692880548108,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.021850544704555015,
      "loss": 0.8036,
      "num_input_tokens_seen": 32286856,
      "step": 55610
    },
    {
      "epoch": 8.283437593089067,
      "grad_norm": 0.02734375,
      "learning_rate": 0.021848810197990995,
      "loss": 0.8038,
      "num_input_tokens_seen": 32289832,
      "step": 55615
    },
    {
      "epoch": 8.284182305630027,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02184707557572329,
      "loss": 0.8071,
      "num_input_tokens_seen": 32292616,
      "step": 55620
    },
    {
      "epoch": 8.284927018170986,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.021845340837781196,
      "loss": 0.7863,
      "num_input_tokens_seen": 32295752,
      "step": 55625
    },
    {
      "epoch": 8.285671730711945,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02184360598419403,
      "loss": 0.805,
      "num_input_tokens_seen": 32298664,
      "step": 55630
    },
    {
      "epoch": 8.286416443252904,
      "grad_norm": 0.01953125,
      "learning_rate": 0.021841871014991094,
      "loss": 0.813,
      "num_input_tokens_seen": 32301480,
      "step": 55635
    },
    {
      "epoch": 8.287161155793864,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.0218401359302017,
      "loss": 0.812,
      "num_input_tokens_seen": 32304296,
      "step": 55640
    },
    {
      "epoch": 8.287905868334823,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.021838400729855167,
      "loss": 0.7982,
      "num_input_tokens_seen": 32307336,
      "step": 55645
    },
    {
      "epoch": 8.288650580875782,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.021836665413980803,
      "loss": 0.7943,
      "num_input_tokens_seen": 32310408,
      "step": 55650
    },
    {
      "epoch": 8.28939529341674,
      "grad_norm": 0.03125,
      "learning_rate": 0.021834929982607925,
      "loss": 0.7987,
      "num_input_tokens_seen": 32313288,
      "step": 55655
    },
    {
      "epoch": 8.290140005957701,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.021833194435765847,
      "loss": 0.7963,
      "num_input_tokens_seen": 32316648,
      "step": 55660
    },
    {
      "epoch": 8.29088471849866,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.0218314587734839,
      "loss": 0.8097,
      "num_input_tokens_seen": 32319752,
      "step": 55665
    },
    {
      "epoch": 8.291629431039619,
      "grad_norm": 0.01458740234375,
      "learning_rate": 0.021829722995791397,
      "loss": 0.799,
      "num_input_tokens_seen": 32322568,
      "step": 55670
    },
    {
      "epoch": 8.292374143580577,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02182798710271766,
      "loss": 0.8134,
      "num_input_tokens_seen": 32325544,
      "step": 55675
    },
    {
      "epoch": 8.293118856121538,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.021826251094292027,
      "loss": 0.8037,
      "num_input_tokens_seen": 32328840,
      "step": 55680
    },
    {
      "epoch": 8.293863568662497,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.021824514970543822,
      "loss": 0.7973,
      "num_input_tokens_seen": 32331688,
      "step": 55685
    },
    {
      "epoch": 8.294608281203455,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02182277873150237,
      "loss": 0.816,
      "num_input_tokens_seen": 32334504,
      "step": 55690
    },
    {
      "epoch": 8.295352993744414,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.021821042377197005,
      "loss": 0.7782,
      "num_input_tokens_seen": 32337672,
      "step": 55695
    },
    {
      "epoch": 8.296097706285375,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02181930590765706,
      "loss": 0.7961,
      "num_input_tokens_seen": 32340360,
      "step": 55700
    },
    {
      "epoch": 8.296842418826333,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.021817569322911876,
      "loss": 0.8183,
      "num_input_tokens_seen": 32343400,
      "step": 55705
    },
    {
      "epoch": 8.297587131367292,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.02181583262299078,
      "loss": 0.7967,
      "num_input_tokens_seen": 32346184,
      "step": 55710
    },
    {
      "epoch": 8.298331843908251,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02181409580792313,
      "loss": 0.8039,
      "num_input_tokens_seen": 32349032,
      "step": 55715
    },
    {
      "epoch": 8.299076556449211,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02181235887773825,
      "loss": 0.7968,
      "num_input_tokens_seen": 32352264,
      "step": 55720
    },
    {
      "epoch": 8.29982126899017,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.021810621832465495,
      "loss": 0.8071,
      "num_input_tokens_seen": 32355272,
      "step": 55725
    },
    {
      "epoch": 8.300565981531129,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.021808884672134203,
      "loss": 0.7954,
      "num_input_tokens_seen": 32357992,
      "step": 55730
    },
    {
      "epoch": 8.301310694072088,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.021807147396773724,
      "loss": 0.8079,
      "num_input_tokens_seen": 32360744,
      "step": 55735
    },
    {
      "epoch": 8.302055406613047,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02180541000641341,
      "loss": 0.8044,
      "num_input_tokens_seen": 32363528,
      "step": 55740
    },
    {
      "epoch": 8.302800119154007,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02180367250108261,
      "loss": 0.7955,
      "num_input_tokens_seen": 32366344,
      "step": 55745
    },
    {
      "epoch": 8.303544831694966,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02180193488081068,
      "loss": 0.8005,
      "num_input_tokens_seen": 32369160,
      "step": 55750
    },
    {
      "epoch": 8.304289544235925,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.02180019714562697,
      "loss": 0.788,
      "num_input_tokens_seen": 32371976,
      "step": 55755
    },
    {
      "epoch": 8.305034256776883,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02179845929556084,
      "loss": 0.8077,
      "num_input_tokens_seen": 32374696,
      "step": 55760
    },
    {
      "epoch": 8.305778969317844,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.021796721330641654,
      "loss": 0.8091,
      "num_input_tokens_seen": 32377448,
      "step": 55765
    },
    {
      "epoch": 8.306523681858803,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.021794983250898767,
      "loss": 0.7989,
      "num_input_tokens_seen": 32380264,
      "step": 55770
    },
    {
      "epoch": 8.307268394399761,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.021793245056361543,
      "loss": 0.8018,
      "num_input_tokens_seen": 32383208,
      "step": 55775
    },
    {
      "epoch": 8.30801310694072,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.021791506747059347,
      "loss": 0.8082,
      "num_input_tokens_seen": 32385960,
      "step": 55780
    },
    {
      "epoch": 8.30875781948168,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02178976832302155,
      "loss": 0.7935,
      "num_input_tokens_seen": 32388680,
      "step": 55785
    },
    {
      "epoch": 8.30950253202264,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02178802978427752,
      "loss": 0.8009,
      "num_input_tokens_seen": 32391336,
      "step": 55790
    },
    {
      "epoch": 8.310247244563598,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02178629113085662,
      "loss": 0.7949,
      "num_input_tokens_seen": 32394344,
      "step": 55795
    },
    {
      "epoch": 8.310991957104557,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.021784552362788232,
      "loss": 0.8003,
      "num_input_tokens_seen": 32397288,
      "step": 55800
    },
    {
      "epoch": 8.311736669645517,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.021782813480101727,
      "loss": 0.808,
      "num_input_tokens_seen": 32400040,
      "step": 55805
    },
    {
      "epoch": 8.312481382186476,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.021781074482826485,
      "loss": 0.8008,
      "num_input_tokens_seen": 32403016,
      "step": 55810
    },
    {
      "epoch": 8.313226094727435,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.021779335370991876,
      "loss": 0.7972,
      "num_input_tokens_seen": 32405992,
      "step": 55815
    },
    {
      "epoch": 8.313970807268394,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.021777596144627293,
      "loss": 0.7976,
      "num_input_tokens_seen": 32408904,
      "step": 55820
    },
    {
      "epoch": 8.314715519809354,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.021775856803762105,
      "loss": 0.7788,
      "num_input_tokens_seen": 32411720,
      "step": 55825
    },
    {
      "epoch": 8.315460232350313,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.021774117348425704,
      "loss": 0.7969,
      "num_input_tokens_seen": 32415016,
      "step": 55830
    },
    {
      "epoch": 8.316204944891272,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02177237777864748,
      "loss": 0.8076,
      "num_input_tokens_seen": 32417896,
      "step": 55835
    },
    {
      "epoch": 8.31694965743223,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.021770638094456805,
      "loss": 0.8046,
      "num_input_tokens_seen": 32420936,
      "step": 55840
    },
    {
      "epoch": 8.317694369973191,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.021768898295883094,
      "loss": 0.7844,
      "num_input_tokens_seen": 32423784,
      "step": 55845
    },
    {
      "epoch": 8.31843908251415,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.02176715838295572,
      "loss": 0.8069,
      "num_input_tokens_seen": 32426536,
      "step": 55850
    },
    {
      "epoch": 8.319183795055109,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.021765418355704086,
      "loss": 0.7674,
      "num_input_tokens_seen": 32429512,
      "step": 55855
    },
    {
      "epoch": 8.319928507596067,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.021763678214157582,
      "loss": 0.8005,
      "num_input_tokens_seen": 32432456,
      "step": 55860
    },
    {
      "epoch": 8.320673220137028,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.021761937958345606,
      "loss": 0.8128,
      "num_input_tokens_seen": 32435656,
      "step": 55865
    },
    {
      "epoch": 8.321417932677987,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.021760197588297565,
      "loss": 0.7994,
      "num_input_tokens_seen": 32438472,
      "step": 55870
    },
    {
      "epoch": 8.322162645218945,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.021758457104042856,
      "loss": 0.8061,
      "num_input_tokens_seen": 32441416,
      "step": 55875
    },
    {
      "epoch": 8.322907357759904,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.021756716505610882,
      "loss": 0.7852,
      "num_input_tokens_seen": 32443976,
      "step": 55880
    },
    {
      "epoch": 8.323652070300863,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.021754975793031053,
      "loss": 0.8643,
      "num_input_tokens_seen": 32446760,
      "step": 55885
    },
    {
      "epoch": 8.324396782841823,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02175323496633277,
      "loss": 0.8124,
      "num_input_tokens_seen": 32449640,
      "step": 55890
    },
    {
      "epoch": 8.325141495382782,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.021751494025545447,
      "loss": 0.795,
      "num_input_tokens_seen": 32452488,
      "step": 55895
    },
    {
      "epoch": 8.325886207923741,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.021749752970698494,
      "loss": 0.8106,
      "num_input_tokens_seen": 32455336,
      "step": 55900
    },
    {
      "epoch": 8.3266309204647,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.021748011801821326,
      "loss": 0.7842,
      "num_input_tokens_seen": 32457864,
      "step": 55905
    },
    {
      "epoch": 8.32737563300566,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.021746270518943355,
      "loss": 0.8189,
      "num_input_tokens_seen": 32460808,
      "step": 55910
    },
    {
      "epoch": 8.328120345546619,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.021744529122094007,
      "loss": 0.7764,
      "num_input_tokens_seen": 32463912,
      "step": 55915
    },
    {
      "epoch": 8.328865058087578,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.021742787611302685,
      "loss": 0.8103,
      "num_input_tokens_seen": 32467080,
      "step": 55920
    },
    {
      "epoch": 8.329609770628537,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.021741045986598824,
      "loss": 0.776,
      "num_input_tokens_seen": 32470152,
      "step": 55925
    },
    {
      "epoch": 8.330354483169497,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.02173930424801184,
      "loss": 0.8279,
      "num_input_tokens_seen": 32473000,
      "step": 55930
    },
    {
      "epoch": 8.331099195710456,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.021737562395571165,
      "loss": 0.8201,
      "num_input_tokens_seen": 32475688,
      "step": 55935
    },
    {
      "epoch": 8.331843908251415,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.021735820429306215,
      "loss": 0.8083,
      "num_input_tokens_seen": 32478600,
      "step": 55940
    },
    {
      "epoch": 8.332588620792373,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.021734078349246427,
      "loss": 0.7746,
      "num_input_tokens_seen": 32481736,
      "step": 55945
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.021732336155421233,
      "loss": 0.7822,
      "num_input_tokens_seen": 32484744,
      "step": 55950
    },
    {
      "epoch": 8.334078045874293,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.021730593847860056,
      "loss": 0.8031,
      "num_input_tokens_seen": 32487688,
      "step": 55955
    },
    {
      "epoch": 8.334822758415251,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.02172885142659234,
      "loss": 0.7951,
      "num_input_tokens_seen": 32490280,
      "step": 55960
    },
    {
      "epoch": 8.33556747095621,
      "grad_norm": 0.041015625,
      "learning_rate": 0.021727108891647518,
      "loss": 0.8161,
      "num_input_tokens_seen": 32493192,
      "step": 55965
    },
    {
      "epoch": 8.33631218349717,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.021725366243055027,
      "loss": 0.7965,
      "num_input_tokens_seen": 32496040,
      "step": 55970
    },
    {
      "epoch": 8.33705689603813,
      "grad_norm": 0.03125,
      "learning_rate": 0.02172362348084431,
      "loss": 0.8028,
      "num_input_tokens_seen": 32499080,
      "step": 55975
    },
    {
      "epoch": 8.337801608579088,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.02172188060504481,
      "loss": 0.7923,
      "num_input_tokens_seen": 32501768,
      "step": 55980
    },
    {
      "epoch": 8.338546321120047,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.021720137615685968,
      "loss": 0.8052,
      "num_input_tokens_seen": 32504456,
      "step": 55985
    },
    {
      "epoch": 8.339291033661008,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.02171839451279723,
      "loss": 0.8016,
      "num_input_tokens_seen": 32507496,
      "step": 55990
    },
    {
      "epoch": 8.340035746201966,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.021716651296408044,
      "loss": 0.7907,
      "num_input_tokens_seen": 32510408,
      "step": 55995
    },
    {
      "epoch": 8.340780458742925,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02171490796654786,
      "loss": 0.8154,
      "num_input_tokens_seen": 32513416,
      "step": 56000
    },
    {
      "epoch": 8.341525171283884,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.021713164523246134,
      "loss": 0.8363,
      "num_input_tokens_seen": 32516232,
      "step": 56005
    },
    {
      "epoch": 8.342269883824844,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02171142096653231,
      "loss": 0.8329,
      "num_input_tokens_seen": 32519240,
      "step": 56010
    },
    {
      "epoch": 8.343014596365803,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.021709677296435855,
      "loss": 0.7919,
      "num_input_tokens_seen": 32521896,
      "step": 56015
    },
    {
      "epoch": 8.343759308906762,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.021707933512986225,
      "loss": 0.7979,
      "num_input_tokens_seen": 32524968,
      "step": 56020
    },
    {
      "epoch": 8.34450402144772,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02170618961621287,
      "loss": 0.8214,
      "num_input_tokens_seen": 32527848,
      "step": 56025
    },
    {
      "epoch": 8.345248733988681,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02170444560614526,
      "loss": 0.8246,
      "num_input_tokens_seen": 32530760,
      "step": 56030
    },
    {
      "epoch": 8.34599344652964,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.021702701482812854,
      "loss": 0.8013,
      "num_input_tokens_seen": 32533512,
      "step": 56035
    },
    {
      "epoch": 8.346738159070599,
      "grad_norm": 0.0234375,
      "learning_rate": 0.021700957246245118,
      "loss": 0.7998,
      "num_input_tokens_seen": 32536168,
      "step": 56040
    },
    {
      "epoch": 8.347482871611557,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.021699212896471524,
      "loss": 0.8072,
      "num_input_tokens_seen": 32539304,
      "step": 56045
    },
    {
      "epoch": 8.348227584152518,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.021697468433521532,
      "loss": 0.8157,
      "num_input_tokens_seen": 32542408,
      "step": 56050
    },
    {
      "epoch": 8.348972296693477,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.021695723857424624,
      "loss": 0.8153,
      "num_input_tokens_seen": 32545448,
      "step": 56055
    },
    {
      "epoch": 8.349717009234435,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.02169397916821026,
      "loss": 0.8029,
      "num_input_tokens_seen": 32548072,
      "step": 56060
    },
    {
      "epoch": 8.350461721775394,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.021692234365907927,
      "loss": 0.8091,
      "num_input_tokens_seen": 32550728,
      "step": 56065
    },
    {
      "epoch": 8.351206434316353,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.021690489450547093,
      "loss": 0.8069,
      "num_input_tokens_seen": 32553896,
      "step": 56070
    },
    {
      "epoch": 8.351951146857314,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.02168874442215724,
      "loss": 0.802,
      "num_input_tokens_seen": 32556520,
      "step": 56075
    },
    {
      "epoch": 8.352695859398272,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.02168699928076785,
      "loss": 0.7965,
      "num_input_tokens_seen": 32559368,
      "step": 56080
    },
    {
      "epoch": 8.353440571939231,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.021685254026408397,
      "loss": 0.7897,
      "num_input_tokens_seen": 32561992,
      "step": 56085
    },
    {
      "epoch": 8.35418528448019,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.021683508659108383,
      "loss": 0.812,
      "num_input_tokens_seen": 32564680,
      "step": 56090
    },
    {
      "epoch": 8.35492999702115,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.021681763178897275,
      "loss": 0.7863,
      "num_input_tokens_seen": 32567528,
      "step": 56095
    },
    {
      "epoch": 8.35567470956211,
      "grad_norm": 0.017578125,
      "learning_rate": 0.021680017585804567,
      "loss": 0.7992,
      "num_input_tokens_seen": 32570184,
      "step": 56100
    },
    {
      "epoch": 8.356419422103068,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.021678271879859757,
      "loss": 0.7944,
      "num_input_tokens_seen": 32573000,
      "step": 56105
    },
    {
      "epoch": 8.357164134644027,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.021676526061092324,
      "loss": 0.8,
      "num_input_tokens_seen": 32575976,
      "step": 56110
    },
    {
      "epoch": 8.357908847184987,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.02167478012953178,
      "loss": 0.8168,
      "num_input_tokens_seen": 32578856,
      "step": 56115
    },
    {
      "epoch": 8.358653559725946,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0216730340852076,
      "loss": 0.8135,
      "num_input_tokens_seen": 32581608,
      "step": 56120
    },
    {
      "epoch": 8.359398272266905,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.021671287928149294,
      "loss": 0.7997,
      "num_input_tokens_seen": 32584456,
      "step": 56125
    },
    {
      "epoch": 8.360142984807863,
      "grad_norm": 0.021484375,
      "learning_rate": 0.02166954165838636,
      "loss": 0.7783,
      "num_input_tokens_seen": 32587336,
      "step": 56130
    },
    {
      "epoch": 8.360887697348824,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0216677952759483,
      "loss": 0.8078,
      "num_input_tokens_seen": 32590664,
      "step": 56135
    },
    {
      "epoch": 8.361632409889783,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.021666048780864613,
      "loss": 0.7983,
      "num_input_tokens_seen": 32593416,
      "step": 56140
    },
    {
      "epoch": 8.362377122430741,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.021664302173164805,
      "loss": 0.7923,
      "num_input_tokens_seen": 32596296,
      "step": 56145
    },
    {
      "epoch": 8.3631218349717,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02166255545287839,
      "loss": 0.8112,
      "num_input_tokens_seen": 32599592,
      "step": 56150
    },
    {
      "epoch": 8.36386654751266,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02166080862003487,
      "loss": 0.8071,
      "num_input_tokens_seen": 32602408,
      "step": 56155
    },
    {
      "epoch": 8.36461126005362,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.021659061674663756,
      "loss": 0.8118,
      "num_input_tokens_seen": 32605128,
      "step": 56160
    },
    {
      "epoch": 8.365355972594578,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.021657314616794564,
      "loss": 0.8053,
      "num_input_tokens_seen": 32608040,
      "step": 56165
    },
    {
      "epoch": 8.366100685135537,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.021655567446456813,
      "loss": 0.8074,
      "num_input_tokens_seen": 32610984,
      "step": 56170
    },
    {
      "epoch": 8.366845397676498,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.021653820163680008,
      "loss": 0.8023,
      "num_input_tokens_seen": 32613992,
      "step": 56175
    },
    {
      "epoch": 8.367590110217456,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.021652072768493673,
      "loss": 0.8174,
      "num_input_tokens_seen": 32616872,
      "step": 56180
    },
    {
      "epoch": 8.368334822758415,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.021650325260927333,
      "loss": 0.8227,
      "num_input_tokens_seen": 32619464,
      "step": 56185
    },
    {
      "epoch": 8.369079535299374,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.021648577641010507,
      "loss": 0.8005,
      "num_input_tokens_seen": 32622600,
      "step": 56190
    },
    {
      "epoch": 8.369824247840334,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.021646829908772716,
      "loss": 0.8108,
      "num_input_tokens_seen": 32625320,
      "step": 56195
    },
    {
      "epoch": 8.370568960381293,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.021645082064243495,
      "loss": 0.8017,
      "num_input_tokens_seen": 32628168,
      "step": 56200
    },
    {
      "epoch": 8.371313672922252,
      "grad_norm": 0.037109375,
      "learning_rate": 0.02164333410745236,
      "loss": 0.7982,
      "num_input_tokens_seen": 32630984,
      "step": 56205
    },
    {
      "epoch": 8.37205838546321,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02164158603842885,
      "loss": 0.8024,
      "num_input_tokens_seen": 32633800,
      "step": 56210
    },
    {
      "epoch": 8.372803098004171,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.021639837857202495,
      "loss": 0.8014,
      "num_input_tokens_seen": 32637000,
      "step": 56215
    },
    {
      "epoch": 8.37354781054513,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.02163808956380283,
      "loss": 0.7952,
      "num_input_tokens_seen": 32639624,
      "step": 56220
    },
    {
      "epoch": 8.374292523086089,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.021636341158259385,
      "loss": 0.8003,
      "num_input_tokens_seen": 32642632,
      "step": 56225
    },
    {
      "epoch": 8.375037235627047,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.021634592640601698,
      "loss": 0.8155,
      "num_input_tokens_seen": 32645640,
      "step": 56230
    },
    {
      "epoch": 8.375781948168008,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.021632844010859315,
      "loss": 0.7904,
      "num_input_tokens_seen": 32648520,
      "step": 56235
    },
    {
      "epoch": 8.376526660708967,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.021631095269061772,
      "loss": 0.7832,
      "num_input_tokens_seen": 32651688,
      "step": 56240
    },
    {
      "epoch": 8.377271373249926,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.021629346415238615,
      "loss": 0.7822,
      "num_input_tokens_seen": 32654504,
      "step": 56245
    },
    {
      "epoch": 8.378016085790884,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.021627597449419392,
      "loss": 0.7854,
      "num_input_tokens_seen": 32657416,
      "step": 56250
    },
    {
      "epoch": 8.378760798331843,
      "grad_norm": 0.025390625,
      "learning_rate": 0.021625848371633644,
      "loss": 0.7978,
      "num_input_tokens_seen": 32660072,
      "step": 56255
    },
    {
      "epoch": 8.379505510872804,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.02162409918191092,
      "loss": 0.8226,
      "num_input_tokens_seen": 32663240,
      "step": 56260
    },
    {
      "epoch": 8.380250223413762,
      "grad_norm": 0.01544189453125,
      "learning_rate": 0.02162234988028077,
      "loss": 0.8236,
      "num_input_tokens_seen": 32666024,
      "step": 56265
    },
    {
      "epoch": 8.380994935954721,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.021620600466772753,
      "loss": 0.7793,
      "num_input_tokens_seen": 32668808,
      "step": 56270
    },
    {
      "epoch": 8.38173964849568,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02161885094141642,
      "loss": 0.7977,
      "num_input_tokens_seen": 32671816,
      "step": 56275
    },
    {
      "epoch": 8.38248436103664,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02161710130424133,
      "loss": 0.8023,
      "num_input_tokens_seen": 32674376,
      "step": 56280
    },
    {
      "epoch": 8.3832290735776,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.021615351555277034,
      "loss": 0.8212,
      "num_input_tokens_seen": 32677192,
      "step": 56285
    },
    {
      "epoch": 8.383973786118558,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.021613601694553102,
      "loss": 0.7889,
      "num_input_tokens_seen": 32680104,
      "step": 56290
    },
    {
      "epoch": 8.384718498659517,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.021611851722099087,
      "loss": 0.8324,
      "num_input_tokens_seen": 32683016,
      "step": 56295
    },
    {
      "epoch": 8.385463211200477,
      "grad_norm": 0.021484375,
      "learning_rate": 0.021610101637944558,
      "loss": 0.7868,
      "num_input_tokens_seen": 32685800,
      "step": 56300
    },
    {
      "epoch": 8.386207923741436,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02160835144211908,
      "loss": 0.8118,
      "num_input_tokens_seen": 32688808,
      "step": 56305
    },
    {
      "epoch": 8.386952636282395,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02160660113465222,
      "loss": 0.8081,
      "num_input_tokens_seen": 32691752,
      "step": 56310
    },
    {
      "epoch": 8.387697348823353,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02160485071557355,
      "loss": 0.8126,
      "num_input_tokens_seen": 32695272,
      "step": 56315
    },
    {
      "epoch": 8.388442061364314,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02160310018491264,
      "loss": 0.7955,
      "num_input_tokens_seen": 32698312,
      "step": 56320
    },
    {
      "epoch": 8.389186773905273,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.021601349542699065,
      "loss": 0.7967,
      "num_input_tokens_seen": 32701224,
      "step": 56325
    },
    {
      "epoch": 8.389931486446232,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.021599598788962397,
      "loss": 0.8133,
      "num_input_tokens_seen": 32704072,
      "step": 56330
    },
    {
      "epoch": 8.39067619898719,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.021597847923732212,
      "loss": 0.8115,
      "num_input_tokens_seen": 32706856,
      "step": 56335
    },
    {
      "epoch": 8.39142091152815,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.021596096947038096,
      "loss": 0.7972,
      "num_input_tokens_seen": 32709640,
      "step": 56340
    },
    {
      "epoch": 8.39216562406911,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.021594345858909625,
      "loss": 0.8165,
      "num_input_tokens_seen": 32712392,
      "step": 56345
    },
    {
      "epoch": 8.392910336610068,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.021592594659376383,
      "loss": 0.7958,
      "num_input_tokens_seen": 32715176,
      "step": 56350
    },
    {
      "epoch": 8.393655049151027,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.021590843348467954,
      "loss": 0.8139,
      "num_input_tokens_seen": 32717896,
      "step": 56355
    },
    {
      "epoch": 8.394399761691988,
      "grad_norm": 0.033203125,
      "learning_rate": 0.021589091926213926,
      "loss": 0.8097,
      "num_input_tokens_seen": 32720424,
      "step": 56360
    },
    {
      "epoch": 8.395144474232946,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.021587340392643886,
      "loss": 0.8159,
      "num_input_tokens_seen": 32723400,
      "step": 56365
    },
    {
      "epoch": 8.395889186773905,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02158558874778743,
      "loss": 0.8076,
      "num_input_tokens_seen": 32726536,
      "step": 56370
    },
    {
      "epoch": 8.396633899314864,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.02158383699167414,
      "loss": 0.8136,
      "num_input_tokens_seen": 32729352,
      "step": 56375
    },
    {
      "epoch": 8.397378611855824,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.021582085124333617,
      "loss": 0.8096,
      "num_input_tokens_seen": 32732264,
      "step": 56380
    },
    {
      "epoch": 8.398123324396783,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.021580333145795456,
      "loss": 0.8184,
      "num_input_tokens_seen": 32735272,
      "step": 56385
    },
    {
      "epoch": 8.398868036937742,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02157858105608925,
      "loss": 0.8006,
      "num_input_tokens_seen": 32737992,
      "step": 56390
    },
    {
      "epoch": 8.3996127494787,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.021576828855244607,
      "loss": 0.8129,
      "num_input_tokens_seen": 32740808,
      "step": 56395
    },
    {
      "epoch": 8.400357462019661,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.021575076543291126,
      "loss": 0.792,
      "num_input_tokens_seen": 32743848,
      "step": 56400
    },
    {
      "epoch": 8.40110217456062,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.02157332412025841,
      "loss": 0.8035,
      "num_input_tokens_seen": 32746952,
      "step": 56405
    },
    {
      "epoch": 8.401846887101579,
      "grad_norm": 0.025390625,
      "learning_rate": 0.021571571586176057,
      "loss": 0.803,
      "num_input_tokens_seen": 32750216,
      "step": 56410
    },
    {
      "epoch": 8.402591599642538,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.02156981894107369,
      "loss": 0.7847,
      "num_input_tokens_seen": 32753480,
      "step": 56415
    },
    {
      "epoch": 8.403336312183498,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.021568066184980906,
      "loss": 0.8395,
      "num_input_tokens_seen": 32756456,
      "step": 56420
    },
    {
      "epoch": 8.404081024724457,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02156631331792732,
      "loss": 0.7999,
      "num_input_tokens_seen": 32759080,
      "step": 56425
    },
    {
      "epoch": 8.404825737265416,
      "grad_norm": 0.041015625,
      "learning_rate": 0.021564560339942543,
      "loss": 0.8135,
      "num_input_tokens_seen": 32762184,
      "step": 56430
    },
    {
      "epoch": 8.405570449806374,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.021562807251056186,
      "loss": 0.7815,
      "num_input_tokens_seen": 32765096,
      "step": 56435
    },
    {
      "epoch": 8.406315162347333,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.021561054051297875,
      "loss": 0.819,
      "num_input_tokens_seen": 32767912,
      "step": 56440
    },
    {
      "epoch": 8.407059874888294,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.02155930074069723,
      "loss": 0.8094,
      "num_input_tokens_seen": 32770696,
      "step": 56445
    },
    {
      "epoch": 8.407804587429252,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02155754731928386,
      "loss": 0.8043,
      "num_input_tokens_seen": 32773768,
      "step": 56450
    },
    {
      "epoch": 8.408549299970211,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.021555793787087392,
      "loss": 0.7925,
      "num_input_tokens_seen": 32776840,
      "step": 56455
    },
    {
      "epoch": 8.40929401251117,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.021554040144137452,
      "loss": 0.8188,
      "num_input_tokens_seen": 32779752,
      "step": 56460
    },
    {
      "epoch": 8.41003872505213,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.021552286390463665,
      "loss": 0.8066,
      "num_input_tokens_seen": 32782568,
      "step": 56465
    },
    {
      "epoch": 8.41078343759309,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.021550532526095657,
      "loss": 0.7847,
      "num_input_tokens_seen": 32785736,
      "step": 56470
    },
    {
      "epoch": 8.411528150134048,
      "grad_norm": 0.015625,
      "learning_rate": 0.021548778551063064,
      "loss": 0.8092,
      "num_input_tokens_seen": 32788744,
      "step": 56475
    },
    {
      "epoch": 8.412272862675007,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.021547024465395512,
      "loss": 0.7963,
      "num_input_tokens_seen": 32791464,
      "step": 56480
    },
    {
      "epoch": 8.413017575215967,
      "grad_norm": 0.033203125,
      "learning_rate": 0.021545270269122632,
      "loss": 0.8049,
      "num_input_tokens_seen": 32794440,
      "step": 56485
    },
    {
      "epoch": 8.413762287756926,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02154351596227407,
      "loss": 0.7966,
      "num_input_tokens_seen": 32797288,
      "step": 56490
    },
    {
      "epoch": 8.414507000297885,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02154176154487945,
      "loss": 0.7977,
      "num_input_tokens_seen": 32800296,
      "step": 56495
    },
    {
      "epoch": 8.415251712838844,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02154000701696842,
      "loss": 0.8011,
      "num_input_tokens_seen": 32802984,
      "step": 56500
    },
    {
      "epoch": 8.415996425379804,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.021538252378570613,
      "loss": 0.8095,
      "num_input_tokens_seen": 32805960,
      "step": 56505
    },
    {
      "epoch": 8.416741137920763,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.021536497629715684,
      "loss": 0.7932,
      "num_input_tokens_seen": 32808808,
      "step": 56510
    },
    {
      "epoch": 8.417485850461722,
      "grad_norm": 0.04296875,
      "learning_rate": 0.021534742770433272,
      "loss": 0.8226,
      "num_input_tokens_seen": 32811560,
      "step": 56515
    },
    {
      "epoch": 8.41823056300268,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.021532987800753016,
      "loss": 0.8061,
      "num_input_tokens_seen": 32814632,
      "step": 56520
    },
    {
      "epoch": 8.418975275543641,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.021531232720704573,
      "loss": 0.8025,
      "num_input_tokens_seen": 32817608,
      "step": 56525
    },
    {
      "epoch": 8.4197199880846,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.021529477530317594,
      "loss": 0.7907,
      "num_input_tokens_seen": 32820488,
      "step": 56530
    },
    {
      "epoch": 8.420464700625558,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.021527722229621727,
      "loss": 0.8165,
      "num_input_tokens_seen": 32823432,
      "step": 56535
    },
    {
      "epoch": 8.421209413166517,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.021525966818646627,
      "loss": 0.8156,
      "num_input_tokens_seen": 32826344,
      "step": 56540
    },
    {
      "epoch": 8.421954125707478,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.021524211297421953,
      "loss": 0.826,
      "num_input_tokens_seen": 32829192,
      "step": 56545
    },
    {
      "epoch": 8.422698838248436,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.021522455665977356,
      "loss": 0.798,
      "num_input_tokens_seen": 32832040,
      "step": 56550
    },
    {
      "epoch": 8.423443550789395,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.021520699924342506,
      "loss": 0.8052,
      "num_input_tokens_seen": 32835368,
      "step": 56555
    },
    {
      "epoch": 8.424188263330354,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.021518944072547053,
      "loss": 0.7914,
      "num_input_tokens_seen": 32838312,
      "step": 56560
    },
    {
      "epoch": 8.424932975871315,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.02151718811062066,
      "loss": 0.8081,
      "num_input_tokens_seen": 32841096,
      "step": 56565
    },
    {
      "epoch": 8.425677688412273,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.021515432038593005,
      "loss": 0.7908,
      "num_input_tokens_seen": 32844040,
      "step": 56570
    },
    {
      "epoch": 8.426422400953232,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.021513675856493747,
      "loss": 0.8038,
      "num_input_tokens_seen": 32846696,
      "step": 56575
    },
    {
      "epoch": 8.42716711349419,
      "grad_norm": 0.03125,
      "learning_rate": 0.021511919564352555,
      "loss": 0.7974,
      "num_input_tokens_seen": 32849544,
      "step": 56580
    },
    {
      "epoch": 8.42791182603515,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.021510163162199104,
      "loss": 0.8015,
      "num_input_tokens_seen": 32852648,
      "step": 56585
    },
    {
      "epoch": 8.42865653857611,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02150840665006306,
      "loss": 0.7959,
      "num_input_tokens_seen": 32856072,
      "step": 56590
    },
    {
      "epoch": 8.429401251117069,
      "grad_norm": 0.037109375,
      "learning_rate": 0.021506650027974097,
      "loss": 0.8264,
      "num_input_tokens_seen": 32858920,
      "step": 56595
    },
    {
      "epoch": 8.430145963658028,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.021504893295961898,
      "loss": 0.8222,
      "num_input_tokens_seen": 32861576,
      "step": 56600
    },
    {
      "epoch": 8.430890676198986,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.021503136454056136,
      "loss": 0.7894,
      "num_input_tokens_seen": 32864584,
      "step": 56605
    },
    {
      "epoch": 8.431635388739947,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.021501379502286498,
      "loss": 0.7916,
      "num_input_tokens_seen": 32867400,
      "step": 56610
    },
    {
      "epoch": 8.432380101280906,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02149962244068266,
      "loss": 0.7952,
      "num_input_tokens_seen": 32870344,
      "step": 56615
    },
    {
      "epoch": 8.433124813821864,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.021497865269274302,
      "loss": 0.8221,
      "num_input_tokens_seen": 32873320,
      "step": 56620
    },
    {
      "epoch": 8.433869526362823,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.02149610798809112,
      "loss": 0.8157,
      "num_input_tokens_seen": 32875976,
      "step": 56625
    },
    {
      "epoch": 8.434614238903784,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02149435059716279,
      "loss": 0.8006,
      "num_input_tokens_seen": 32878824,
      "step": 56630
    },
    {
      "epoch": 8.435358951444742,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.021492593096519012,
      "loss": 0.8093,
      "num_input_tokens_seen": 32881608,
      "step": 56635
    },
    {
      "epoch": 8.436103663985701,
      "grad_norm": 0.037109375,
      "learning_rate": 0.02149083548618947,
      "loss": 0.7975,
      "num_input_tokens_seen": 32884456,
      "step": 56640
    },
    {
      "epoch": 8.43684837652666,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02148907776620386,
      "loss": 0.8214,
      "num_input_tokens_seen": 32887336,
      "step": 56645
    },
    {
      "epoch": 8.43759308906762,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.021487319936591878,
      "loss": 0.7898,
      "num_input_tokens_seen": 32890184,
      "step": 56650
    },
    {
      "epoch": 8.43833780160858,
      "grad_norm": 0.041015625,
      "learning_rate": 0.021485561997383218,
      "loss": 0.7867,
      "num_input_tokens_seen": 32893128,
      "step": 56655
    },
    {
      "epoch": 8.439082514149538,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.021483803948607577,
      "loss": 0.8056,
      "num_input_tokens_seen": 32896136,
      "step": 56660
    },
    {
      "epoch": 8.439827226690497,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02148204579029466,
      "loss": 0.7964,
      "num_input_tokens_seen": 32898856,
      "step": 56665
    },
    {
      "epoch": 8.440571939231457,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02148028752247417,
      "loss": 0.7925,
      "num_input_tokens_seen": 32901672,
      "step": 56670
    },
    {
      "epoch": 8.441316651772416,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.021478529145175805,
      "loss": 0.8171,
      "num_input_tokens_seen": 32904680,
      "step": 56675
    },
    {
      "epoch": 8.442061364313375,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02147677065842928,
      "loss": 0.8143,
      "num_input_tokens_seen": 32907624,
      "step": 56680
    },
    {
      "epoch": 8.442806076854334,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02147501206226429,
      "loss": 0.8182,
      "num_input_tokens_seen": 32910632,
      "step": 56685
    },
    {
      "epoch": 8.443550789395294,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.021473253356710555,
      "loss": 0.7932,
      "num_input_tokens_seen": 32913608,
      "step": 56690
    },
    {
      "epoch": 8.444295501936253,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.021471494541797784,
      "loss": 0.8136,
      "num_input_tokens_seen": 32916392,
      "step": 56695
    },
    {
      "epoch": 8.445040214477212,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.021469735617555687,
      "loss": 0.8079,
      "num_input_tokens_seen": 32919528,
      "step": 56700
    },
    {
      "epoch": 8.44578492701817,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02146797658401399,
      "loss": 0.7835,
      "num_input_tokens_seen": 32922120,
      "step": 56705
    },
    {
      "epoch": 8.446529639559131,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.021466217441202397,
      "loss": 0.7785,
      "num_input_tokens_seen": 32925000,
      "step": 56710
    },
    {
      "epoch": 8.44727435210009,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.021464458189150635,
      "loss": 0.8101,
      "num_input_tokens_seen": 32928104,
      "step": 56715
    },
    {
      "epoch": 8.448019064641048,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.021462698827888416,
      "loss": 0.8159,
      "num_input_tokens_seen": 32931016,
      "step": 56720
    },
    {
      "epoch": 8.448763777182007,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.021460939357445474,
      "loss": 0.8171,
      "num_input_tokens_seen": 32934344,
      "step": 56725
    },
    {
      "epoch": 8.449508489722968,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.021459179777851525,
      "loss": 0.7975,
      "num_input_tokens_seen": 32936904,
      "step": 56730
    },
    {
      "epoch": 8.450253202263927,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.021457420089136304,
      "loss": 0.792,
      "num_input_tokens_seen": 32939720,
      "step": 56735
    },
    {
      "epoch": 8.450997914804885,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02145566029132953,
      "loss": 0.7934,
      "num_input_tokens_seen": 32942760,
      "step": 56740
    },
    {
      "epoch": 8.451742627345844,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.021453900384460937,
      "loss": 0.7912,
      "num_input_tokens_seen": 32945608,
      "step": 56745
    },
    {
      "epoch": 8.452487339886805,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.021452140368560257,
      "loss": 0.8017,
      "num_input_tokens_seen": 32948776,
      "step": 56750
    },
    {
      "epoch": 8.453232052427763,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02145038024365722,
      "loss": 0.817,
      "num_input_tokens_seen": 32951720,
      "step": 56755
    },
    {
      "epoch": 8.453976764968722,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.021448620009781566,
      "loss": 0.7925,
      "num_input_tokens_seen": 32954664,
      "step": 56760
    },
    {
      "epoch": 8.45472147750968,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.021446859666963036,
      "loss": 0.7937,
      "num_input_tokens_seen": 32957576,
      "step": 56765
    },
    {
      "epoch": 8.45546619005064,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.021445099215231356,
      "loss": 0.8054,
      "num_input_tokens_seen": 32960680,
      "step": 56770
    },
    {
      "epoch": 8.4562109025916,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02144333865461628,
      "loss": 0.8058,
      "num_input_tokens_seen": 32963528,
      "step": 56775
    },
    {
      "epoch": 8.456955615132559,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.021441577985147554,
      "loss": 0.7978,
      "num_input_tokens_seen": 32966344,
      "step": 56780
    },
    {
      "epoch": 8.457700327673518,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0214398172068549,
      "loss": 0.789,
      "num_input_tokens_seen": 32969512,
      "step": 56785
    },
    {
      "epoch": 8.458445040214476,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.021438056319768087,
      "loss": 0.7857,
      "num_input_tokens_seen": 32972168,
      "step": 56790
    },
    {
      "epoch": 8.459189752755437,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.021436295323916853,
      "loss": 0.8431,
      "num_input_tokens_seen": 32975048,
      "step": 56795
    },
    {
      "epoch": 8.459934465296396,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.021434534219330954,
      "loss": 0.8027,
      "num_input_tokens_seen": 32977864,
      "step": 56800
    },
    {
      "epoch": 8.460679177837354,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02143277300604014,
      "loss": 0.7968,
      "num_input_tokens_seen": 32980712,
      "step": 56805
    },
    {
      "epoch": 8.461423890378313,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.02143101168407416,
      "loss": 0.8355,
      "num_input_tokens_seen": 32983432,
      "step": 56810
    },
    {
      "epoch": 8.462168602919274,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02142925025346278,
      "loss": 0.8343,
      "num_input_tokens_seen": 32986472,
      "step": 56815
    },
    {
      "epoch": 8.462913315460233,
      "grad_norm": 0.01300048828125,
      "learning_rate": 0.021427488714235745,
      "loss": 0.8105,
      "num_input_tokens_seen": 32989064,
      "step": 56820
    },
    {
      "epoch": 8.463658028001191,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.021425727066422824,
      "loss": 0.8088,
      "num_input_tokens_seen": 32992168,
      "step": 56825
    },
    {
      "epoch": 8.46440274054215,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.02142396531005377,
      "loss": 0.7976,
      "num_input_tokens_seen": 32994856,
      "step": 56830
    },
    {
      "epoch": 8.46514745308311,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.02142220344515836,
      "loss": 0.8016,
      "num_input_tokens_seen": 32997992,
      "step": 56835
    },
    {
      "epoch": 8.46589216562407,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.021420441471766343,
      "loss": 0.8096,
      "num_input_tokens_seen": 33000584,
      "step": 56840
    },
    {
      "epoch": 8.466636878165028,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.021418679389907498,
      "loss": 0.796,
      "num_input_tokens_seen": 33003304,
      "step": 56845
    },
    {
      "epoch": 8.467381590705987,
      "grad_norm": 0.046875,
      "learning_rate": 0.02141691719961158,
      "loss": 0.8091,
      "num_input_tokens_seen": 33006504,
      "step": 56850
    },
    {
      "epoch": 8.468126303246947,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.021415154900908378,
      "loss": 0.8007,
      "num_input_tokens_seen": 33009128,
      "step": 56855
    },
    {
      "epoch": 8.468871015787906,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.02141339249382765,
      "loss": 0.8071,
      "num_input_tokens_seen": 33011848,
      "step": 56860
    },
    {
      "epoch": 8.469615728328865,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02141162997839917,
      "loss": 0.8038,
      "num_input_tokens_seen": 33014536,
      "step": 56865
    },
    {
      "epoch": 8.470360440869824,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02140986735465272,
      "loss": 0.8056,
      "num_input_tokens_seen": 33017672,
      "step": 56870
    },
    {
      "epoch": 8.471105153410784,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.021408104622618076,
      "loss": 0.8098,
      "num_input_tokens_seen": 33020232,
      "step": 56875
    },
    {
      "epoch": 8.471849865951743,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.021406341782325023,
      "loss": 0.812,
      "num_input_tokens_seen": 33022888,
      "step": 56880
    },
    {
      "epoch": 8.472594578492702,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02140457883380333,
      "loss": 0.7951,
      "num_input_tokens_seen": 33025768,
      "step": 56885
    },
    {
      "epoch": 8.47333929103366,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.021402815777082786,
      "loss": 0.8056,
      "num_input_tokens_seen": 33028680,
      "step": 56890
    },
    {
      "epoch": 8.474084003574621,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.021401052612193182,
      "loss": 0.8104,
      "num_input_tokens_seen": 33031880,
      "step": 56895
    },
    {
      "epoch": 8.47482871611558,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.0213992893391643,
      "loss": 0.8066,
      "num_input_tokens_seen": 33034760,
      "step": 56900
    },
    {
      "epoch": 8.475573428656539,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.021397525958025923,
      "loss": 0.8079,
      "num_input_tokens_seen": 33037576,
      "step": 56905
    },
    {
      "epoch": 8.476318141197497,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.021395762468807852,
      "loss": 0.8089,
      "num_input_tokens_seen": 33040520,
      "step": 56910
    },
    {
      "epoch": 8.477062853738458,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.021393998871539872,
      "loss": 0.8052,
      "num_input_tokens_seen": 33043496,
      "step": 56915
    },
    {
      "epoch": 8.477807566279417,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.02139223516625178,
      "loss": 0.8091,
      "num_input_tokens_seen": 33046376,
      "step": 56920
    },
    {
      "epoch": 8.478552278820375,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.021390471352973374,
      "loss": 0.7937,
      "num_input_tokens_seen": 33049224,
      "step": 56925
    },
    {
      "epoch": 8.479296991361334,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.021388707431734443,
      "loss": 0.8113,
      "num_input_tokens_seen": 33052136,
      "step": 56930
    },
    {
      "epoch": 8.480041703902295,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.021386943402564803,
      "loss": 0.8288,
      "num_input_tokens_seen": 33054792,
      "step": 56935
    },
    {
      "epoch": 8.480786416443253,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02138517926549424,
      "loss": 0.7993,
      "num_input_tokens_seen": 33057704,
      "step": 56940
    },
    {
      "epoch": 8.481531128984212,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.021383415020552562,
      "loss": 0.8158,
      "num_input_tokens_seen": 33060680,
      "step": 56945
    },
    {
      "epoch": 8.482275841525171,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.02138165066776958,
      "loss": 0.8087,
      "num_input_tokens_seen": 33063592,
      "step": 56950
    },
    {
      "epoch": 8.48302055406613,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.02137988620717509,
      "loss": 0.8063,
      "num_input_tokens_seen": 33066472,
      "step": 56955
    },
    {
      "epoch": 8.48376526660709,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.021378121638798906,
      "loss": 0.8016,
      "num_input_tokens_seen": 33069352,
      "step": 56960
    },
    {
      "epoch": 8.484509979148049,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02137635696267084,
      "loss": 0.7951,
      "num_input_tokens_seen": 33072200,
      "step": 56965
    },
    {
      "epoch": 8.485254691689008,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02137459217882071,
      "loss": 0.8005,
      "num_input_tokens_seen": 33075304,
      "step": 56970
    },
    {
      "epoch": 8.485999404229966,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02137282728727832,
      "loss": 0.805,
      "num_input_tokens_seen": 33078472,
      "step": 56975
    },
    {
      "epoch": 8.486744116770927,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.021371062288073492,
      "loss": 0.797,
      "num_input_tokens_seen": 33081512,
      "step": 56980
    },
    {
      "epoch": 8.487488829311886,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.02136929718123604,
      "loss": 0.8045,
      "num_input_tokens_seen": 33084232,
      "step": 56985
    },
    {
      "epoch": 8.488233541852845,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.021367531966795784,
      "loss": 0.8133,
      "num_input_tokens_seen": 33087176,
      "step": 56990
    },
    {
      "epoch": 8.488978254393803,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.021365766644782548,
      "loss": 0.8116,
      "num_input_tokens_seen": 33090024,
      "step": 56995
    },
    {
      "epoch": 8.489722966934764,
      "grad_norm": 0.04296875,
      "learning_rate": 0.021364001215226157,
      "loss": 0.8039,
      "num_input_tokens_seen": 33093064,
      "step": 57000
    },
    {
      "epoch": 8.490467679475723,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.021362235678156427,
      "loss": 0.7965,
      "num_input_tokens_seen": 33095752,
      "step": 57005
    },
    {
      "epoch": 8.491212392016681,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.0213604700336032,
      "loss": 0.7987,
      "num_input_tokens_seen": 33098472,
      "step": 57010
    },
    {
      "epoch": 8.49195710455764,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.021358704281596294,
      "loss": 0.7904,
      "num_input_tokens_seen": 33101096,
      "step": 57015
    },
    {
      "epoch": 8.4927018170986,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02135693842216554,
      "loss": 0.8042,
      "num_input_tokens_seen": 33104040,
      "step": 57020
    },
    {
      "epoch": 8.49344652963956,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.021355172455340774,
      "loss": 0.7969,
      "num_input_tokens_seen": 33106920,
      "step": 57025
    },
    {
      "epoch": 8.494191242180518,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.021353406381151824,
      "loss": 0.793,
      "num_input_tokens_seen": 33109992,
      "step": 57030
    },
    {
      "epoch": 8.494935954721477,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.021351640199628534,
      "loss": 0.8012,
      "num_input_tokens_seen": 33114216,
      "step": 57035
    },
    {
      "epoch": 8.495680667262437,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02134987391080074,
      "loss": 0.8138,
      "num_input_tokens_seen": 33117128,
      "step": 57040
    },
    {
      "epoch": 8.496425379803396,
      "grad_norm": 0.01416015625,
      "learning_rate": 0.02134810751469828,
      "loss": 0.7913,
      "num_input_tokens_seen": 33119656,
      "step": 57045
    },
    {
      "epoch": 8.497170092344355,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02134634101135099,
      "loss": 0.7889,
      "num_input_tokens_seen": 33122408,
      "step": 57050
    },
    {
      "epoch": 8.497914804885314,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.021344574400788723,
      "loss": 0.7908,
      "num_input_tokens_seen": 33125128,
      "step": 57055
    },
    {
      "epoch": 8.498659517426274,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02134280768304132,
      "loss": 0.8013,
      "num_input_tokens_seen": 33127912,
      "step": 57060
    },
    {
      "epoch": 8.499404229967233,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.021341040858138625,
      "loss": 0.8018,
      "num_input_tokens_seen": 33130952,
      "step": 57065
    },
    {
      "epoch": 8.500148942508192,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.021339273926110493,
      "loss": 0.8132,
      "num_input_tokens_seen": 33133896,
      "step": 57070
    },
    {
      "epoch": 8.50089365504915,
      "grad_norm": 0.04296875,
      "learning_rate": 0.021337506886986764,
      "loss": 0.8118,
      "num_input_tokens_seen": 33136968,
      "step": 57075
    },
    {
      "epoch": 8.501638367590111,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.021335739740797305,
      "loss": 0.796,
      "num_input_tokens_seen": 33139976,
      "step": 57080
    },
    {
      "epoch": 8.50238308013107,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.02133397248757196,
      "loss": 0.8075,
      "num_input_tokens_seen": 33142856,
      "step": 57085
    },
    {
      "epoch": 8.503127792672029,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02133220512734059,
      "loss": 0.7945,
      "num_input_tokens_seen": 33145704,
      "step": 57090
    },
    {
      "epoch": 8.503872505212987,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.021330437660133046,
      "loss": 0.8077,
      "num_input_tokens_seen": 33148552,
      "step": 57095
    },
    {
      "epoch": 8.504617217753946,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.021328670085979193,
      "loss": 0.7899,
      "num_input_tokens_seen": 33151112,
      "step": 57100
    },
    {
      "epoch": 8.505361930294907,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.02132690240490889,
      "loss": 0.843,
      "num_input_tokens_seen": 33153832,
      "step": 57105
    },
    {
      "epoch": 8.506106642835865,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.021325134616952006,
      "loss": 0.8059,
      "num_input_tokens_seen": 33156744,
      "step": 57110
    },
    {
      "epoch": 8.506851355376824,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.021323366722138394,
      "loss": 0.8115,
      "num_input_tokens_seen": 33159528,
      "step": 57115
    },
    {
      "epoch": 8.507596067917785,
      "grad_norm": 0.033203125,
      "learning_rate": 0.021321598720497933,
      "loss": 0.8199,
      "num_input_tokens_seen": 33162408,
      "step": 57120
    },
    {
      "epoch": 8.508340780458743,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.021319830612060484,
      "loss": 0.7942,
      "num_input_tokens_seen": 33165352,
      "step": 57125
    },
    {
      "epoch": 8.509085492999702,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.021318062396855922,
      "loss": 0.7905,
      "num_input_tokens_seen": 33168488,
      "step": 57130
    },
    {
      "epoch": 8.509830205540661,
      "grad_norm": 0.02734375,
      "learning_rate": 0.021316294074914116,
      "loss": 0.8341,
      "num_input_tokens_seen": 33171336,
      "step": 57135
    },
    {
      "epoch": 8.51057491808162,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.02131452564626494,
      "loss": 0.7926,
      "num_input_tokens_seen": 33174216,
      "step": 57140
    },
    {
      "epoch": 8.51131963062258,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.021312757110938278,
      "loss": 0.8009,
      "num_input_tokens_seen": 33177224,
      "step": 57145
    },
    {
      "epoch": 8.512064343163539,
      "grad_norm": 0.021484375,
      "learning_rate": 0.021310988468963996,
      "loss": 0.7973,
      "num_input_tokens_seen": 33179976,
      "step": 57150
    },
    {
      "epoch": 8.512809055704498,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.021309219720371975,
      "loss": 0.8119,
      "num_input_tokens_seen": 33182824,
      "step": 57155
    },
    {
      "epoch": 8.513553768245457,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.021307450865192097,
      "loss": 0.8038,
      "num_input_tokens_seen": 33185704,
      "step": 57160
    },
    {
      "epoch": 8.514298480786417,
      "grad_norm": 0.01275634765625,
      "learning_rate": 0.021305681903454252,
      "loss": 0.8103,
      "num_input_tokens_seen": 33188328,
      "step": 57165
    },
    {
      "epoch": 8.515043193327376,
      "grad_norm": 0.029296875,
      "learning_rate": 0.021303912835188323,
      "loss": 0.8203,
      "num_input_tokens_seen": 33191176,
      "step": 57170
    },
    {
      "epoch": 8.515787905868335,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.02130214366042419,
      "loss": 0.8035,
      "num_input_tokens_seen": 33194056,
      "step": 57175
    },
    {
      "epoch": 8.516532618409293,
      "grad_norm": 0.01385498046875,
      "learning_rate": 0.021300374379191748,
      "loss": 0.8077,
      "num_input_tokens_seen": 33197064,
      "step": 57180
    },
    {
      "epoch": 8.517277330950254,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.02129860499152088,
      "loss": 0.8013,
      "num_input_tokens_seen": 33199816,
      "step": 57185
    },
    {
      "epoch": 8.518022043491213,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.021296835497441484,
      "loss": 0.8048,
      "num_input_tokens_seen": 33202728,
      "step": 57190
    },
    {
      "epoch": 8.518766756032171,
      "grad_norm": 0.01348876953125,
      "learning_rate": 0.02129506589698345,
      "loss": 0.7999,
      "num_input_tokens_seen": 33205416,
      "step": 57195
    },
    {
      "epoch": 8.51951146857313,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.021293296190176676,
      "loss": 0.7933,
      "num_input_tokens_seen": 33208584,
      "step": 57200
    },
    {
      "epoch": 8.52025618111409,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.021291526377051062,
      "loss": 0.8052,
      "num_input_tokens_seen": 33211720,
      "step": 57205
    },
    {
      "epoch": 8.52100089365505,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.021289756457636505,
      "loss": 0.8166,
      "num_input_tokens_seen": 33214888,
      "step": 57210
    },
    {
      "epoch": 8.521745606196008,
      "grad_norm": 0.0150146484375,
      "learning_rate": 0.021287986431962905,
      "loss": 0.8057,
      "num_input_tokens_seen": 33218024,
      "step": 57215
    },
    {
      "epoch": 8.522490318736967,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.021286216300060164,
      "loss": 0.8118,
      "num_input_tokens_seen": 33221096,
      "step": 57220
    },
    {
      "epoch": 8.523235031277927,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.021284446061958187,
      "loss": 0.8215,
      "num_input_tokens_seen": 33224424,
      "step": 57225
    },
    {
      "epoch": 8.523979743818886,
      "grad_norm": 0.013916015625,
      "learning_rate": 0.02128267571768688,
      "loss": 0.8177,
      "num_input_tokens_seen": 33227048,
      "step": 57230
    },
    {
      "epoch": 8.524724456359845,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.021280905267276155,
      "loss": 0.8101,
      "num_input_tokens_seen": 33230184,
      "step": 57235
    },
    {
      "epoch": 8.525469168900804,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.021279134710755914,
      "loss": 0.798,
      "num_input_tokens_seen": 33233256,
      "step": 57240
    },
    {
      "epoch": 8.526213881441764,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02127736404815608,
      "loss": 0.7887,
      "num_input_tokens_seen": 33236264,
      "step": 57245
    },
    {
      "epoch": 8.526958593982723,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.021275593279506556,
      "loss": 0.8081,
      "num_input_tokens_seen": 33239144,
      "step": 57250
    },
    {
      "epoch": 8.527703306523682,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.021273822404837266,
      "loss": 0.7899,
      "num_input_tokens_seen": 33241736,
      "step": 57255
    },
    {
      "epoch": 8.52844801906464,
      "grad_norm": 0.03125,
      "learning_rate": 0.02127205142417812,
      "loss": 0.8067,
      "num_input_tokens_seen": 33244520,
      "step": 57260
    },
    {
      "epoch": 8.529192731605601,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.021270280337559042,
      "loss": 0.7972,
      "num_input_tokens_seen": 33247528,
      "step": 57265
    },
    {
      "epoch": 8.52993744414656,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.021268509145009946,
      "loss": 0.804,
      "num_input_tokens_seen": 33250248,
      "step": 57270
    },
    {
      "epoch": 8.530682156687519,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.021266737846560767,
      "loss": 0.8006,
      "num_input_tokens_seen": 33252808,
      "step": 57275
    },
    {
      "epoch": 8.531426869228477,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.021264966442241413,
      "loss": 0.788,
      "num_input_tokens_seen": 33255496,
      "step": 57280
    },
    {
      "epoch": 8.532171581769436,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.02126319493208182,
      "loss": 0.8053,
      "num_input_tokens_seen": 33258376,
      "step": 57285
    },
    {
      "epoch": 8.532916294310397,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.021261423316111916,
      "loss": 0.7915,
      "num_input_tokens_seen": 33260872,
      "step": 57290
    },
    {
      "epoch": 8.533661006851355,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.02125965159436163,
      "loss": 0.7956,
      "num_input_tokens_seen": 33263752,
      "step": 57295
    },
    {
      "epoch": 8.534405719392314,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.02125787976686089,
      "loss": 0.8012,
      "num_input_tokens_seen": 33266248,
      "step": 57300
    },
    {
      "epoch": 8.535150431933273,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.021256107833639633,
      "loss": 0.8025,
      "num_input_tokens_seen": 33269000,
      "step": 57305
    },
    {
      "epoch": 8.535895144474233,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.021254335794727793,
      "loss": 0.8072,
      "num_input_tokens_seen": 33271688,
      "step": 57310
    },
    {
      "epoch": 8.536639857015192,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.021252563650155307,
      "loss": 0.8016,
      "num_input_tokens_seen": 33274472,
      "step": 57315
    },
    {
      "epoch": 8.537384569556151,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.021250791399952108,
      "loss": 0.8002,
      "num_input_tokens_seen": 33277704,
      "step": 57320
    },
    {
      "epoch": 8.53812928209711,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.021249019044148144,
      "loss": 0.7936,
      "num_input_tokens_seen": 33280552,
      "step": 57325
    },
    {
      "epoch": 8.53887399463807,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.021247246582773358,
      "loss": 0.7927,
      "num_input_tokens_seen": 33283496,
      "step": 57330
    },
    {
      "epoch": 8.539618707179029,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.021245474015857687,
      "loss": 0.8269,
      "num_input_tokens_seen": 33286696,
      "step": 57335
    },
    {
      "epoch": 8.540363419719988,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.021243701343431082,
      "loss": 0.8192,
      "num_input_tokens_seen": 33289448,
      "step": 57340
    },
    {
      "epoch": 8.541108132260947,
      "grad_norm": 0.0140380859375,
      "learning_rate": 0.02124192856552349,
      "loss": 0.8015,
      "num_input_tokens_seen": 33292232,
      "step": 57345
    },
    {
      "epoch": 8.541852844801907,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.021240155682164856,
      "loss": 0.7903,
      "num_input_tokens_seen": 33294760,
      "step": 57350
    },
    {
      "epoch": 8.542597557342866,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.021238382693385134,
      "loss": 0.8145,
      "num_input_tokens_seen": 33297544,
      "step": 57355
    },
    {
      "epoch": 8.543342269883825,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.021236609599214275,
      "loss": 0.8029,
      "num_input_tokens_seen": 33300296,
      "step": 57360
    },
    {
      "epoch": 8.544086982424783,
      "grad_norm": 0.025390625,
      "learning_rate": 0.021234836399682235,
      "loss": 0.7922,
      "num_input_tokens_seen": 33302984,
      "step": 57365
    },
    {
      "epoch": 8.544831694965744,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.021233063094818976,
      "loss": 0.7947,
      "num_input_tokens_seen": 33305928,
      "step": 57370
    },
    {
      "epoch": 8.545576407506703,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.02123128968465445,
      "loss": 0.795,
      "num_input_tokens_seen": 33308936,
      "step": 57375
    },
    {
      "epoch": 8.546321120047661,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.021229516169218615,
      "loss": 0.8007,
      "num_input_tokens_seen": 33311880,
      "step": 57380
    },
    {
      "epoch": 8.54706583258862,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.02122774254854144,
      "loss": 0.7931,
      "num_input_tokens_seen": 33315048,
      "step": 57385
    },
    {
      "epoch": 8.54781054512958,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.021225968822652878,
      "loss": 0.7903,
      "num_input_tokens_seen": 33317960,
      "step": 57390
    },
    {
      "epoch": 8.54855525767054,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.021224194991582904,
      "loss": 0.7939,
      "num_input_tokens_seen": 33320808,
      "step": 57395
    },
    {
      "epoch": 8.549299970211498,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.02122242105536148,
      "loss": 0.8192,
      "num_input_tokens_seen": 33324872,
      "step": 57400
    },
    {
      "epoch": 8.550044682752457,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.02122064701401858,
      "loss": 0.7832,
      "num_input_tokens_seen": 33327272,
      "step": 57405
    },
    {
      "epoch": 8.550789395293418,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.021218872867584167,
      "loss": 0.8078,
      "num_input_tokens_seen": 33329864,
      "step": 57410
    },
    {
      "epoch": 8.551534107834376,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.021217098616088222,
      "loss": 0.7903,
      "num_input_tokens_seen": 33332488,
      "step": 57415
    },
    {
      "epoch": 8.552278820375335,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.021215324259560714,
      "loss": 0.7927,
      "num_input_tokens_seen": 33335720,
      "step": 57420
    },
    {
      "epoch": 8.553023532916294,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.021213549798031615,
      "loss": 0.7947,
      "num_input_tokens_seen": 33338632,
      "step": 57425
    },
    {
      "epoch": 8.553768245457253,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02121177523153091,
      "loss": 0.8247,
      "num_input_tokens_seen": 33341576,
      "step": 57430
    },
    {
      "epoch": 8.554512957998213,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.021210000560088576,
      "loss": 0.7963,
      "num_input_tokens_seen": 33344456,
      "step": 57435
    },
    {
      "epoch": 8.555257670539172,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.021208225783734595,
      "loss": 0.8046,
      "num_input_tokens_seen": 33347336,
      "step": 57440
    },
    {
      "epoch": 8.55600238308013,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.021206450902498945,
      "loss": 0.8056,
      "num_input_tokens_seen": 33350472,
      "step": 57445
    },
    {
      "epoch": 8.556747095621091,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02120467591641162,
      "loss": 0.8144,
      "num_input_tokens_seen": 33353256,
      "step": 57450
    },
    {
      "epoch": 8.55749180816205,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.0212029008255026,
      "loss": 0.8045,
      "num_input_tokens_seen": 33355848,
      "step": 57455
    },
    {
      "epoch": 8.558236520703009,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02120112562980188,
      "loss": 0.7991,
      "num_input_tokens_seen": 33358600,
      "step": 57460
    },
    {
      "epoch": 8.558981233243967,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.021199350329339433,
      "loss": 0.7935,
      "num_input_tokens_seen": 33361480,
      "step": 57465
    },
    {
      "epoch": 8.559725945784926,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.02119757492414527,
      "loss": 0.8142,
      "num_input_tokens_seen": 33364232,
      "step": 57470
    },
    {
      "epoch": 8.560470658325887,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.02119579941424938,
      "loss": 0.8281,
      "num_input_tokens_seen": 33367176,
      "step": 57475
    },
    {
      "epoch": 8.561215370866845,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.02119402379968175,
      "loss": 0.8015,
      "num_input_tokens_seen": 33370216,
      "step": 57480
    },
    {
      "epoch": 8.561960083407804,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02119224808047239,
      "loss": 0.786,
      "num_input_tokens_seen": 33373000,
      "step": 57485
    },
    {
      "epoch": 8.562704795948763,
      "grad_norm": 0.033203125,
      "learning_rate": 0.021190472256651285,
      "loss": 0.802,
      "num_input_tokens_seen": 33376136,
      "step": 57490
    },
    {
      "epoch": 8.563449508489724,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.021188696328248446,
      "loss": 0.7944,
      "num_input_tokens_seen": 33378920,
      "step": 57495
    },
    {
      "epoch": 8.564194221030682,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.021186920295293874,
      "loss": 0.7929,
      "num_input_tokens_seen": 33381928,
      "step": 57500
    },
    {
      "epoch": 8.564938933571641,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.02118514415781757,
      "loss": 0.8025,
      "num_input_tokens_seen": 33384840,
      "step": 57505
    },
    {
      "epoch": 8.5656836461126,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.021183367915849544,
      "loss": 0.7897,
      "num_input_tokens_seen": 33387656,
      "step": 57510
    },
    {
      "epoch": 8.56642835865356,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.021181591569419803,
      "loss": 0.7858,
      "num_input_tokens_seen": 33390856,
      "step": 57515
    },
    {
      "epoch": 8.567173071194519,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02117981511855835,
      "loss": 0.779,
      "num_input_tokens_seen": 33393448,
      "step": 57520
    },
    {
      "epoch": 8.567917783735478,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.021178038563295202,
      "loss": 0.8024,
      "num_input_tokens_seen": 33396392,
      "step": 57525
    },
    {
      "epoch": 8.568662496276437,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02117626190366037,
      "loss": 0.8224,
      "num_input_tokens_seen": 33399304,
      "step": 57530
    },
    {
      "epoch": 8.569407208817397,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.021174485139683875,
      "loss": 0.7961,
      "num_input_tokens_seen": 33402088,
      "step": 57535
    },
    {
      "epoch": 8.570151921358356,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.021172708271395727,
      "loss": 0.788,
      "num_input_tokens_seen": 33404808,
      "step": 57540
    },
    {
      "epoch": 8.570896633899315,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.021170931298825954,
      "loss": 0.786,
      "num_input_tokens_seen": 33407720,
      "step": 57545
    },
    {
      "epoch": 8.571641346440273,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.02116915422200456,
      "loss": 0.8022,
      "num_input_tokens_seen": 33410792,
      "step": 57550
    },
    {
      "epoch": 8.572386058981234,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.021167377040961576,
      "loss": 0.8034,
      "num_input_tokens_seen": 33413544,
      "step": 57555
    },
    {
      "epoch": 8.573130771522193,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.021165599755727026,
      "loss": 0.819,
      "num_input_tokens_seen": 33416520,
      "step": 57560
    },
    {
      "epoch": 8.573875484063151,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.021163822366330933,
      "loss": 0.8152,
      "num_input_tokens_seen": 33419336,
      "step": 57565
    },
    {
      "epoch": 8.57462019660411,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.021162044872803328,
      "loss": 0.7914,
      "num_input_tokens_seen": 33421896,
      "step": 57570
    },
    {
      "epoch": 8.57536490914507,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.021160267275174238,
      "loss": 0.8134,
      "num_input_tokens_seen": 33424840,
      "step": 57575
    },
    {
      "epoch": 8.57610962168603,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.021158489573473695,
      "loss": 0.8014,
      "num_input_tokens_seen": 33427784,
      "step": 57580
    },
    {
      "epoch": 8.576854334226988,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.021156711767731724,
      "loss": 0.7922,
      "num_input_tokens_seen": 33430536,
      "step": 57585
    },
    {
      "epoch": 8.577599046767947,
      "grad_norm": 0.021484375,
      "learning_rate": 0.021154933857978366,
      "loss": 0.8173,
      "num_input_tokens_seen": 33433704,
      "step": 57590
    },
    {
      "epoch": 8.578343759308908,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.02115315584424366,
      "loss": 0.7984,
      "num_input_tokens_seen": 33436584,
      "step": 57595
    },
    {
      "epoch": 8.579088471849866,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.02115137772655764,
      "loss": 0.8009,
      "num_input_tokens_seen": 33439432,
      "step": 57600
    },
    {
      "epoch": 8.579833184390825,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.021149599504950344,
      "loss": 0.799,
      "num_input_tokens_seen": 33442504,
      "step": 57605
    },
    {
      "epoch": 8.580577896931784,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02114782117945181,
      "loss": 0.7975,
      "num_input_tokens_seen": 33445480,
      "step": 57610
    },
    {
      "epoch": 8.581322609472743,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.021146042750092093,
      "loss": 0.7944,
      "num_input_tokens_seen": 33448392,
      "step": 57615
    },
    {
      "epoch": 8.582067322013703,
      "grad_norm": 0.033203125,
      "learning_rate": 0.021144264216901226,
      "loss": 0.8061,
      "num_input_tokens_seen": 33451208,
      "step": 57620
    },
    {
      "epoch": 8.582812034554662,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.021142485579909265,
      "loss": 0.8216,
      "num_input_tokens_seen": 33454088,
      "step": 57625
    },
    {
      "epoch": 8.58355674709562,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.021140706839146248,
      "loss": 0.7792,
      "num_input_tokens_seen": 33456968,
      "step": 57630
    },
    {
      "epoch": 8.584301459636581,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.021138927994642225,
      "loss": 0.8083,
      "num_input_tokens_seen": 33459912,
      "step": 57635
    },
    {
      "epoch": 8.58504617217754,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.021137149046427262,
      "loss": 0.7829,
      "num_input_tokens_seen": 33463048,
      "step": 57640
    },
    {
      "epoch": 8.585790884718499,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.021135369994531403,
      "loss": 0.7891,
      "num_input_tokens_seen": 33465736,
      "step": 57645
    },
    {
      "epoch": 8.586535597259457,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.0211335908389847,
      "loss": 0.8086,
      "num_input_tokens_seen": 33468712,
      "step": 57650
    },
    {
      "epoch": 8.587280309800416,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.021131811579817215,
      "loss": 0.805,
      "num_input_tokens_seen": 33471592,
      "step": 57655
    },
    {
      "epoch": 8.588025022341377,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.021130032217059003,
      "loss": 0.8072,
      "num_input_tokens_seen": 33474760,
      "step": 57660
    },
    {
      "epoch": 8.588769734882336,
      "grad_norm": 0.01214599609375,
      "learning_rate": 0.021128252750740126,
      "loss": 0.7786,
      "num_input_tokens_seen": 33477864,
      "step": 57665
    },
    {
      "epoch": 8.589514447423294,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.021126473180890653,
      "loss": 0.7867,
      "num_input_tokens_seen": 33481096,
      "step": 57670
    },
    {
      "epoch": 8.590259159964253,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.02112469350754064,
      "loss": 0.7999,
      "num_input_tokens_seen": 33483624,
      "step": 57675
    },
    {
      "epoch": 8.591003872505214,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.021122913730720155,
      "loss": 0.797,
      "num_input_tokens_seen": 33486536,
      "step": 57680
    },
    {
      "epoch": 8.591748585046172,
      "grad_norm": 0.01397705078125,
      "learning_rate": 0.021121133850459266,
      "loss": 0.7922,
      "num_input_tokens_seen": 33489256,
      "step": 57685
    },
    {
      "epoch": 8.592493297587131,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.021119353866788038,
      "loss": 0.7594,
      "num_input_tokens_seen": 33492040,
      "step": 57690
    },
    {
      "epoch": 8.59323801012809,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.021117573779736547,
      "loss": 0.7822,
      "num_input_tokens_seen": 33495080,
      "step": 57695
    },
    {
      "epoch": 8.59398272266905,
      "grad_norm": 0.033203125,
      "learning_rate": 0.021115793589334873,
      "loss": 0.7781,
      "num_input_tokens_seen": 33498056,
      "step": 57700
    },
    {
      "epoch": 8.59472743521001,
      "grad_norm": 0.03125,
      "learning_rate": 0.021114013295613073,
      "loss": 0.7861,
      "num_input_tokens_seen": 33500872,
      "step": 57705
    },
    {
      "epoch": 8.595472147750968,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.021112232898601237,
      "loss": 0.8209,
      "num_input_tokens_seen": 33504008,
      "step": 57710
    },
    {
      "epoch": 8.596216860291927,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.021110452398329434,
      "loss": 0.8268,
      "num_input_tokens_seen": 33506760,
      "step": 57715
    },
    {
      "epoch": 8.596961572832887,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.021108671794827748,
      "loss": 0.8299,
      "num_input_tokens_seen": 33509608,
      "step": 57720
    },
    {
      "epoch": 8.597706285373846,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.021106891088126266,
      "loss": 0.8032,
      "num_input_tokens_seen": 33512648,
      "step": 57725
    },
    {
      "epoch": 8.598450997914805,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02110511027825506,
      "loss": 0.7822,
      "num_input_tokens_seen": 33515528,
      "step": 57730
    },
    {
      "epoch": 8.599195710455763,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.021103329365244224,
      "loss": 0.7631,
      "num_input_tokens_seen": 33518632,
      "step": 57735
    },
    {
      "epoch": 8.599940422996724,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.021101548349123842,
      "loss": 0.8123,
      "num_input_tokens_seen": 33521352,
      "step": 57740
    },
    {
      "epoch": 8.600685135537683,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.021099767229924005,
      "loss": 0.7996,
      "num_input_tokens_seen": 33524008,
      "step": 57745
    },
    {
      "epoch": 8.601429848078642,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.021097986007674794,
      "loss": 0.8188,
      "num_input_tokens_seen": 33527048,
      "step": 57750
    },
    {
      "epoch": 8.6021745606196,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.021096204682406312,
      "loss": 0.7745,
      "num_input_tokens_seen": 33530024,
      "step": 57755
    },
    {
      "epoch": 8.60291927316056,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.021094423254148644,
      "loss": 0.7958,
      "num_input_tokens_seen": 33532872,
      "step": 57760
    },
    {
      "epoch": 8.60366398570152,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.021092641722931887,
      "loss": 0.7964,
      "num_input_tokens_seen": 33535752,
      "step": 57765
    },
    {
      "epoch": 8.604408698242478,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.021090860088786148,
      "loss": 0.821,
      "num_input_tokens_seen": 33538696,
      "step": 57770
    },
    {
      "epoch": 8.605153410783437,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02108907835174151,
      "loss": 0.802,
      "num_input_tokens_seen": 33541448,
      "step": 57775
    },
    {
      "epoch": 8.605898123324398,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.021087296511828085,
      "loss": 0.8162,
      "num_input_tokens_seen": 33544648,
      "step": 57780
    },
    {
      "epoch": 8.606642835865356,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.021085514569075977,
      "loss": 0.7823,
      "num_input_tokens_seen": 33547560,
      "step": 57785
    },
    {
      "epoch": 8.607387548406315,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02108373252351528,
      "loss": 0.7912,
      "num_input_tokens_seen": 33550312,
      "step": 57790
    },
    {
      "epoch": 8.608132260947274,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.021081950375176108,
      "loss": 0.8018,
      "num_input_tokens_seen": 33553032,
      "step": 57795
    },
    {
      "epoch": 8.608876973488233,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.021080168124088562,
      "loss": 0.7939,
      "num_input_tokens_seen": 33555880,
      "step": 57800
    },
    {
      "epoch": 8.609621686029193,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02107838577028276,
      "loss": 0.7875,
      "num_input_tokens_seen": 33558792,
      "step": 57805
    },
    {
      "epoch": 8.610366398570152,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.021076603313788807,
      "loss": 0.8031,
      "num_input_tokens_seen": 33561864,
      "step": 57810
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.021074820754636816,
      "loss": 0.8222,
      "num_input_tokens_seen": 33564680,
      "step": 57815
    },
    {
      "epoch": 8.61185582365207,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.0210730380928569,
      "loss": 0.7935,
      "num_input_tokens_seen": 33567848,
      "step": 57820
    },
    {
      "epoch": 8.61260053619303,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.021071255328479182,
      "loss": 0.7981,
      "num_input_tokens_seen": 33570696,
      "step": 57825
    },
    {
      "epoch": 8.613345248733989,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.021069472461533775,
      "loss": 0.8417,
      "num_input_tokens_seen": 33573512,
      "step": 57830
    },
    {
      "epoch": 8.614089961274948,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.021067689492050796,
      "loss": 0.8108,
      "num_input_tokens_seen": 33576296,
      "step": 57835
    },
    {
      "epoch": 8.614834673815906,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.021065906420060378,
      "loss": 0.7789,
      "num_input_tokens_seen": 33579304,
      "step": 57840
    },
    {
      "epoch": 8.615579386356867,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.02106412324559263,
      "loss": 0.8222,
      "num_input_tokens_seen": 33582152,
      "step": 57845
    },
    {
      "epoch": 8.616324098897826,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02106233996867768,
      "loss": 0.8116,
      "num_input_tokens_seen": 33584968,
      "step": 57850
    },
    {
      "epoch": 8.617068811438784,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.021060556589345664,
      "loss": 0.8097,
      "num_input_tokens_seen": 33587784,
      "step": 57855
    },
    {
      "epoch": 8.617813523979743,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.021058773107626702,
      "loss": 0.7821,
      "num_input_tokens_seen": 33590760,
      "step": 57860
    },
    {
      "epoch": 8.618558236520704,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.021056989523550926,
      "loss": 0.7762,
      "num_input_tokens_seen": 33593640,
      "step": 57865
    },
    {
      "epoch": 8.619302949061662,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.02105520583714847,
      "loss": 0.8011,
      "num_input_tokens_seen": 33596648,
      "step": 57870
    },
    {
      "epoch": 8.620047661602621,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.021053422048449465,
      "loss": 0.7825,
      "num_input_tokens_seen": 33599592,
      "step": 57875
    },
    {
      "epoch": 8.62079237414358,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02105163815748404,
      "loss": 0.7941,
      "num_input_tokens_seen": 33602472,
      "step": 57880
    },
    {
      "epoch": 8.62153708668454,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.021049854164282348,
      "loss": 0.7999,
      "num_input_tokens_seen": 33605480,
      "step": 57885
    },
    {
      "epoch": 8.6222817992255,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.021048070068874512,
      "loss": 0.7877,
      "num_input_tokens_seen": 33608392,
      "step": 57890
    },
    {
      "epoch": 8.623026511766458,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.021046285871290683,
      "loss": 0.8255,
      "num_input_tokens_seen": 33611336,
      "step": 57895
    },
    {
      "epoch": 8.623771224307417,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.021044501571560997,
      "loss": 0.8092,
      "num_input_tokens_seen": 33614376,
      "step": 57900
    },
    {
      "epoch": 8.624515936848377,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.021042717169715605,
      "loss": 0.8056,
      "num_input_tokens_seen": 33617288,
      "step": 57905
    },
    {
      "epoch": 8.625260649389336,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.021040932665784648,
      "loss": 0.8144,
      "num_input_tokens_seen": 33620232,
      "step": 57910
    },
    {
      "epoch": 8.626005361930295,
      "grad_norm": 0.033203125,
      "learning_rate": 0.021039148059798268,
      "loss": 0.7842,
      "num_input_tokens_seen": 33623304,
      "step": 57915
    },
    {
      "epoch": 8.626750074471254,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02103736335178662,
      "loss": 0.7768,
      "num_input_tokens_seen": 33626120,
      "step": 57920
    },
    {
      "epoch": 8.627494787012214,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.021035578541779856,
      "loss": 0.7742,
      "num_input_tokens_seen": 33628776,
      "step": 57925
    },
    {
      "epoch": 8.628239499553173,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.021033793629808124,
      "loss": 0.8368,
      "num_input_tokens_seen": 33631720,
      "step": 57930
    },
    {
      "epoch": 8.628984212094132,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.021032008615901588,
      "loss": 0.8012,
      "num_input_tokens_seen": 33634696,
      "step": 57935
    },
    {
      "epoch": 8.62972892463509,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.021030223500090393,
      "loss": 0.7745,
      "num_input_tokens_seen": 33637480,
      "step": 57940
    },
    {
      "epoch": 8.63047363717605,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.021028438282404704,
      "loss": 0.8115,
      "num_input_tokens_seen": 33640968,
      "step": 57945
    },
    {
      "epoch": 8.63121834971701,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.021026652962874672,
      "loss": 0.8149,
      "num_input_tokens_seen": 33644104,
      "step": 57950
    },
    {
      "epoch": 8.631963062257968,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.021024867541530465,
      "loss": 0.7915,
      "num_input_tokens_seen": 33646856,
      "step": 57955
    },
    {
      "epoch": 8.632707774798927,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.021023082018402243,
      "loss": 0.8033,
      "num_input_tokens_seen": 33649384,
      "step": 57960
    },
    {
      "epoch": 8.633452487339888,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.021021296393520173,
      "loss": 0.799,
      "num_input_tokens_seen": 33652328,
      "step": 57965
    },
    {
      "epoch": 8.634197199880846,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.021019510666914425,
      "loss": 0.7794,
      "num_input_tokens_seen": 33655208,
      "step": 57970
    },
    {
      "epoch": 8.634941912421805,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.021017724838615155,
      "loss": 0.7951,
      "num_input_tokens_seen": 33657928,
      "step": 57975
    },
    {
      "epoch": 8.635686624962764,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.021015938908652543,
      "loss": 0.7957,
      "num_input_tokens_seen": 33660904,
      "step": 57980
    },
    {
      "epoch": 8.636431337503723,
      "grad_norm": 0.02734375,
      "learning_rate": 0.021014152877056762,
      "loss": 0.7773,
      "num_input_tokens_seen": 33663720,
      "step": 57985
    },
    {
      "epoch": 8.637176050044683,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.021012366743857973,
      "loss": 0.7747,
      "num_input_tokens_seen": 33666440,
      "step": 57990
    },
    {
      "epoch": 8.637920762585642,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.02101058050908636,
      "loss": 0.824,
      "num_input_tokens_seen": 33669128,
      "step": 57995
    },
    {
      "epoch": 8.6386654751266,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.021008794172772103,
      "loss": 0.7944,
      "num_input_tokens_seen": 33671816,
      "step": 58000
    },
    {
      "epoch": 8.63941018766756,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02100700773494537,
      "loss": 0.8064,
      "num_input_tokens_seen": 33674696,
      "step": 58005
    },
    {
      "epoch": 8.64015490020852,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.021005221195636352,
      "loss": 0.7665,
      "num_input_tokens_seen": 33677640,
      "step": 58010
    },
    {
      "epoch": 8.640899612749479,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02100343455487522,
      "loss": 0.7895,
      "num_input_tokens_seen": 33680648,
      "step": 58015
    },
    {
      "epoch": 8.641644325290438,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.021001647812692165,
      "loss": 0.7833,
      "num_input_tokens_seen": 33683944,
      "step": 58020
    },
    {
      "epoch": 8.642389037831396,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.020999860969117372,
      "loss": 0.8246,
      "num_input_tokens_seen": 33686760,
      "step": 58025
    },
    {
      "epoch": 8.643133750372357,
      "grad_norm": 0.021484375,
      "learning_rate": 0.020998074024181024,
      "loss": 0.7673,
      "num_input_tokens_seen": 33689608,
      "step": 58030
    },
    {
      "epoch": 8.643878462913316,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.020996286977913312,
      "loss": 0.7931,
      "num_input_tokens_seen": 33692904,
      "step": 58035
    },
    {
      "epoch": 8.644623175454274,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.020994499830344425,
      "loss": 0.8262,
      "num_input_tokens_seen": 33695784,
      "step": 58040
    },
    {
      "epoch": 8.645367887995233,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.020992712581504556,
      "loss": 0.8088,
      "num_input_tokens_seen": 33698760,
      "step": 58045
    },
    {
      "epoch": 8.646112600536194,
      "grad_norm": 0.015380859375,
      "learning_rate": 0.020990925231423898,
      "loss": 0.81,
      "num_input_tokens_seen": 33701768,
      "step": 58050
    },
    {
      "epoch": 8.646857313077152,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.020989137780132647,
      "loss": 0.8251,
      "num_input_tokens_seen": 33704424,
      "step": 58055
    },
    {
      "epoch": 8.647602025618111,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.020987350227661003,
      "loss": 0.8049,
      "num_input_tokens_seen": 33707240,
      "step": 58060
    },
    {
      "epoch": 8.64834673815907,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.02098556257403916,
      "loss": 0.8125,
      "num_input_tokens_seen": 33710088,
      "step": 58065
    },
    {
      "epoch": 8.64909145070003,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02098377481929732,
      "loss": 0.8231,
      "num_input_tokens_seen": 33713256,
      "step": 58070
    },
    {
      "epoch": 8.64983616324099,
      "grad_norm": 0.01318359375,
      "learning_rate": 0.02098198696346569,
      "loss": 0.7825,
      "num_input_tokens_seen": 33716008,
      "step": 58075
    },
    {
      "epoch": 8.650580875781948,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.020980199006574465,
      "loss": 0.8006,
      "num_input_tokens_seen": 33719176,
      "step": 58080
    },
    {
      "epoch": 8.651325588322907,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02097841094865386,
      "loss": 0.8048,
      "num_input_tokens_seen": 33721992,
      "step": 58085
    },
    {
      "epoch": 8.652070300863867,
      "grad_norm": 0.0234375,
      "learning_rate": 0.020976622789734077,
      "loss": 0.8196,
      "num_input_tokens_seen": 33724648,
      "step": 58090
    },
    {
      "epoch": 8.652815013404826,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02097483452984532,
      "loss": 0.7878,
      "num_input_tokens_seen": 33727624,
      "step": 58095
    },
    {
      "epoch": 8.653559725945785,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.02097304616901782,
      "loss": 0.7755,
      "num_input_tokens_seen": 33730248,
      "step": 58100
    },
    {
      "epoch": 8.654304438486744,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.020971257707281764,
      "loss": 0.7879,
      "num_input_tokens_seen": 33733000,
      "step": 58105
    },
    {
      "epoch": 8.655049151027704,
      "grad_norm": 0.029296875,
      "learning_rate": 0.020969469144667387,
      "loss": 0.7875,
      "num_input_tokens_seen": 33735880,
      "step": 58110
    },
    {
      "epoch": 8.655793863568663,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.02096768048120489,
      "loss": 0.794,
      "num_input_tokens_seen": 33738536,
      "step": 58115
    },
    {
      "epoch": 8.656538576109622,
      "grad_norm": 0.033203125,
      "learning_rate": 0.020965891716924496,
      "loss": 0.8019,
      "num_input_tokens_seen": 33741256,
      "step": 58120
    },
    {
      "epoch": 8.65728328865058,
      "grad_norm": 0.01312255859375,
      "learning_rate": 0.020964102851856424,
      "loss": 0.7874,
      "num_input_tokens_seen": 33744168,
      "step": 58125
    },
    {
      "epoch": 8.65802800119154,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.020962313886030894,
      "loss": 0.841,
      "num_input_tokens_seen": 33746888,
      "step": 58130
    },
    {
      "epoch": 8.6587727137325,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02096052481947814,
      "loss": 0.7753,
      "num_input_tokens_seen": 33749832,
      "step": 58135
    },
    {
      "epoch": 8.659517426273458,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02095873565222837,
      "loss": 0.7926,
      "num_input_tokens_seen": 33752872,
      "step": 58140
    },
    {
      "epoch": 8.660262138814417,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.020956946384311823,
      "loss": 0.8116,
      "num_input_tokens_seen": 33755848,
      "step": 58145
    },
    {
      "epoch": 8.661006851355378,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02095515701575872,
      "loss": 0.8149,
      "num_input_tokens_seen": 33758824,
      "step": 58150
    },
    {
      "epoch": 8.661751563896336,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.020953367546599287,
      "loss": 0.7862,
      "num_input_tokens_seen": 33761480,
      "step": 58155
    },
    {
      "epoch": 8.662496276437295,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02095157797686376,
      "loss": 0.8274,
      "num_input_tokens_seen": 33764360,
      "step": 58160
    },
    {
      "epoch": 8.663240988978254,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.020949788306582375,
      "loss": 0.8175,
      "num_input_tokens_seen": 33767336,
      "step": 58165
    },
    {
      "epoch": 8.663985701519213,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02094799853578536,
      "loss": 0.8117,
      "num_input_tokens_seen": 33770376,
      "step": 58170
    },
    {
      "epoch": 8.664730414060173,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02094620866450296,
      "loss": 0.8001,
      "num_input_tokens_seen": 33773384,
      "step": 58175
    },
    {
      "epoch": 8.665475126601132,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.020944418692765407,
      "loss": 0.7898,
      "num_input_tokens_seen": 33776296,
      "step": 58180
    },
    {
      "epoch": 8.66621983914209,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.020942628620602942,
      "loss": 0.8076,
      "num_input_tokens_seen": 33779208,
      "step": 58185
    },
    {
      "epoch": 8.66696455168305,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.020940838448045805,
      "loss": 0.8169,
      "num_input_tokens_seen": 33782280,
      "step": 58190
    },
    {
      "epoch": 8.66770926422401,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.020939048175124236,
      "loss": 0.806,
      "num_input_tokens_seen": 33785192,
      "step": 58195
    },
    {
      "epoch": 8.668453976764969,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.020937257801868493,
      "loss": 0.805,
      "num_input_tokens_seen": 33788040,
      "step": 58200
    },
    {
      "epoch": 8.669198689305928,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02093546732830881,
      "loss": 0.7913,
      "num_input_tokens_seen": 33791080,
      "step": 58205
    },
    {
      "epoch": 8.669943401846886,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02093367675447544,
      "loss": 0.7977,
      "num_input_tokens_seen": 33793896,
      "step": 58210
    },
    {
      "epoch": 8.670688114387847,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.020931886080398625,
      "loss": 0.7731,
      "num_input_tokens_seen": 33796808,
      "step": 58215
    },
    {
      "epoch": 8.671432826928806,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02093009530610863,
      "loss": 0.7829,
      "num_input_tokens_seen": 33799912,
      "step": 58220
    },
    {
      "epoch": 8.672177539469764,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0209283044316357,
      "loss": 0.8099,
      "num_input_tokens_seen": 33803048,
      "step": 58225
    },
    {
      "epoch": 8.672922252010723,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.020926513457010097,
      "loss": 0.8039,
      "num_input_tokens_seen": 33806056,
      "step": 58230
    },
    {
      "epoch": 8.673666964551684,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.020924722382262066,
      "loss": 0.7951,
      "num_input_tokens_seen": 33809192,
      "step": 58235
    },
    {
      "epoch": 8.674411677092642,
      "grad_norm": 0.03125,
      "learning_rate": 0.020922931207421877,
      "loss": 0.7996,
      "num_input_tokens_seen": 33811848,
      "step": 58240
    },
    {
      "epoch": 8.675156389633601,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.020921139932519782,
      "loss": 0.8164,
      "num_input_tokens_seen": 33814920,
      "step": 58245
    },
    {
      "epoch": 8.67590110217456,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.020919348557586044,
      "loss": 0.8127,
      "num_input_tokens_seen": 33817704,
      "step": 58250
    },
    {
      "epoch": 8.67664581471552,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.02091755708265093,
      "loss": 0.8053,
      "num_input_tokens_seen": 33820488,
      "step": 58255
    },
    {
      "epoch": 8.67739052725648,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.020915765507744703,
      "loss": 0.7977,
      "num_input_tokens_seen": 33823400,
      "step": 58260
    },
    {
      "epoch": 8.678135239797438,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.020913973832897635,
      "loss": 0.8047,
      "num_input_tokens_seen": 33826216,
      "step": 58265
    },
    {
      "epoch": 8.678879952338397,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.020912182058139984,
      "loss": 0.7885,
      "num_input_tokens_seen": 33829064,
      "step": 58270
    },
    {
      "epoch": 8.679624664879357,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.02091039018350203,
      "loss": 0.7992,
      "num_input_tokens_seen": 33832104,
      "step": 58275
    },
    {
      "epoch": 8.680369377420316,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.02090859820901404,
      "loss": 0.8066,
      "num_input_tokens_seen": 33835016,
      "step": 58280
    },
    {
      "epoch": 8.681114089961275,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.020906806134706288,
      "loss": 0.7855,
      "num_input_tokens_seen": 33838248,
      "step": 58285
    },
    {
      "epoch": 8.681858802502234,
      "grad_norm": 0.01531982421875,
      "learning_rate": 0.020905013960609047,
      "loss": 0.7893,
      "num_input_tokens_seen": 33841160,
      "step": 58290
    },
    {
      "epoch": 8.682603515043194,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.020903221686752595,
      "loss": 0.8029,
      "num_input_tokens_seen": 33844008,
      "step": 58295
    },
    {
      "epoch": 8.683348227584153,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.02090142931316722,
      "loss": 0.8041,
      "num_input_tokens_seen": 33846696,
      "step": 58300
    },
    {
      "epoch": 8.684092940125112,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.020899636839883193,
      "loss": 0.789,
      "num_input_tokens_seen": 33849736,
      "step": 58305
    },
    {
      "epoch": 8.68483765266607,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.020897844266930796,
      "loss": 0.7708,
      "num_input_tokens_seen": 33852552,
      "step": 58310
    },
    {
      "epoch": 8.68558236520703,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.020896051594340312,
      "loss": 0.8339,
      "num_input_tokens_seen": 33855400,
      "step": 58315
    },
    {
      "epoch": 8.68632707774799,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.02089425882214203,
      "loss": 0.8242,
      "num_input_tokens_seen": 33858248,
      "step": 58320
    },
    {
      "epoch": 8.687071790288948,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.020892465950366236,
      "loss": 0.7861,
      "num_input_tokens_seen": 33860904,
      "step": 58325
    },
    {
      "epoch": 8.687816502829907,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02089067297904322,
      "loss": 0.7811,
      "num_input_tokens_seen": 33863816,
      "step": 58330
    },
    {
      "epoch": 8.688561215370868,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.020888879908203272,
      "loss": 0.8119,
      "num_input_tokens_seen": 33867048,
      "step": 58335
    },
    {
      "epoch": 8.689305927911827,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.020887086737876683,
      "loss": 0.8051,
      "num_input_tokens_seen": 33870088,
      "step": 58340
    },
    {
      "epoch": 8.690050640452785,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.020885293468093744,
      "loss": 0.8093,
      "num_input_tokens_seen": 33873576,
      "step": 58345
    },
    {
      "epoch": 8.690795352993744,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.020883500098884754,
      "loss": 0.8146,
      "num_input_tokens_seen": 33876520,
      "step": 58350
    },
    {
      "epoch": 8.691540065534703,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.020881706630280012,
      "loss": 0.7726,
      "num_input_tokens_seen": 33879368,
      "step": 58355
    },
    {
      "epoch": 8.692284778075663,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.02087991306230981,
      "loss": 0.794,
      "num_input_tokens_seen": 33882536,
      "step": 58360
    },
    {
      "epoch": 8.693029490616622,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.020878119395004457,
      "loss": 0.8032,
      "num_input_tokens_seen": 33885576,
      "step": 58365
    },
    {
      "epoch": 8.69377420315758,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.020876325628394254,
      "loss": 0.7812,
      "num_input_tokens_seen": 33888456,
      "step": 58370
    },
    {
      "epoch": 8.69451891569854,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.020874531762509495,
      "loss": 0.8227,
      "num_input_tokens_seen": 33891304,
      "step": 58375
    },
    {
      "epoch": 8.6952636282395,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.020872737797380494,
      "loss": 0.8266,
      "num_input_tokens_seen": 33894024,
      "step": 58380
    },
    {
      "epoch": 8.696008340780459,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.020870943733037557,
      "loss": 0.7907,
      "num_input_tokens_seen": 33896744,
      "step": 58385
    },
    {
      "epoch": 8.696753053321418,
      "grad_norm": 0.02734375,
      "learning_rate": 0.020869149569510994,
      "loss": 0.8217,
      "num_input_tokens_seen": 33899880,
      "step": 58390
    },
    {
      "epoch": 8.697497765862376,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.020867355306831117,
      "loss": 0.7787,
      "num_input_tokens_seen": 33903080,
      "step": 58395
    },
    {
      "epoch": 8.698242478403337,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.02086556094502823,
      "loss": 0.7912,
      "num_input_tokens_seen": 33906024,
      "step": 58400
    },
    {
      "epoch": 8.698987190944296,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.020863766484132656,
      "loss": 0.7999,
      "num_input_tokens_seen": 33908904,
      "step": 58405
    },
    {
      "epoch": 8.699731903485254,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.020861971924174706,
      "loss": 0.8008,
      "num_input_tokens_seen": 33911944,
      "step": 58410
    },
    {
      "epoch": 8.700476616026213,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.020860177265184694,
      "loss": 0.8051,
      "num_input_tokens_seen": 33914824,
      "step": 58415
    },
    {
      "epoch": 8.701221328567174,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.020858382507192946,
      "loss": 0.8189,
      "num_input_tokens_seen": 33917544,
      "step": 58420
    },
    {
      "epoch": 8.701966041108133,
      "grad_norm": 0.02734375,
      "learning_rate": 0.020856587650229783,
      "loss": 0.7627,
      "num_input_tokens_seen": 33920168,
      "step": 58425
    },
    {
      "epoch": 8.702710753649091,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.02085479269432552,
      "loss": 0.8231,
      "num_input_tokens_seen": 33923176,
      "step": 58430
    },
    {
      "epoch": 8.70345546619005,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.020852997639510488,
      "loss": 0.8035,
      "num_input_tokens_seen": 33926120,
      "step": 58435
    },
    {
      "epoch": 8.70420017873101,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02085120248581501,
      "loss": 0.8245,
      "num_input_tokens_seen": 33928744,
      "step": 58440
    },
    {
      "epoch": 8.70494489127197,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.020849407233269408,
      "loss": 0.8052,
      "num_input_tokens_seen": 33931464,
      "step": 58445
    },
    {
      "epoch": 8.705689603812928,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02084761188190402,
      "loss": 0.7979,
      "num_input_tokens_seen": 33934024,
      "step": 58450
    },
    {
      "epoch": 8.706434316353887,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.020845816431749167,
      "loss": 0.8138,
      "num_input_tokens_seen": 33936840,
      "step": 58455
    },
    {
      "epoch": 8.707179028894847,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02084402088283519,
      "loss": 0.8015,
      "num_input_tokens_seen": 33939688,
      "step": 58460
    },
    {
      "epoch": 8.707923741435806,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.020842225235192427,
      "loss": 0.7882,
      "num_input_tokens_seen": 33942408,
      "step": 58465
    },
    {
      "epoch": 8.708668453976765,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.0208404294888512,
      "loss": 0.7888,
      "num_input_tokens_seen": 33945608,
      "step": 58470
    },
    {
      "epoch": 8.709413166517724,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.020838633643841854,
      "loss": 0.808,
      "num_input_tokens_seen": 33948584,
      "step": 58475
    },
    {
      "epoch": 8.710157879058684,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.020836837700194726,
      "loss": 0.8211,
      "num_input_tokens_seen": 33951368,
      "step": 58480
    },
    {
      "epoch": 8.710902591599643,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.020835041657940152,
      "loss": 0.793,
      "num_input_tokens_seen": 33954344,
      "step": 58485
    },
    {
      "epoch": 8.711647304140602,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.020833245517108488,
      "loss": 0.8076,
      "num_input_tokens_seen": 33957192,
      "step": 58490
    },
    {
      "epoch": 8.71239201668156,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.020831449277730063,
      "loss": 0.7811,
      "num_input_tokens_seen": 33959912,
      "step": 58495
    },
    {
      "epoch": 8.71313672922252,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.020829652939835235,
      "loss": 0.8002,
      "num_input_tokens_seen": 33963016,
      "step": 58500
    },
    {
      "epoch": 8.71388144176348,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.020827856503454343,
      "loss": 0.7949,
      "num_input_tokens_seen": 33966184,
      "step": 58505
    },
    {
      "epoch": 8.714626154304439,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.020826059968617743,
      "loss": 0.7926,
      "num_input_tokens_seen": 33969768,
      "step": 58510
    },
    {
      "epoch": 8.715370866845397,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.020824263335355774,
      "loss": 0.7999,
      "num_input_tokens_seen": 33972520,
      "step": 58515
    },
    {
      "epoch": 8.716115579386356,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.020822466603698797,
      "loss": 0.821,
      "num_input_tokens_seen": 33975592,
      "step": 58520
    },
    {
      "epoch": 8.716860291927317,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.020820669773677164,
      "loss": 0.7933,
      "num_input_tokens_seen": 33978600,
      "step": 58525
    },
    {
      "epoch": 8.717605004468275,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.020818872845321232,
      "loss": 0.7928,
      "num_input_tokens_seen": 33981640,
      "step": 58530
    },
    {
      "epoch": 8.718349717009234,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.020817075818661362,
      "loss": 0.7836,
      "num_input_tokens_seen": 33984648,
      "step": 58535
    },
    {
      "epoch": 8.719094429550193,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.020815278693727905,
      "loss": 0.7992,
      "num_input_tokens_seen": 33987336,
      "step": 58540
    },
    {
      "epoch": 8.719839142091153,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.02081348147055122,
      "loss": 0.8058,
      "num_input_tokens_seen": 33990248,
      "step": 58545
    },
    {
      "epoch": 8.720583854632112,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.02081168414916168,
      "loss": 0.7826,
      "num_input_tokens_seen": 33993320,
      "step": 58550
    },
    {
      "epoch": 8.721328567173071,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.020809886729589642,
      "loss": 0.803,
      "num_input_tokens_seen": 33996232,
      "step": 58555
    },
    {
      "epoch": 8.72207327971403,
      "grad_norm": 0.0169677734375,
      "learning_rate": 0.02080808921186547,
      "loss": 0.7852,
      "num_input_tokens_seen": 33999016,
      "step": 58560
    },
    {
      "epoch": 8.72281799225499,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.020806291596019537,
      "loss": 0.8046,
      "num_input_tokens_seen": 34002056,
      "step": 58565
    },
    {
      "epoch": 8.723562704795949,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02080449388208221,
      "loss": 0.808,
      "num_input_tokens_seen": 34005256,
      "step": 58570
    },
    {
      "epoch": 8.724307417336908,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.020802696070083854,
      "loss": 0.8214,
      "num_input_tokens_seen": 34007656,
      "step": 58575
    },
    {
      "epoch": 8.725052129877866,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.020800898160054846,
      "loss": 0.7893,
      "num_input_tokens_seen": 34010600,
      "step": 58580
    },
    {
      "epoch": 8.725796842418827,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.02079910015202556,
      "loss": 0.799,
      "num_input_tokens_seen": 34013640,
      "step": 58585
    },
    {
      "epoch": 8.726541554959786,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.020797302046026374,
      "loss": 0.8023,
      "num_input_tokens_seen": 34016552,
      "step": 58590
    },
    {
      "epoch": 8.727286267500745,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02079550384208766,
      "loss": 0.8041,
      "num_input_tokens_seen": 34019720,
      "step": 58595
    },
    {
      "epoch": 8.728030980041703,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.0207937055402398,
      "loss": 0.7963,
      "num_input_tokens_seen": 34022376,
      "step": 58600
    },
    {
      "epoch": 8.728775692582664,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.020791907140513172,
      "loss": 0.8199,
      "num_input_tokens_seen": 34025320,
      "step": 58605
    },
    {
      "epoch": 8.729520405123623,
      "grad_norm": 0.01495361328125,
      "learning_rate": 0.020790108642938156,
      "loss": 0.7896,
      "num_input_tokens_seen": 34028008,
      "step": 58610
    },
    {
      "epoch": 8.730265117664581,
      "grad_norm": 0.02734375,
      "learning_rate": 0.020788310047545142,
      "loss": 0.8255,
      "num_input_tokens_seen": 34030792,
      "step": 58615
    },
    {
      "epoch": 8.73100983020554,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.020786511354364514,
      "loss": 0.7776,
      "num_input_tokens_seen": 34033800,
      "step": 58620
    },
    {
      "epoch": 8.7317545427465,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.020784712563426656,
      "loss": 0.7962,
      "num_input_tokens_seen": 34036520,
      "step": 58625
    },
    {
      "epoch": 8.73249925528746,
      "grad_norm": 0.025390625,
      "learning_rate": 0.020782913674761965,
      "loss": 0.8484,
      "num_input_tokens_seen": 34039240,
      "step": 58630
    },
    {
      "epoch": 8.733243967828418,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.02078111468840082,
      "loss": 0.816,
      "num_input_tokens_seen": 34041864,
      "step": 58635
    },
    {
      "epoch": 8.733988680369377,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.020779315604373617,
      "loss": 0.7941,
      "num_input_tokens_seen": 34044648,
      "step": 58640
    },
    {
      "epoch": 8.734733392910336,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.02077751642271075,
      "loss": 0.8065,
      "num_input_tokens_seen": 34047560,
      "step": 58645
    },
    {
      "epoch": 8.735478105451296,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.020775717143442615,
      "loss": 0.7956,
      "num_input_tokens_seen": 34050152,
      "step": 58650
    },
    {
      "epoch": 8.736222817992255,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02077391776659961,
      "loss": 0.8057,
      "num_input_tokens_seen": 34053096,
      "step": 58655
    },
    {
      "epoch": 8.736967530533214,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.020772118292212132,
      "loss": 0.8047,
      "num_input_tokens_seen": 34056040,
      "step": 58660
    },
    {
      "epoch": 8.737712243074174,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02077031872031059,
      "loss": 0.8248,
      "num_input_tokens_seen": 34058984,
      "step": 58665
    },
    {
      "epoch": 8.738456955615133,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.02076851905092537,
      "loss": 0.7998,
      "num_input_tokens_seen": 34061960,
      "step": 58670
    },
    {
      "epoch": 8.739201668156092,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.020766719284086882,
      "loss": 0.797,
      "num_input_tokens_seen": 34064872,
      "step": 58675
    },
    {
      "epoch": 8.73994638069705,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.020764919419825535,
      "loss": 0.8138,
      "num_input_tokens_seen": 34067496,
      "step": 58680
    },
    {
      "epoch": 8.74069109323801,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.020763119458171733,
      "loss": 0.7971,
      "num_input_tokens_seen": 34070504,
      "step": 58685
    },
    {
      "epoch": 8.74143580577897,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.020761319399155886,
      "loss": 0.8185,
      "num_input_tokens_seen": 34073576,
      "step": 58690
    },
    {
      "epoch": 8.742180518319929,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.020759519242808402,
      "loss": 0.8159,
      "num_input_tokens_seen": 34076328,
      "step": 58695
    },
    {
      "epoch": 8.742925230860887,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.020757718989159697,
      "loss": 0.7841,
      "num_input_tokens_seen": 34078952,
      "step": 58700
    },
    {
      "epoch": 8.743669943401846,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.020755918638240178,
      "loss": 0.8084,
      "num_input_tokens_seen": 34081800,
      "step": 58705
    },
    {
      "epoch": 8.744414655942807,
      "grad_norm": 0.01513671875,
      "learning_rate": 0.020754118190080266,
      "loss": 0.7939,
      "num_input_tokens_seen": 34085128,
      "step": 58710
    },
    {
      "epoch": 8.745159368483765,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.020752317644710378,
      "loss": 0.8054,
      "num_input_tokens_seen": 34087816,
      "step": 58715
    },
    {
      "epoch": 8.745904081024724,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.020750517002160924,
      "loss": 0.8133,
      "num_input_tokens_seen": 34090920,
      "step": 58720
    },
    {
      "epoch": 8.746648793565683,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.020748716262462333,
      "loss": 0.8085,
      "num_input_tokens_seen": 34093608,
      "step": 58725
    },
    {
      "epoch": 8.747393506106643,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.020746915425645024,
      "loss": 0.8142,
      "num_input_tokens_seen": 34096552,
      "step": 58730
    },
    {
      "epoch": 8.748138218647602,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02074511449173942,
      "loss": 0.7898,
      "num_input_tokens_seen": 34099336,
      "step": 58735
    },
    {
      "epoch": 8.748882931188561,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02074331346077594,
      "loss": 0.7959,
      "num_input_tokens_seen": 34102280,
      "step": 58740
    },
    {
      "epoch": 8.74962764372952,
      "grad_norm": 0.0234375,
      "learning_rate": 0.020741512332785022,
      "loss": 0.7841,
      "num_input_tokens_seen": 34105192,
      "step": 58745
    },
    {
      "epoch": 8.75037235627048,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.02073971110779709,
      "loss": 0.78,
      "num_input_tokens_seen": 34108040,
      "step": 58750
    },
    {
      "epoch": 8.751117068811439,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.02073790978584257,
      "loss": 0.812,
      "num_input_tokens_seen": 34111112,
      "step": 58755
    },
    {
      "epoch": 8.751861781352398,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.020736108366951896,
      "loss": 0.8078,
      "num_input_tokens_seen": 34114408,
      "step": 58760
    },
    {
      "epoch": 8.752606493893357,
      "grad_norm": 0.029296875,
      "learning_rate": 0.020734306851155498,
      "loss": 0.8066,
      "num_input_tokens_seen": 34117160,
      "step": 58765
    },
    {
      "epoch": 8.753351206434317,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02073250523848382,
      "loss": 0.8253,
      "num_input_tokens_seen": 34120072,
      "step": 58770
    },
    {
      "epoch": 8.754095918975276,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.02073070352896729,
      "loss": 0.816,
      "num_input_tokens_seen": 34123112,
      "step": 58775
    },
    {
      "epoch": 8.754840631516235,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.020728901722636344,
      "loss": 0.8134,
      "num_input_tokens_seen": 34125864,
      "step": 58780
    },
    {
      "epoch": 8.755585344057193,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.02072709981952143,
      "loss": 0.7937,
      "num_input_tokens_seen": 34129032,
      "step": 58785
    },
    {
      "epoch": 8.756330056598154,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.020725297819652988,
      "loss": 0.8038,
      "num_input_tokens_seen": 34131848,
      "step": 58790
    },
    {
      "epoch": 8.757074769139113,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.020723495723061452,
      "loss": 0.805,
      "num_input_tokens_seen": 34134920,
      "step": 58795
    },
    {
      "epoch": 8.757819481680071,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02072169352977728,
      "loss": 0.7953,
      "num_input_tokens_seen": 34137672,
      "step": 58800
    },
    {
      "epoch": 8.75856419422103,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.020719891239830906,
      "loss": 0.8071,
      "num_input_tokens_seen": 34141160,
      "step": 58805
    },
    {
      "epoch": 8.75930890676199,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.02071808885325278,
      "loss": 0.8147,
      "num_input_tokens_seen": 34144232,
      "step": 58810
    },
    {
      "epoch": 8.76005361930295,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.020716286370073358,
      "loss": 0.796,
      "num_input_tokens_seen": 34147304,
      "step": 58815
    },
    {
      "epoch": 8.760798331843908,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.020714483790323087,
      "loss": 0.7978,
      "num_input_tokens_seen": 34149736,
      "step": 58820
    },
    {
      "epoch": 8.761543044384867,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.020712681114032423,
      "loss": 0.8099,
      "num_input_tokens_seen": 34152552,
      "step": 58825
    },
    {
      "epoch": 8.762287756925826,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.020710878341231816,
      "loss": 0.7844,
      "num_input_tokens_seen": 34155400,
      "step": 58830
    },
    {
      "epoch": 8.763032469466786,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.020709075471951725,
      "loss": 0.7815,
      "num_input_tokens_seen": 34158568,
      "step": 58835
    },
    {
      "epoch": 8.763777182007745,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.020707272506222604,
      "loss": 0.8031,
      "num_input_tokens_seen": 34161352,
      "step": 58840
    },
    {
      "epoch": 8.764521894548704,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.020705469444074916,
      "loss": 0.7981,
      "num_input_tokens_seen": 34164328,
      "step": 58845
    },
    {
      "epoch": 8.765266607089664,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.02070366628553912,
      "loss": 0.8001,
      "num_input_tokens_seen": 34167176,
      "step": 58850
    },
    {
      "epoch": 8.766011319630623,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.020701863030645674,
      "loss": 0.8121,
      "num_input_tokens_seen": 34170088,
      "step": 58855
    },
    {
      "epoch": 8.766756032171582,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.020700059679425054,
      "loss": 0.8169,
      "num_input_tokens_seen": 34172808,
      "step": 58860
    },
    {
      "epoch": 8.76750074471254,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.02069825623190772,
      "loss": 0.7649,
      "num_input_tokens_seen": 34175624,
      "step": 58865
    },
    {
      "epoch": 8.7682454572535,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.020696452688124138,
      "loss": 0.8068,
      "num_input_tokens_seen": 34178504,
      "step": 58870
    },
    {
      "epoch": 8.76899016979446,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.020694649048104773,
      "loss": 0.8049,
      "num_input_tokens_seen": 34181224,
      "step": 58875
    },
    {
      "epoch": 8.769734882335419,
      "grad_norm": 0.01373291015625,
      "learning_rate": 0.020692845311880103,
      "loss": 0.8171,
      "num_input_tokens_seen": 34184072,
      "step": 58880
    },
    {
      "epoch": 8.770479594876377,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.020691041479480594,
      "loss": 0.7859,
      "num_input_tokens_seen": 34186600,
      "step": 58885
    },
    {
      "epoch": 8.771224307417336,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.020689237550936726,
      "loss": 0.7839,
      "num_input_tokens_seen": 34189160,
      "step": 58890
    },
    {
      "epoch": 8.771969019958297,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.020687433526278966,
      "loss": 0.8052,
      "num_input_tokens_seen": 34192072,
      "step": 58895
    },
    {
      "epoch": 8.772713732499255,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02068562940553781,
      "loss": 0.7952,
      "num_input_tokens_seen": 34194920,
      "step": 58900
    },
    {
      "epoch": 8.773458445040214,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.020683825188743715,
      "loss": 0.7963,
      "num_input_tokens_seen": 34197672,
      "step": 58905
    },
    {
      "epoch": 8.774203157581173,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.020682020875927173,
      "loss": 0.8005,
      "num_input_tokens_seen": 34200648,
      "step": 58910
    },
    {
      "epoch": 8.774947870122134,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.020680216467118664,
      "loss": 0.8179,
      "num_input_tokens_seen": 34203848,
      "step": 58915
    },
    {
      "epoch": 8.775692582663092,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.020678411962348667,
      "loss": 0.8055,
      "num_input_tokens_seen": 34206888,
      "step": 58920
    },
    {
      "epoch": 8.776437295204051,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.020676607361647673,
      "loss": 0.8018,
      "num_input_tokens_seen": 34209768,
      "step": 58925
    },
    {
      "epoch": 8.77718200774501,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.020674802665046163,
      "loss": 0.8018,
      "num_input_tokens_seen": 34212648,
      "step": 58930
    },
    {
      "epoch": 8.77792672028597,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02067299787257464,
      "loss": 0.7944,
      "num_input_tokens_seen": 34215560,
      "step": 58935
    },
    {
      "epoch": 8.778671432826929,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.020671192984263575,
      "loss": 0.7904,
      "num_input_tokens_seen": 34218568,
      "step": 58940
    },
    {
      "epoch": 8.779416145367888,
      "grad_norm": 0.03125,
      "learning_rate": 0.02066938800014347,
      "loss": 0.7975,
      "num_input_tokens_seen": 34221288,
      "step": 58945
    },
    {
      "epoch": 8.780160857908847,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.020667582920244815,
      "loss": 0.7852,
      "num_input_tokens_seen": 34224072,
      "step": 58950
    },
    {
      "epoch": 8.780905570449807,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.020665777744598112,
      "loss": 0.7867,
      "num_input_tokens_seen": 34227240,
      "step": 58955
    },
    {
      "epoch": 8.781650282990766,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02066397247323385,
      "loss": 0.8062,
      "num_input_tokens_seen": 34230344,
      "step": 58960
    },
    {
      "epoch": 8.782394995531725,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.020662167106182534,
      "loss": 0.7651,
      "num_input_tokens_seen": 34233544,
      "step": 58965
    },
    {
      "epoch": 8.783139708072683,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02066036164347465,
      "loss": 0.7789,
      "num_input_tokens_seen": 34236296,
      "step": 58970
    },
    {
      "epoch": 8.783884420613644,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.020658556085140715,
      "loss": 0.8078,
      "num_input_tokens_seen": 34239240,
      "step": 58975
    },
    {
      "epoch": 8.784629133154603,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.020656750431211228,
      "loss": 0.8026,
      "num_input_tokens_seen": 34242024,
      "step": 58980
    },
    {
      "epoch": 8.785373845695561,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.020654944681716685,
      "loss": 0.791,
      "num_input_tokens_seen": 34245032,
      "step": 58985
    },
    {
      "epoch": 8.78611855823652,
      "grad_norm": 0.03515625,
      "learning_rate": 0.020653138836687603,
      "loss": 0.8424,
      "num_input_tokens_seen": 34247752,
      "step": 58990
    },
    {
      "epoch": 8.78686327077748,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.02065133289615449,
      "loss": 0.7714,
      "num_input_tokens_seen": 34250856,
      "step": 58995
    },
    {
      "epoch": 8.78760798331844,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.02064952686014785,
      "loss": 0.7765,
      "num_input_tokens_seen": 34253768,
      "step": 59000
    },
    {
      "epoch": 8.788352695859398,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.02064772072869819,
      "loss": 0.8155,
      "num_input_tokens_seen": 34256872,
      "step": 59005
    },
    {
      "epoch": 8.789097408400357,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.020645914501836032,
      "loss": 0.858,
      "num_input_tokens_seen": 34259720,
      "step": 59010
    },
    {
      "epoch": 8.789842120941316,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.020644108179591884,
      "loss": 0.8035,
      "num_input_tokens_seen": 34262408,
      "step": 59015
    },
    {
      "epoch": 8.790586833482276,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.020642301761996266,
      "loss": 0.7838,
      "num_input_tokens_seen": 34265416,
      "step": 59020
    },
    {
      "epoch": 8.791331546023235,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.020640495249079696,
      "loss": 0.7827,
      "num_input_tokens_seen": 34268776,
      "step": 59025
    },
    {
      "epoch": 8.792076258564194,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02063868864087269,
      "loss": 0.84,
      "num_input_tokens_seen": 34271752,
      "step": 59030
    },
    {
      "epoch": 8.792820971105153,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.020636881937405777,
      "loss": 0.8118,
      "num_input_tokens_seen": 34274568,
      "step": 59035
    },
    {
      "epoch": 8.793565683646113,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.020635075138709467,
      "loss": 0.8167,
      "num_input_tokens_seen": 34278088,
      "step": 59040
    },
    {
      "epoch": 8.794310396187072,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.020633268244814287,
      "loss": 0.7898,
      "num_input_tokens_seen": 34281032,
      "step": 59045
    },
    {
      "epoch": 8.79505510872803,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.020631461255750765,
      "loss": 0.7883,
      "num_input_tokens_seen": 34283976,
      "step": 59050
    },
    {
      "epoch": 8.79579982126899,
      "grad_norm": 0.025390625,
      "learning_rate": 0.02062965417154943,
      "loss": 0.8093,
      "num_input_tokens_seen": 34286952,
      "step": 59055
    },
    {
      "epoch": 8.79654453380995,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.020627846992240812,
      "loss": 0.7963,
      "num_input_tokens_seen": 34289928,
      "step": 59060
    },
    {
      "epoch": 8.797289246350909,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.02062603971785544,
      "loss": 0.8102,
      "num_input_tokens_seen": 34293224,
      "step": 59065
    },
    {
      "epoch": 8.798033958891867,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.020624232348423845,
      "loss": 0.7958,
      "num_input_tokens_seen": 34295880,
      "step": 59070
    },
    {
      "epoch": 8.798778671432826,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02062242488397656,
      "loss": 0.7942,
      "num_input_tokens_seen": 34298792,
      "step": 59075
    },
    {
      "epoch": 8.799523383973787,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.020620617324544118,
      "loss": 0.7935,
      "num_input_tokens_seen": 34301576,
      "step": 59080
    },
    {
      "epoch": 8.800268096514746,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.02061880967015706,
      "loss": 0.8073,
      "num_input_tokens_seen": 34304744,
      "step": 59085
    },
    {
      "epoch": 8.801012809055704,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.020617001920845923,
      "loss": 0.8329,
      "num_input_tokens_seen": 34307784,
      "step": 59090
    },
    {
      "epoch": 8.801757521596663,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02061519407664125,
      "loss": 0.8093,
      "num_input_tokens_seen": 34310600,
      "step": 59095
    },
    {
      "epoch": 8.802502234137624,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.020613386137573575,
      "loss": 0.8079,
      "num_input_tokens_seen": 34313384,
      "step": 59100
    },
    {
      "epoch": 8.803246946678582,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.020611578103673454,
      "loss": 0.7715,
      "num_input_tokens_seen": 34316392,
      "step": 59105
    },
    {
      "epoch": 8.803991659219541,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.020609769974971416,
      "loss": 0.7877,
      "num_input_tokens_seen": 34319112,
      "step": 59110
    },
    {
      "epoch": 8.8047363717605,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02060796175149802,
      "loss": 0.8065,
      "num_input_tokens_seen": 34322088,
      "step": 59115
    },
    {
      "epoch": 8.80548108430146,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.020606153433283814,
      "loss": 0.7818,
      "num_input_tokens_seen": 34324712,
      "step": 59120
    },
    {
      "epoch": 8.80622579684242,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.02060434502035934,
      "loss": 0.8035,
      "num_input_tokens_seen": 34327656,
      "step": 59125
    },
    {
      "epoch": 8.806970509383378,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.020602536512755152,
      "loss": 0.7854,
      "num_input_tokens_seen": 34330472,
      "step": 59130
    },
    {
      "epoch": 8.807715221924337,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.0206007279105018,
      "loss": 0.8043,
      "num_input_tokens_seen": 34333192,
      "step": 59135
    },
    {
      "epoch": 8.808459934465297,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02059891921362985,
      "loss": 0.8054,
      "num_input_tokens_seen": 34336200,
      "step": 59140
    },
    {
      "epoch": 8.809204647006256,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.020597110422169843,
      "loss": 0.8111,
      "num_input_tokens_seen": 34339112,
      "step": 59145
    },
    {
      "epoch": 8.809949359547215,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.020595301536152348,
      "loss": 0.8342,
      "num_input_tokens_seen": 34341928,
      "step": 59150
    },
    {
      "epoch": 8.810694072088173,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.020593492555607922,
      "loss": 0.8307,
      "num_input_tokens_seen": 34344776,
      "step": 59155
    },
    {
      "epoch": 8.811438784629132,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.020591683480567124,
      "loss": 0.7883,
      "num_input_tokens_seen": 34347816,
      "step": 59160
    },
    {
      "epoch": 8.812183497170093,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.020589874311060515,
      "loss": 0.7626,
      "num_input_tokens_seen": 34350568,
      "step": 59165
    },
    {
      "epoch": 8.812928209711052,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.020588065047118662,
      "loss": 0.7728,
      "num_input_tokens_seen": 34353384,
      "step": 59170
    },
    {
      "epoch": 8.81367292225201,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.020586255688772124,
      "loss": 0.7991,
      "num_input_tokens_seen": 34356200,
      "step": 59175
    },
    {
      "epoch": 8.81441763479297,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02058444623605148,
      "loss": 0.7817,
      "num_input_tokens_seen": 34359240,
      "step": 59180
    },
    {
      "epoch": 8.81516234733393,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.02058263668898729,
      "loss": 0.8096,
      "num_input_tokens_seen": 34361992,
      "step": 59185
    },
    {
      "epoch": 8.815907059874888,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.020580827047610127,
      "loss": 0.831,
      "num_input_tokens_seen": 34364776,
      "step": 59190
    },
    {
      "epoch": 8.816651772415847,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.020579017311950564,
      "loss": 0.8235,
      "num_input_tokens_seen": 34367880,
      "step": 59195
    },
    {
      "epoch": 8.817396484956806,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.020577207482039176,
      "loss": 0.8244,
      "num_input_tokens_seen": 34371048,
      "step": 59200
    },
    {
      "epoch": 8.818141197497766,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.020575397557906532,
      "loss": 0.7913,
      "num_input_tokens_seen": 34373736,
      "step": 59205
    },
    {
      "epoch": 8.818885910038725,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.020573587539583216,
      "loss": 0.779,
      "num_input_tokens_seen": 34376648,
      "step": 59210
    },
    {
      "epoch": 8.819630622579684,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.020571777427099803,
      "loss": 0.7768,
      "num_input_tokens_seen": 34379368,
      "step": 59215
    },
    {
      "epoch": 8.820375335120643,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.02056996722048687,
      "loss": 0.772,
      "num_input_tokens_seen": 34382152,
      "step": 59220
    },
    {
      "epoch": 8.821120047661603,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.020568156919775003,
      "loss": 0.7998,
      "num_input_tokens_seen": 34385032,
      "step": 59225
    },
    {
      "epoch": 8.821864760202562,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.020566346524994783,
      "loss": 0.7987,
      "num_input_tokens_seen": 34388072,
      "step": 59230
    },
    {
      "epoch": 8.82260947274352,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.020564536036176806,
      "loss": 0.7872,
      "num_input_tokens_seen": 34390856,
      "step": 59235
    },
    {
      "epoch": 8.82335418528448,
      "grad_norm": 0.033203125,
      "learning_rate": 0.020562725453351638,
      "loss": 0.8017,
      "num_input_tokens_seen": 34393672,
      "step": 59240
    },
    {
      "epoch": 8.82409889782544,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.020560914776549876,
      "loss": 0.7969,
      "num_input_tokens_seen": 34396616,
      "step": 59245
    },
    {
      "epoch": 8.824843610366399,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.020559104005802114,
      "loss": 0.8118,
      "num_input_tokens_seen": 34399336,
      "step": 59250
    },
    {
      "epoch": 8.825588322907358,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.020557293141138938,
      "loss": 0.7695,
      "num_input_tokens_seen": 34402184,
      "step": 59255
    },
    {
      "epoch": 8.826333035448316,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.020555482182590946,
      "loss": 0.7831,
      "num_input_tokens_seen": 34405064,
      "step": 59260
    },
    {
      "epoch": 8.827077747989277,
      "grad_norm": 0.01483154296875,
      "learning_rate": 0.02055367113018873,
      "loss": 0.8301,
      "num_input_tokens_seen": 34407656,
      "step": 59265
    },
    {
      "epoch": 8.827822460530236,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.02055185998396288,
      "loss": 0.8108,
      "num_input_tokens_seen": 34410376,
      "step": 59270
    },
    {
      "epoch": 8.828567173071194,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.020550048743944,
      "loss": 0.7853,
      "num_input_tokens_seen": 34413352,
      "step": 59275
    },
    {
      "epoch": 8.829311885612153,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.02054823741016269,
      "loss": 0.8227,
      "num_input_tokens_seen": 34416424,
      "step": 59280
    },
    {
      "epoch": 8.830056598153114,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02054642598264954,
      "loss": 0.7935,
      "num_input_tokens_seen": 34419624,
      "step": 59285
    },
    {
      "epoch": 8.830801310694072,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.020544614461435167,
      "loss": 0.8093,
      "num_input_tokens_seen": 34422280,
      "step": 59290
    },
    {
      "epoch": 8.831546023235031,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.02054280284655017,
      "loss": 0.788,
      "num_input_tokens_seen": 34425032,
      "step": 59295
    },
    {
      "epoch": 8.83229073577599,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.02054099113802515,
      "loss": 0.8186,
      "num_input_tokens_seen": 34427848,
      "step": 59300
    },
    {
      "epoch": 8.83303544831695,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.020539179335890712,
      "loss": 0.797,
      "num_input_tokens_seen": 34430600,
      "step": 59305
    },
    {
      "epoch": 8.83378016085791,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.020537367440177475,
      "loss": 0.7893,
      "num_input_tokens_seen": 34433384,
      "step": 59310
    },
    {
      "epoch": 8.834524873398868,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.020535555450916038,
      "loss": 0.809,
      "num_input_tokens_seen": 34436264,
      "step": 59315
    },
    {
      "epoch": 8.835269585939827,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02053374336813702,
      "loss": 0.7976,
      "num_input_tokens_seen": 34439112,
      "step": 59320
    },
    {
      "epoch": 8.836014298480787,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.020531931191871033,
      "loss": 0.7959,
      "num_input_tokens_seen": 34442056,
      "step": 59325
    },
    {
      "epoch": 8.836759011021746,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.02053011892214869,
      "loss": 0.8253,
      "num_input_tokens_seen": 34444808,
      "step": 59330
    },
    {
      "epoch": 8.837503723562705,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02052830655900061,
      "loss": 0.7976,
      "num_input_tokens_seen": 34448040,
      "step": 59335
    },
    {
      "epoch": 8.838248436103664,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.020526494102457406,
      "loss": 0.8177,
      "num_input_tokens_seen": 34451016,
      "step": 59340
    },
    {
      "epoch": 8.838993148644622,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0205246815525497,
      "loss": 0.7887,
      "num_input_tokens_seen": 34454088,
      "step": 59345
    },
    {
      "epoch": 8.839737861185583,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.020522868909308115,
      "loss": 0.7999,
      "num_input_tokens_seen": 34456936,
      "step": 59350
    },
    {
      "epoch": 8.840482573726542,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.02052105617276328,
      "loss": 0.826,
      "num_input_tokens_seen": 34460168,
      "step": 59355
    },
    {
      "epoch": 8.8412272862675,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.020519243342945803,
      "loss": 0.8226,
      "num_input_tokens_seen": 34463112,
      "step": 59360
    },
    {
      "epoch": 8.84197199880846,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.020517430419886327,
      "loss": 0.7939,
      "num_input_tokens_seen": 34465896,
      "step": 59365
    },
    {
      "epoch": 8.84271671134942,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.020515617403615467,
      "loss": 0.8145,
      "num_input_tokens_seen": 34468776,
      "step": 59370
    },
    {
      "epoch": 8.843461423890378,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.020513804294163857,
      "loss": 0.7874,
      "num_input_tokens_seen": 34471624,
      "step": 59375
    },
    {
      "epoch": 8.844206136431337,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02051199109156213,
      "loss": 0.8057,
      "num_input_tokens_seen": 34474536,
      "step": 59380
    },
    {
      "epoch": 8.844950848972296,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02051017779584091,
      "loss": 0.8294,
      "num_input_tokens_seen": 34477512,
      "step": 59385
    },
    {
      "epoch": 8.845695561513256,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.020508364407030844,
      "loss": 0.8075,
      "num_input_tokens_seen": 34480168,
      "step": 59390
    },
    {
      "epoch": 8.846440274054215,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.020506550925162556,
      "loss": 0.8117,
      "num_input_tokens_seen": 34483144,
      "step": 59395
    },
    {
      "epoch": 8.847184986595174,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02050473735026669,
      "loss": 0.8194,
      "num_input_tokens_seen": 34485832,
      "step": 59400
    },
    {
      "epoch": 8.847929699136133,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.02050292368237388,
      "loss": 0.8148,
      "num_input_tokens_seen": 34488872,
      "step": 59405
    },
    {
      "epoch": 8.848674411677093,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.020501109921514764,
      "loss": 0.7882,
      "num_input_tokens_seen": 34491720,
      "step": 59410
    },
    {
      "epoch": 8.849419124218052,
      "grad_norm": 0.029296875,
      "learning_rate": 0.020499296067719986,
      "loss": 0.8207,
      "num_input_tokens_seen": 34494632,
      "step": 59415
    },
    {
      "epoch": 8.85016383675901,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.02049748212102019,
      "loss": 0.7999,
      "num_input_tokens_seen": 34497672,
      "step": 59420
    },
    {
      "epoch": 8.85090854929997,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.020495668081446025,
      "loss": 0.7911,
      "num_input_tokens_seen": 34500424,
      "step": 59425
    },
    {
      "epoch": 8.85165326184093,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.020493853949028133,
      "loss": 0.8016,
      "num_input_tokens_seen": 34503208,
      "step": 59430
    },
    {
      "epoch": 8.852397974381889,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02049203972379716,
      "loss": 0.8075,
      "num_input_tokens_seen": 34506024,
      "step": 59435
    },
    {
      "epoch": 8.853142686922848,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.02049022540578376,
      "loss": 0.7938,
      "num_input_tokens_seen": 34508968,
      "step": 59440
    },
    {
      "epoch": 8.853887399463806,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.02048841099501858,
      "loss": 0.8097,
      "num_input_tokens_seen": 34511656,
      "step": 59445
    },
    {
      "epoch": 8.854632112004767,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.020486596491532276,
      "loss": 0.8192,
      "num_input_tokens_seen": 34514696,
      "step": 59450
    },
    {
      "epoch": 8.855376824545726,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.020484781895355495,
      "loss": 0.7994,
      "num_input_tokens_seen": 34517480,
      "step": 59455
    },
    {
      "epoch": 8.856121537086684,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.020482967206518907,
      "loss": 0.8188,
      "num_input_tokens_seen": 34520488,
      "step": 59460
    },
    {
      "epoch": 8.856866249627643,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.02048115242505316,
      "loss": 0.7786,
      "num_input_tokens_seen": 34523208,
      "step": 59465
    },
    {
      "epoch": 8.857610962168604,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.020479337550988904,
      "loss": 0.8185,
      "num_input_tokens_seen": 34526248,
      "step": 59470
    },
    {
      "epoch": 8.858355674709562,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.02047752258435682,
      "loss": 0.803,
      "num_input_tokens_seen": 34529000,
      "step": 59475
    },
    {
      "epoch": 8.859100387250521,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.020475707525187552,
      "loss": 0.8,
      "num_input_tokens_seen": 34531976,
      "step": 59480
    },
    {
      "epoch": 8.85984509979148,
      "grad_norm": 0.021484375,
      "learning_rate": 0.020473892373511774,
      "loss": 0.7831,
      "num_input_tokens_seen": 34534728,
      "step": 59485
    },
    {
      "epoch": 8.86058981233244,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.020472077129360146,
      "loss": 0.8007,
      "num_input_tokens_seen": 34537800,
      "step": 59490
    },
    {
      "epoch": 8.8613345248734,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.020470261792763338,
      "loss": 0.7768,
      "num_input_tokens_seen": 34540392,
      "step": 59495
    },
    {
      "epoch": 8.862079237414358,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.020468446363752016,
      "loss": 0.7803,
      "num_input_tokens_seen": 34543240,
      "step": 59500
    },
    {
      "epoch": 8.862823949955317,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.02046663084235685,
      "loss": 0.8179,
      "num_input_tokens_seen": 34545960,
      "step": 59505
    },
    {
      "epoch": 8.863568662496277,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.02046481522860851,
      "loss": 0.8113,
      "num_input_tokens_seen": 34548616,
      "step": 59510
    },
    {
      "epoch": 8.864313375037236,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.02046299952253768,
      "loss": 0.82,
      "num_input_tokens_seen": 34551432,
      "step": 59515
    },
    {
      "epoch": 8.865058087578195,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.020461183724175016,
      "loss": 0.7993,
      "num_input_tokens_seen": 34554216,
      "step": 59520
    },
    {
      "epoch": 8.865802800119154,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.020459367833551206,
      "loss": 0.8056,
      "num_input_tokens_seen": 34557576,
      "step": 59525
    },
    {
      "epoch": 8.866547512660112,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.020457551850696927,
      "loss": 0.8001,
      "num_input_tokens_seen": 34560328,
      "step": 59530
    },
    {
      "epoch": 8.867292225201073,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.020455735775642855,
      "loss": 0.7967,
      "num_input_tokens_seen": 34563176,
      "step": 59535
    },
    {
      "epoch": 8.868036937742032,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02045391960841967,
      "loss": 0.813,
      "num_input_tokens_seen": 34566184,
      "step": 59540
    },
    {
      "epoch": 8.86878165028299,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.020452103349058056,
      "loss": 0.7998,
      "num_input_tokens_seen": 34569000,
      "step": 59545
    },
    {
      "epoch": 8.86952636282395,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.0204502869975887,
      "loss": 0.8059,
      "num_input_tokens_seen": 34571720,
      "step": 59550
    },
    {
      "epoch": 8.87027107536491,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.020448470554042284,
      "loss": 0.8041,
      "num_input_tokens_seen": 34574376,
      "step": 59555
    },
    {
      "epoch": 8.871015787905868,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.020446654018449497,
      "loss": 0.8027,
      "num_input_tokens_seen": 34577032,
      "step": 59560
    },
    {
      "epoch": 8.871760500446827,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.020444837390841023,
      "loss": 0.7998,
      "num_input_tokens_seen": 34579944,
      "step": 59565
    },
    {
      "epoch": 8.872505212987786,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.020443020671247557,
      "loss": 0.8048,
      "num_input_tokens_seen": 34582504,
      "step": 59570
    },
    {
      "epoch": 8.873249925528746,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02044120385969979,
      "loss": 0.8005,
      "num_input_tokens_seen": 34585288,
      "step": 59575
    },
    {
      "epoch": 8.873994638069705,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.02043938695622841,
      "loss": 0.793,
      "num_input_tokens_seen": 34588168,
      "step": 59580
    },
    {
      "epoch": 8.874739350610664,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02043756996086412,
      "loss": 0.7969,
      "num_input_tokens_seen": 34590792,
      "step": 59585
    },
    {
      "epoch": 8.875484063151623,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.020435752873637605,
      "loss": 0.7834,
      "num_input_tokens_seen": 34593576,
      "step": 59590
    },
    {
      "epoch": 8.876228775692583,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.020433935694579578,
      "loss": 0.8035,
      "num_input_tokens_seen": 34596616,
      "step": 59595
    },
    {
      "epoch": 8.876973488233542,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.02043211842372073,
      "loss": 0.7933,
      "num_input_tokens_seen": 34599656,
      "step": 59600
    },
    {
      "epoch": 8.8777182007745,
      "grad_norm": 0.0147705078125,
      "learning_rate": 0.020430301061091758,
      "loss": 0.8042,
      "num_input_tokens_seen": 34602408,
      "step": 59605
    },
    {
      "epoch": 8.87846291331546,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.020428483606723367,
      "loss": 0.7989,
      "num_input_tokens_seen": 34605448,
      "step": 59610
    },
    {
      "epoch": 8.87920762585642,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.02042666606064627,
      "loss": 0.7982,
      "num_input_tokens_seen": 34608520,
      "step": 59615
    },
    {
      "epoch": 8.879952338397379,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.02042484842289116,
      "loss": 0.8085,
      "num_input_tokens_seen": 34611752,
      "step": 59620
    },
    {
      "epoch": 8.880697050938338,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02042303069348875,
      "loss": 0.8166,
      "num_input_tokens_seen": 34614568,
      "step": 59625
    },
    {
      "epoch": 8.881441763479296,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.02042121287246975,
      "loss": 0.7511,
      "num_input_tokens_seen": 34617256,
      "step": 59630
    },
    {
      "epoch": 8.882186476020257,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.020419394959864865,
      "loss": 0.7886,
      "num_input_tokens_seen": 34620168,
      "step": 59635
    },
    {
      "epoch": 8.882931188561216,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.020417576955704814,
      "loss": 0.8079,
      "num_input_tokens_seen": 34623048,
      "step": 59640
    },
    {
      "epoch": 8.883675901102174,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.020415758860020304,
      "loss": 0.7828,
      "num_input_tokens_seen": 34625896,
      "step": 59645
    },
    {
      "epoch": 8.884420613643133,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.02041394067284205,
      "loss": 0.8304,
      "num_input_tokens_seen": 34628744,
      "step": 59650
    },
    {
      "epoch": 8.885165326184094,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.020412122394200777,
      "loss": 0.8491,
      "num_input_tokens_seen": 34631464,
      "step": 59655
    },
    {
      "epoch": 8.885910038725052,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.020410304024127197,
      "loss": 0.8016,
      "num_input_tokens_seen": 34634376,
      "step": 59660
    },
    {
      "epoch": 8.886654751266011,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.020408485562652023,
      "loss": 0.8022,
      "num_input_tokens_seen": 34637192,
      "step": 59665
    },
    {
      "epoch": 8.88739946380697,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.020406667009805986,
      "loss": 0.8152,
      "num_input_tokens_seen": 34639848,
      "step": 59670
    },
    {
      "epoch": 8.88814417634793,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.020404848365619805,
      "loss": 0.7878,
      "num_input_tokens_seen": 34642472,
      "step": 59675
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.020403029630124206,
      "loss": 0.8058,
      "num_input_tokens_seen": 34645288,
      "step": 59680
    },
    {
      "epoch": 8.889633601429848,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.020401210803349907,
      "loss": 0.7962,
      "num_input_tokens_seen": 34648136,
      "step": 59685
    },
    {
      "epoch": 8.890378313970807,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.020399391885327647,
      "loss": 0.8005,
      "num_input_tokens_seen": 34650856,
      "step": 59690
    },
    {
      "epoch": 8.891123026511767,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.020397572876088146,
      "loss": 0.8195,
      "num_input_tokens_seen": 34653800,
      "step": 59695
    },
    {
      "epoch": 8.891867739052726,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.020395753775662134,
      "loss": 0.8248,
      "num_input_tokens_seen": 34656712,
      "step": 59700
    },
    {
      "epoch": 8.892612451593685,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.02039393458408035,
      "loss": 0.8184,
      "num_input_tokens_seen": 34659528,
      "step": 59705
    },
    {
      "epoch": 8.893357164134644,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.020392115301373522,
      "loss": 0.8179,
      "num_input_tokens_seen": 34662216,
      "step": 59710
    },
    {
      "epoch": 8.894101876675602,
      "grad_norm": 0.03125,
      "learning_rate": 0.020390295927572386,
      "loss": 0.8072,
      "num_input_tokens_seen": 34665192,
      "step": 59715
    },
    {
      "epoch": 8.894846589216563,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02038847646270768,
      "loss": 0.8162,
      "num_input_tokens_seen": 34667784,
      "step": 59720
    },
    {
      "epoch": 8.895591301757522,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.02038665690681014,
      "loss": 0.8028,
      "num_input_tokens_seen": 34670824,
      "step": 59725
    },
    {
      "epoch": 8.89633601429848,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.020384837259910513,
      "loss": 0.7954,
      "num_input_tokens_seen": 34673608,
      "step": 59730
    },
    {
      "epoch": 8.89708072683944,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.020383017522039526,
      "loss": 0.8141,
      "num_input_tokens_seen": 34676584,
      "step": 59735
    },
    {
      "epoch": 8.8978254393804,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.020381197693227927,
      "loss": 0.7867,
      "num_input_tokens_seen": 34679368,
      "step": 59740
    },
    {
      "epoch": 8.898570151921358,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.020379377773506464,
      "loss": 0.8037,
      "num_input_tokens_seen": 34682248,
      "step": 59745
    },
    {
      "epoch": 8.899314864462317,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.020377557762905882,
      "loss": 0.8207,
      "num_input_tokens_seen": 34685032,
      "step": 59750
    },
    {
      "epoch": 8.900059577003276,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.020375737661456925,
      "loss": 0.8029,
      "num_input_tokens_seen": 34687688,
      "step": 59755
    },
    {
      "epoch": 8.900804289544237,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.02037391746919035,
      "loss": 0.7822,
      "num_input_tokens_seen": 34690728,
      "step": 59760
    },
    {
      "epoch": 8.901549002085195,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.0203720971861369,
      "loss": 0.8049,
      "num_input_tokens_seen": 34693864,
      "step": 59765
    },
    {
      "epoch": 8.902293714626154,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.020370276812327323,
      "loss": 0.7944,
      "num_input_tokens_seen": 34696904,
      "step": 59770
    },
    {
      "epoch": 8.903038427167113,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.02036845634779238,
      "loss": 0.7994,
      "num_input_tokens_seen": 34699944,
      "step": 59775
    },
    {
      "epoch": 8.903783139708073,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.020366635792562822,
      "loss": 0.8147,
      "num_input_tokens_seen": 34703048,
      "step": 59780
    },
    {
      "epoch": 8.904527852249032,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.020364815146669404,
      "loss": 0.8073,
      "num_input_tokens_seen": 34705928,
      "step": 59785
    },
    {
      "epoch": 8.90527256478999,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.020362994410142894,
      "loss": 0.8273,
      "num_input_tokens_seen": 34708968,
      "step": 59790
    },
    {
      "epoch": 8.90601727733095,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02036117358301404,
      "loss": 0.7983,
      "num_input_tokens_seen": 34711816,
      "step": 59795
    },
    {
      "epoch": 8.90676198987191,
      "grad_norm": 0.041015625,
      "learning_rate": 0.02035935266531361,
      "loss": 0.8022,
      "num_input_tokens_seen": 34714440,
      "step": 59800
    },
    {
      "epoch": 8.907506702412869,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.020357531657072363,
      "loss": 0.7953,
      "num_input_tokens_seen": 34717320,
      "step": 59805
    },
    {
      "epoch": 8.908251414953828,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.02035571055832106,
      "loss": 0.8028,
      "num_input_tokens_seen": 34720328,
      "step": 59810
    },
    {
      "epoch": 8.908996127494786,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02035388936909047,
      "loss": 0.8051,
      "num_input_tokens_seen": 34723560,
      "step": 59815
    },
    {
      "epoch": 8.909740840035747,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.020352068089411365,
      "loss": 0.7883,
      "num_input_tokens_seen": 34726568,
      "step": 59820
    },
    {
      "epoch": 8.910485552576706,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.02035024671931451,
      "loss": 0.8084,
      "num_input_tokens_seen": 34729320,
      "step": 59825
    },
    {
      "epoch": 8.911230265117664,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.02034842525883067,
      "loss": 0.8111,
      "num_input_tokens_seen": 34732104,
      "step": 59830
    },
    {
      "epoch": 8.911974977658623,
      "grad_norm": 0.015625,
      "learning_rate": 0.020346603707990622,
      "loss": 0.8007,
      "num_input_tokens_seen": 34734856,
      "step": 59835
    },
    {
      "epoch": 8.912719690199584,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.02034478206682514,
      "loss": 0.7894,
      "num_input_tokens_seen": 34737576,
      "step": 59840
    },
    {
      "epoch": 8.913464402740543,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.020342960335364996,
      "loss": 0.8004,
      "num_input_tokens_seen": 34740584,
      "step": 59845
    },
    {
      "epoch": 8.914209115281501,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.020341138513640968,
      "loss": 0.7957,
      "num_input_tokens_seen": 34743656,
      "step": 59850
    },
    {
      "epoch": 8.91495382782246,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.020339316601683834,
      "loss": 0.8051,
      "num_input_tokens_seen": 34746568,
      "step": 59855
    },
    {
      "epoch": 8.915698540363419,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.020337494599524372,
      "loss": 0.7988,
      "num_input_tokens_seen": 34749736,
      "step": 59860
    },
    {
      "epoch": 8.91644325290438,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.020335672507193363,
      "loss": 0.8016,
      "num_input_tokens_seen": 34752648,
      "step": 59865
    },
    {
      "epoch": 8.917187965445338,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.02033385032472159,
      "loss": 0.7935,
      "num_input_tokens_seen": 34755880,
      "step": 59870
    },
    {
      "epoch": 8.917932677986297,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.020332028052139836,
      "loss": 0.7853,
      "num_input_tokens_seen": 34758696,
      "step": 59875
    },
    {
      "epoch": 8.918677390527257,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02033020568947889,
      "loss": 0.802,
      "num_input_tokens_seen": 34761480,
      "step": 59880
    },
    {
      "epoch": 8.919422103068216,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.020328383236769533,
      "loss": 0.8105,
      "num_input_tokens_seen": 34764744,
      "step": 59885
    },
    {
      "epoch": 8.920166815609175,
      "grad_norm": 0.029296875,
      "learning_rate": 0.02032656069404256,
      "loss": 0.8025,
      "num_input_tokens_seen": 34767496,
      "step": 59890
    },
    {
      "epoch": 8.920911528150134,
      "grad_norm": 0.03515625,
      "learning_rate": 0.02032473806132876,
      "loss": 0.7843,
      "num_input_tokens_seen": 34771560,
      "step": 59895
    },
    {
      "epoch": 8.921656240691092,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.020322915338658914,
      "loss": 0.7966,
      "num_input_tokens_seen": 34774504,
      "step": 59900
    },
    {
      "epoch": 8.922400953232053,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.020321092526063827,
      "loss": 0.7725,
      "num_input_tokens_seen": 34777352,
      "step": 59905
    },
    {
      "epoch": 8.923145665773012,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.02031926962357429,
      "loss": 0.7892,
      "num_input_tokens_seen": 34780008,
      "step": 59910
    },
    {
      "epoch": 8.92389037831397,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0203174466312211,
      "loss": 0.8186,
      "num_input_tokens_seen": 34783144,
      "step": 59915
    },
    {
      "epoch": 8.92463509085493,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.020315623549035055,
      "loss": 0.7918,
      "num_input_tokens_seen": 34785960,
      "step": 59920
    },
    {
      "epoch": 8.92537980339589,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.02031380037704695,
      "loss": 0.8192,
      "num_input_tokens_seen": 34788680,
      "step": 59925
    },
    {
      "epoch": 8.926124515936849,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.02031197711528759,
      "loss": 0.7665,
      "num_input_tokens_seen": 34791400,
      "step": 59930
    },
    {
      "epoch": 8.926869228477807,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.020310153763787777,
      "loss": 0.7905,
      "num_input_tokens_seen": 34794440,
      "step": 59935
    },
    {
      "epoch": 8.927613941018766,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.020308330322578307,
      "loss": 0.8169,
      "num_input_tokens_seen": 34797768,
      "step": 59940
    },
    {
      "epoch": 8.928358653559727,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.020306506791689993,
      "loss": 0.8287,
      "num_input_tokens_seen": 34800456,
      "step": 59945
    },
    {
      "epoch": 8.929103366100685,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.020304683171153647,
      "loss": 0.8062,
      "num_input_tokens_seen": 34803464,
      "step": 59950
    },
    {
      "epoch": 8.929848078641644,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.020302859461000064,
      "loss": 0.8105,
      "num_input_tokens_seen": 34806472,
      "step": 59955
    },
    {
      "epoch": 8.930592791182603,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.020301035661260063,
      "loss": 0.8127,
      "num_input_tokens_seen": 34809480,
      "step": 59960
    },
    {
      "epoch": 8.931337503723563,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.02029921177196445,
      "loss": 0.7942,
      "num_input_tokens_seen": 34812328,
      "step": 59965
    },
    {
      "epoch": 8.932082216264522,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.020297387793144044,
      "loss": 0.8056,
      "num_input_tokens_seen": 34814952,
      "step": 59970
    },
    {
      "epoch": 8.932826928805481,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.020295563724829654,
      "loss": 0.7992,
      "num_input_tokens_seen": 34817512,
      "step": 59975
    },
    {
      "epoch": 8.93357164134644,
      "grad_norm": 0.0546875,
      "learning_rate": 0.020293739567052087,
      "loss": 0.803,
      "num_input_tokens_seen": 34820424,
      "step": 59980
    },
    {
      "epoch": 8.9343163538874,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.02029191531984218,
      "loss": 0.8042,
      "num_input_tokens_seen": 34823112,
      "step": 59985
    },
    {
      "epoch": 8.935061066428359,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.020290090983230747,
      "loss": 0.8158,
      "num_input_tokens_seen": 34825800,
      "step": 59990
    },
    {
      "epoch": 8.935805778969318,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.020288266557248594,
      "loss": 0.7934,
      "num_input_tokens_seen": 34828648,
      "step": 59995
    },
    {
      "epoch": 8.936550491510276,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.020286442041926558,
      "loss": 0.7917,
      "num_input_tokens_seen": 34831496,
      "step": 60000
    },
    {
      "epoch": 8.937295204051237,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.02028461743729545,
      "loss": 0.7869,
      "num_input_tokens_seen": 34834248,
      "step": 60005
    },
    {
      "epoch": 8.938039916592196,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.020282792743386107,
      "loss": 0.7965,
      "num_input_tokens_seen": 34837064,
      "step": 60010
    },
    {
      "epoch": 8.938784629133155,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.020280967960229348,
      "loss": 0.8226,
      "num_input_tokens_seen": 34839944,
      "step": 60015
    },
    {
      "epoch": 8.939529341674113,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.020279143087856,
      "loss": 0.7746,
      "num_input_tokens_seen": 34843176,
      "step": 60020
    },
    {
      "epoch": 8.940274054215074,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0202773181262969,
      "loss": 0.7845,
      "num_input_tokens_seen": 34845960,
      "step": 60025
    },
    {
      "epoch": 8.941018766756033,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.020275493075582872,
      "loss": 0.8116,
      "num_input_tokens_seen": 34848872,
      "step": 60030
    },
    {
      "epoch": 8.941763479296991,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.020273667935744748,
      "loss": 0.8289,
      "num_input_tokens_seen": 34852168,
      "step": 60035
    },
    {
      "epoch": 8.94250819183795,
      "grad_norm": 0.259765625,
      "learning_rate": 0.02027184270681337,
      "loss": 0.8131,
      "num_input_tokens_seen": 34855144,
      "step": 60040
    },
    {
      "epoch": 8.943252904378909,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.02027001738881956,
      "loss": 0.8203,
      "num_input_tokens_seen": 34858152,
      "step": 60045
    },
    {
      "epoch": 8.94399761691987,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.020268191981794168,
      "loss": 0.8164,
      "num_input_tokens_seen": 34860872,
      "step": 60050
    },
    {
      "epoch": 8.944742329460828,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.02026636648576802,
      "loss": 0.7982,
      "num_input_tokens_seen": 34863560,
      "step": 60055
    },
    {
      "epoch": 8.945487042001787,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.020264540900771972,
      "loss": 0.8165,
      "num_input_tokens_seen": 34866600,
      "step": 60060
    },
    {
      "epoch": 8.946231754542747,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.020262715226836852,
      "loss": 0.7972,
      "num_input_tokens_seen": 34869608,
      "step": 60065
    },
    {
      "epoch": 8.946976467083706,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.020260889463993504,
      "loss": 0.8179,
      "num_input_tokens_seen": 34872584,
      "step": 60070
    },
    {
      "epoch": 8.947721179624665,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02025906361227278,
      "loss": 0.8048,
      "num_input_tokens_seen": 34875560,
      "step": 60075
    },
    {
      "epoch": 8.948465892165624,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.02025723767170552,
      "loss": 0.7948,
      "num_input_tokens_seen": 34878216,
      "step": 60080
    },
    {
      "epoch": 8.949210604706582,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.02025541164232257,
      "loss": 0.7936,
      "num_input_tokens_seen": 34880872,
      "step": 60085
    },
    {
      "epoch": 8.949955317247543,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.020253585524154787,
      "loss": 0.8135,
      "num_input_tokens_seen": 34884072,
      "step": 60090
    },
    {
      "epoch": 8.950700029788502,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.020251759317233015,
      "loss": 0.8079,
      "num_input_tokens_seen": 34886952,
      "step": 60095
    },
    {
      "epoch": 8.95144474232946,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.020249933021588104,
      "loss": 0.7943,
      "num_input_tokens_seen": 34889928,
      "step": 60100
    },
    {
      "epoch": 8.95218945487042,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.02024810663725091,
      "loss": 0.7951,
      "num_input_tokens_seen": 34892584,
      "step": 60105
    },
    {
      "epoch": 8.95293416741138,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.020246280164252287,
      "loss": 0.7995,
      "num_input_tokens_seen": 34895432,
      "step": 60110
    },
    {
      "epoch": 8.953678879952339,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.020244453602623096,
      "loss": 0.7962,
      "num_input_tokens_seen": 34898280,
      "step": 60115
    },
    {
      "epoch": 8.954423592493297,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.02024262695239419,
      "loss": 0.8009,
      "num_input_tokens_seen": 34901256,
      "step": 60120
    },
    {
      "epoch": 8.955168305034256,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.020240800213596435,
      "loss": 0.8355,
      "num_input_tokens_seen": 34904072,
      "step": 60125
    },
    {
      "epoch": 8.955913017575217,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.020238973386260684,
      "loss": 0.8013,
      "num_input_tokens_seen": 34906888,
      "step": 60130
    },
    {
      "epoch": 8.956657730116175,
      "grad_norm": 0.03125,
      "learning_rate": 0.0202371464704178,
      "loss": 0.791,
      "num_input_tokens_seen": 34909992,
      "step": 60135
    },
    {
      "epoch": 8.957402442657134,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.02023531946609865,
      "loss": 0.7823,
      "num_input_tokens_seen": 34912840,
      "step": 60140
    },
    {
      "epoch": 8.958147155198093,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0202334923733341,
      "loss": 0.7907,
      "num_input_tokens_seen": 34915944,
      "step": 60145
    },
    {
      "epoch": 8.958891867739053,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.02023166519215501,
      "loss": 0.7782,
      "num_input_tokens_seen": 34918696,
      "step": 60150
    },
    {
      "epoch": 8.959636580280012,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02022983792259226,
      "loss": 0.8354,
      "num_input_tokens_seen": 34921736,
      "step": 60155
    },
    {
      "epoch": 8.960381292820971,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.020228010564676713,
      "loss": 0.8216,
      "num_input_tokens_seen": 34924872,
      "step": 60160
    },
    {
      "epoch": 8.96112600536193,
      "grad_norm": 0.021484375,
      "learning_rate": 0.020226183118439243,
      "loss": 0.7935,
      "num_input_tokens_seen": 34927656,
      "step": 60165
    },
    {
      "epoch": 8.96187071790289,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.020224355583910718,
      "loss": 0.7936,
      "num_input_tokens_seen": 34930632,
      "step": 60170
    },
    {
      "epoch": 8.962615430443849,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.020222527961122012,
      "loss": 0.7799,
      "num_input_tokens_seen": 34933736,
      "step": 60175
    },
    {
      "epoch": 8.963360142984808,
      "grad_norm": 0.03125,
      "learning_rate": 0.020220700250104005,
      "loss": 0.8146,
      "num_input_tokens_seen": 34936552,
      "step": 60180
    },
    {
      "epoch": 8.964104855525767,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.020218872450887576,
      "loss": 0.8069,
      "num_input_tokens_seen": 34939432,
      "step": 60185
    },
    {
      "epoch": 8.964849568066727,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.0202170445635036,
      "loss": 0.821,
      "num_input_tokens_seen": 34941992,
      "step": 60190
    },
    {
      "epoch": 8.965594280607686,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02021521658798296,
      "loss": 0.823,
      "num_input_tokens_seen": 34944808,
      "step": 60195
    },
    {
      "epoch": 8.966338993148645,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.020213388524356536,
      "loss": 0.8038,
      "num_input_tokens_seen": 34948072,
      "step": 60200
    },
    {
      "epoch": 8.967083705689603,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.02021156037265521,
      "loss": 0.8173,
      "num_input_tokens_seen": 34950696,
      "step": 60205
    },
    {
      "epoch": 8.967828418230564,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.020209732132909868,
      "loss": 0.8108,
      "num_input_tokens_seen": 34953640,
      "step": 60210
    },
    {
      "epoch": 8.968573130771523,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0202079038051514,
      "loss": 0.7986,
      "num_input_tokens_seen": 34956392,
      "step": 60215
    },
    {
      "epoch": 8.969317843312481,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.020206075389410685,
      "loss": 0.8111,
      "num_input_tokens_seen": 34959336,
      "step": 60220
    },
    {
      "epoch": 8.97006255585344,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.020204246885718622,
      "loss": 0.7977,
      "num_input_tokens_seen": 34962024,
      "step": 60225
    },
    {
      "epoch": 8.970807268394399,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.020202418294106097,
      "loss": 0.8016,
      "num_input_tokens_seen": 34964968,
      "step": 60230
    },
    {
      "epoch": 8.97155198093536,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.020200589614604,
      "loss": 0.8046,
      "num_input_tokens_seen": 34967912,
      "step": 60235
    },
    {
      "epoch": 8.972296693476318,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.020198760847243227,
      "loss": 0.8155,
      "num_input_tokens_seen": 34970632,
      "step": 60240
    },
    {
      "epoch": 8.973041406017277,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.02019693199205468,
      "loss": 0.8031,
      "num_input_tokens_seen": 34973320,
      "step": 60245
    },
    {
      "epoch": 8.973786118558236,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.020195103049069237,
      "loss": 0.8052,
      "num_input_tokens_seen": 34976264,
      "step": 60250
    },
    {
      "epoch": 8.974530831099196,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.020193274018317817,
      "loss": 0.8041,
      "num_input_tokens_seen": 34978984,
      "step": 60255
    },
    {
      "epoch": 8.975275543640155,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.02019144489983131,
      "loss": 0.801,
      "num_input_tokens_seen": 34981928,
      "step": 60260
    },
    {
      "epoch": 8.976020256181114,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.020189615693640615,
      "loss": 0.7933,
      "num_input_tokens_seen": 34984680,
      "step": 60265
    },
    {
      "epoch": 8.976764968722073,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.020187786399776635,
      "loss": 0.809,
      "num_input_tokens_seen": 34987400,
      "step": 60270
    },
    {
      "epoch": 8.977509681263033,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02018595701827028,
      "loss": 0.8017,
      "num_input_tokens_seen": 34990376,
      "step": 60275
    },
    {
      "epoch": 8.978254393803992,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.020184127549152452,
      "loss": 0.8122,
      "num_input_tokens_seen": 34993128,
      "step": 60280
    },
    {
      "epoch": 8.97899910634495,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.020182297992454058,
      "loss": 0.8087,
      "num_input_tokens_seen": 34995848,
      "step": 60285
    },
    {
      "epoch": 8.97974381888591,
      "grad_norm": 0.0145263671875,
      "learning_rate": 0.020180468348206006,
      "loss": 0.8032,
      "num_input_tokens_seen": 34998568,
      "step": 60290
    },
    {
      "epoch": 8.98048853142687,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.020178638616439207,
      "loss": 0.7947,
      "num_input_tokens_seen": 35001512,
      "step": 60295
    },
    {
      "epoch": 8.981233243967829,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.020176808797184573,
      "loss": 0.8168,
      "num_input_tokens_seen": 35004680,
      "step": 60300
    },
    {
      "epoch": 8.981977956508787,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.02017497889047301,
      "loss": 0.7879,
      "num_input_tokens_seen": 35007560,
      "step": 60305
    },
    {
      "epoch": 8.982722669049746,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.020173148896335445,
      "loss": 0.8083,
      "num_input_tokens_seen": 35010344,
      "step": 60310
    },
    {
      "epoch": 8.983467381590707,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.020171318814802785,
      "loss": 0.7915,
      "num_input_tokens_seen": 35013160,
      "step": 60315
    },
    {
      "epoch": 8.984212094131665,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.02016948864590595,
      "loss": 0.8169,
      "num_input_tokens_seen": 35016008,
      "step": 60320
    },
    {
      "epoch": 8.984956806672624,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.020167658389675856,
      "loss": 0.7969,
      "num_input_tokens_seen": 35019208,
      "step": 60325
    },
    {
      "epoch": 8.985701519213583,
      "grad_norm": 0.02734375,
      "learning_rate": 0.02016582804614343,
      "loss": 0.8128,
      "num_input_tokens_seen": 35021928,
      "step": 60330
    },
    {
      "epoch": 8.986446231754543,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.020163997615339584,
      "loss": 0.8117,
      "num_input_tokens_seen": 35024904,
      "step": 60335
    },
    {
      "epoch": 8.987190944295502,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.02016216709729525,
      "loss": 0.8195,
      "num_input_tokens_seen": 35027848,
      "step": 60340
    },
    {
      "epoch": 8.987935656836461,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.02016033649204135,
      "loss": 0.8014,
      "num_input_tokens_seen": 35030856,
      "step": 60345
    },
    {
      "epoch": 8.98868036937742,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.020158505799608806,
      "loss": 0.7912,
      "num_input_tokens_seen": 35033800,
      "step": 60350
    },
    {
      "epoch": 8.98942508191838,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.020156675020028554,
      "loss": 0.8042,
      "num_input_tokens_seen": 35036712,
      "step": 60355
    },
    {
      "epoch": 8.990169794459339,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.020154844153331516,
      "loss": 0.8192,
      "num_input_tokens_seen": 35039336,
      "step": 60360
    },
    {
      "epoch": 8.990914507000298,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.020153013199548625,
      "loss": 0.8071,
      "num_input_tokens_seen": 35042152,
      "step": 60365
    },
    {
      "epoch": 8.991659219541257,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.020151182158710812,
      "loss": 0.7946,
      "num_input_tokens_seen": 35044936,
      "step": 60370
    },
    {
      "epoch": 8.992403932082215,
      "grad_norm": 0.021484375,
      "learning_rate": 0.020149351030849017,
      "loss": 0.7939,
      "num_input_tokens_seen": 35047816,
      "step": 60375
    },
    {
      "epoch": 8.993148644623176,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.020147519815994164,
      "loss": 0.82,
      "num_input_tokens_seen": 35050856,
      "step": 60380
    },
    {
      "epoch": 8.993893357164135,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.020145688514177196,
      "loss": 0.7908,
      "num_input_tokens_seen": 35053672,
      "step": 60385
    },
    {
      "epoch": 8.994638069705093,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.020143857125429057,
      "loss": 0.7995,
      "num_input_tokens_seen": 35056360,
      "step": 60390
    },
    {
      "epoch": 8.995382782246054,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.020142025649780675,
      "loss": 0.7864,
      "num_input_tokens_seen": 35059336,
      "step": 60395
    },
    {
      "epoch": 8.996127494787013,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.020140194087262996,
      "loss": 0.7866,
      "num_input_tokens_seen": 35062120,
      "step": 60400
    },
    {
      "epoch": 8.996872207327971,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.020138362437906964,
      "loss": 0.7781,
      "num_input_tokens_seen": 35064904,
      "step": 60405
    },
    {
      "epoch": 8.99761691986893,
      "grad_norm": 0.02734375,
      "learning_rate": 0.020136530701743517,
      "loss": 0.7967,
      "num_input_tokens_seen": 35068200,
      "step": 60410
    },
    {
      "epoch": 8.998361632409889,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.020134698878803606,
      "loss": 0.8183,
      "num_input_tokens_seen": 35071304,
      "step": 60415
    },
    {
      "epoch": 8.99910634495085,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.020132866969118177,
      "loss": 0.7962,
      "num_input_tokens_seen": 35074088,
      "step": 60420
    },
    {
      "epoch": 8.999851057491808,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.02013103497271818,
      "loss": 0.8132,
      "num_input_tokens_seen": 35076936,
      "step": 60425
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.8010132908821106,
      "eval_runtime": 70.6792,
      "eval_samples_per_second": 42.219,
      "eval_steps_per_second": 10.555,
      "num_input_tokens_seen": 35077032,
      "step": 60426
    },
    {
      "epoch": 9.000595770032767,
      "grad_norm": 0.025390625,
      "learning_rate": 0.020129202889634556,
      "loss": 0.7883,
      "num_input_tokens_seen": 35079464,
      "step": 60430
    },
    {
      "epoch": 9.001340482573726,
      "grad_norm": 0.012939453125,
      "learning_rate": 0.020127370719898268,
      "loss": 0.7973,
      "num_input_tokens_seen": 35082408,
      "step": 60435
    },
    {
      "epoch": 9.002085195114686,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.020125538463540257,
      "loss": 0.804,
      "num_input_tokens_seen": 35085608,
      "step": 60440
    },
    {
      "epoch": 9.002829907655645,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.02012370612059149,
      "loss": 0.7779,
      "num_input_tokens_seen": 35088904,
      "step": 60445
    },
    {
      "epoch": 9.003574620196604,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.020121873691082908,
      "loss": 0.7915,
      "num_input_tokens_seen": 35091816,
      "step": 60450
    },
    {
      "epoch": 9.004319332737563,
      "grad_norm": 0.01434326171875,
      "learning_rate": 0.02012004117504548,
      "loss": 0.7941,
      "num_input_tokens_seen": 35094760,
      "step": 60455
    },
    {
      "epoch": 9.005064045278523,
      "grad_norm": 0.03515625,
      "learning_rate": 0.020118208572510166,
      "loss": 0.7911,
      "num_input_tokens_seen": 35097512,
      "step": 60460
    },
    {
      "epoch": 9.005808757819482,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.02011637588350791,
      "loss": 0.8061,
      "num_input_tokens_seen": 35100456,
      "step": 60465
    },
    {
      "epoch": 9.00655347036044,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.020114543108069687,
      "loss": 0.8059,
      "num_input_tokens_seen": 35103304,
      "step": 60470
    },
    {
      "epoch": 9.0072981829014,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.020112710246226455,
      "loss": 0.8099,
      "num_input_tokens_seen": 35106216,
      "step": 60475
    },
    {
      "epoch": 9.00804289544236,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.020110877298009183,
      "loss": 0.8109,
      "num_input_tokens_seen": 35109384,
      "step": 60480
    },
    {
      "epoch": 9.008787607983319,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.02010904426344883,
      "loss": 0.778,
      "num_input_tokens_seen": 35112328,
      "step": 60485
    },
    {
      "epoch": 9.009532320524277,
      "grad_norm": 0.0234375,
      "learning_rate": 0.02010721114257637,
      "loss": 0.799,
      "num_input_tokens_seen": 35115176,
      "step": 60490
    },
    {
      "epoch": 9.010277033065236,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.02010537793542277,
      "loss": 0.792,
      "num_input_tokens_seen": 35118088,
      "step": 60495
    },
    {
      "epoch": 9.011021745606197,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.020103544642018994,
      "loss": 0.7882,
      "num_input_tokens_seen": 35120904,
      "step": 60500
    },
    {
      "epoch": 9.011766458147155,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.02010171126239602,
      "loss": 0.78,
      "num_input_tokens_seen": 35123592,
      "step": 60505
    },
    {
      "epoch": 9.012511170688114,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.02009987779658482,
      "loss": 0.8028,
      "num_input_tokens_seen": 35126408,
      "step": 60510
    },
    {
      "epoch": 9.013255883229073,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.020098044244616364,
      "loss": 0.8017,
      "num_input_tokens_seen": 35129512,
      "step": 60515
    },
    {
      "epoch": 9.014000595770034,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.020096210606521638,
      "loss": 0.7709,
      "num_input_tokens_seen": 35132360,
      "step": 60520
    },
    {
      "epoch": 9.014745308310992,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.020094376882331615,
      "loss": 0.8211,
      "num_input_tokens_seen": 35135272,
      "step": 60525
    },
    {
      "epoch": 9.015490020851951,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.020092543072077265,
      "loss": 0.8051,
      "num_input_tokens_seen": 35138248,
      "step": 60530
    },
    {
      "epoch": 9.01623473339291,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.02009070917578958,
      "loss": 0.7777,
      "num_input_tokens_seen": 35140936,
      "step": 60535
    },
    {
      "epoch": 9.01697944593387,
      "grad_norm": 0.029296875,
      "learning_rate": 0.020088875193499535,
      "loss": 0.8249,
      "num_input_tokens_seen": 35143752,
      "step": 60540
    },
    {
      "epoch": 9.017724158474829,
      "grad_norm": 0.029296875,
      "learning_rate": 0.020087041125238115,
      "loss": 0.7949,
      "num_input_tokens_seen": 35146536,
      "step": 60545
    },
    {
      "epoch": 9.018468871015788,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.020085206971036308,
      "loss": 0.7812,
      "num_input_tokens_seen": 35149800,
      "step": 60550
    },
    {
      "epoch": 9.019213583556747,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.020083372730925097,
      "loss": 0.8142,
      "num_input_tokens_seen": 35152872,
      "step": 60555
    },
    {
      "epoch": 9.019958296097707,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.020081538404935467,
      "loss": 0.7756,
      "num_input_tokens_seen": 35155752,
      "step": 60560
    },
    {
      "epoch": 9.020703008638666,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.02007970399309842,
      "loss": 0.8221,
      "num_input_tokens_seen": 35158792,
      "step": 60565
    },
    {
      "epoch": 9.021447721179625,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.020077869495444926,
      "loss": 0.7895,
      "num_input_tokens_seen": 35161832,
      "step": 60570
    },
    {
      "epoch": 9.022192433720583,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02007603491200599,
      "loss": 0.7758,
      "num_input_tokens_seen": 35164616,
      "step": 60575
    },
    {
      "epoch": 9.022937146261542,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.020074200242812604,
      "loss": 0.7953,
      "num_input_tokens_seen": 35167304,
      "step": 60580
    },
    {
      "epoch": 9.023681858802503,
      "grad_norm": 0.03125,
      "learning_rate": 0.020072365487895764,
      "loss": 0.7821,
      "num_input_tokens_seen": 35170312,
      "step": 60585
    },
    {
      "epoch": 9.024426571343461,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.020070530647286467,
      "loss": 0.7882,
      "num_input_tokens_seen": 35173288,
      "step": 60590
    },
    {
      "epoch": 9.02517128388442,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.020068695721015706,
      "loss": 0.8193,
      "num_input_tokens_seen": 35176424,
      "step": 60595
    },
    {
      "epoch": 9.025915996425379,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.020066860709114478,
      "loss": 0.826,
      "num_input_tokens_seen": 35179656,
      "step": 60600
    },
    {
      "epoch": 9.02666070896634,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.020065025611613795,
      "loss": 0.7762,
      "num_input_tokens_seen": 35182568,
      "step": 60605
    },
    {
      "epoch": 9.027405421507298,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.02006319042854465,
      "loss": 0.8104,
      "num_input_tokens_seen": 35185768,
      "step": 60610
    },
    {
      "epoch": 9.028150134048257,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.02006135515993804,
      "loss": 0.798,
      "num_input_tokens_seen": 35188456,
      "step": 60615
    },
    {
      "epoch": 9.028894846589216,
      "grad_norm": 0.041015625,
      "learning_rate": 0.020059519805824988,
      "loss": 0.8186,
      "num_input_tokens_seen": 35191720,
      "step": 60620
    },
    {
      "epoch": 9.029639559130176,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.02005768436623649,
      "loss": 0.8091,
      "num_input_tokens_seen": 35194664,
      "step": 60625
    },
    {
      "epoch": 9.030384271671135,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.020055848841203556,
      "loss": 0.803,
      "num_input_tokens_seen": 35197576,
      "step": 60630
    },
    {
      "epoch": 9.031128984212094,
      "grad_norm": 0.033203125,
      "learning_rate": 0.02005401323075719,
      "loss": 0.8121,
      "num_input_tokens_seen": 35200552,
      "step": 60635
    },
    {
      "epoch": 9.031873696753053,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.020052177534928408,
      "loss": 0.815,
      "num_input_tokens_seen": 35203432,
      "step": 60640
    },
    {
      "epoch": 9.032618409294013,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.020050341753748223,
      "loss": 0.817,
      "num_input_tokens_seen": 35206088,
      "step": 60645
    },
    {
      "epoch": 9.033363121834972,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.020048505887247648,
      "loss": 0.8041,
      "num_input_tokens_seen": 35208968,
      "step": 60650
    },
    {
      "epoch": 9.03410783437593,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.020046669935457697,
      "loss": 0.818,
      "num_input_tokens_seen": 35211496,
      "step": 60655
    },
    {
      "epoch": 9.03485254691689,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.020044833898409383,
      "loss": 0.7812,
      "num_input_tokens_seen": 35214184,
      "step": 60660
    },
    {
      "epoch": 9.03559725945785,
      "grad_norm": 0.01470947265625,
      "learning_rate": 0.020042997776133734,
      "loss": 0.7901,
      "num_input_tokens_seen": 35216968,
      "step": 60665
    },
    {
      "epoch": 9.036341971998809,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.020041161568661755,
      "loss": 0.8049,
      "num_input_tokens_seen": 35219592,
      "step": 60670
    },
    {
      "epoch": 9.037086684539767,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.020039325276024478,
      "loss": 0.8173,
      "num_input_tokens_seen": 35222568,
      "step": 60675
    },
    {
      "epoch": 9.037831397080726,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.020037488898252926,
      "loss": 0.8119,
      "num_input_tokens_seen": 35225384,
      "step": 60680
    },
    {
      "epoch": 9.038576109621687,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.020035652435378117,
      "loss": 0.8074,
      "num_input_tokens_seen": 35228200,
      "step": 60685
    },
    {
      "epoch": 9.039320822162646,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.02003381588743108,
      "loss": 0.8071,
      "num_input_tokens_seen": 35231016,
      "step": 60690
    },
    {
      "epoch": 9.040065534703604,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.020031979254442837,
      "loss": 0.7914,
      "num_input_tokens_seen": 35233896,
      "step": 60695
    },
    {
      "epoch": 9.040810247244563,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.020030142536444414,
      "loss": 0.8156,
      "num_input_tokens_seen": 35236552,
      "step": 60700
    },
    {
      "epoch": 9.041554959785524,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.020028305733466853,
      "loss": 0.8108,
      "num_input_tokens_seen": 35239304,
      "step": 60705
    },
    {
      "epoch": 9.042299672326482,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.020026468845541174,
      "loss": 0.7958,
      "num_input_tokens_seen": 35241864,
      "step": 60710
    },
    {
      "epoch": 9.043044384867441,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.02002463187269841,
      "loss": 0.8193,
      "num_input_tokens_seen": 35244744,
      "step": 60715
    },
    {
      "epoch": 9.0437890974084,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.020022794814969602,
      "loss": 0.8037,
      "num_input_tokens_seen": 35247656,
      "step": 60720
    },
    {
      "epoch": 9.04453380994936,
      "grad_norm": 0.0155029296875,
      "learning_rate": 0.020020957672385778,
      "loss": 0.788,
      "num_input_tokens_seen": 35250408,
      "step": 60725
    },
    {
      "epoch": 9.04527852249032,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.02001912044497798,
      "loss": 0.7965,
      "num_input_tokens_seen": 35253192,
      "step": 60730
    },
    {
      "epoch": 9.046023235031278,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.020017283132777244,
      "loss": 0.8028,
      "num_input_tokens_seen": 35256136,
      "step": 60735
    },
    {
      "epoch": 9.046767947572237,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.0200154457358146,
      "loss": 0.7951,
      "num_input_tokens_seen": 35259112,
      "step": 60740
    },
    {
      "epoch": 9.047512660113195,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.020013608254121105,
      "loss": 0.8034,
      "num_input_tokens_seen": 35262184,
      "step": 60745
    },
    {
      "epoch": 9.048257372654156,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.020011770687727793,
      "loss": 0.8355,
      "num_input_tokens_seen": 35264968,
      "step": 60750
    },
    {
      "epoch": 9.049002085195115,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.02000993303666571,
      "loss": 0.7989,
      "num_input_tokens_seen": 35267976,
      "step": 60755
    },
    {
      "epoch": 9.049746797736073,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.020008095300965898,
      "loss": 0.7946,
      "num_input_tokens_seen": 35271400,
      "step": 60760
    },
    {
      "epoch": 9.050491510277032,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.020006257480659408,
      "loss": 0.8043,
      "num_input_tokens_seen": 35274312,
      "step": 60765
    },
    {
      "epoch": 9.051236222817993,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.020004419575777286,
      "loss": 0.8188,
      "num_input_tokens_seen": 35277096,
      "step": 60770
    },
    {
      "epoch": 9.051980935358952,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.02000258158635058,
      "loss": 0.8089,
      "num_input_tokens_seen": 35280072,
      "step": 60775
    },
    {
      "epoch": 9.05272564789991,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.02000074351241034,
      "loss": 0.8037,
      "num_input_tokens_seen": 35282920,
      "step": 60780
    },
    {
      "epoch": 9.053470360440869,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.019998905353987625,
      "loss": 0.7914,
      "num_input_tokens_seen": 35285864,
      "step": 60785
    },
    {
      "epoch": 9.05421507298183,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.019997067111113483,
      "loss": 0.7996,
      "num_input_tokens_seen": 35288552,
      "step": 60790
    },
    {
      "epoch": 9.054959785522788,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.019995228783818972,
      "loss": 0.7865,
      "num_input_tokens_seen": 35291176,
      "step": 60795
    },
    {
      "epoch": 9.055704498063747,
      "grad_norm": 0.0234375,
      "learning_rate": 0.019993390372135145,
      "loss": 0.8349,
      "num_input_tokens_seen": 35294120,
      "step": 60800
    },
    {
      "epoch": 9.056449210604706,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.019991551876093066,
      "loss": 0.8014,
      "num_input_tokens_seen": 35297128,
      "step": 60805
    },
    {
      "epoch": 9.057193923145666,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.019989713295723792,
      "loss": 0.8096,
      "num_input_tokens_seen": 35300104,
      "step": 60810
    },
    {
      "epoch": 9.057938635686625,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.019987874631058382,
      "loss": 0.8217,
      "num_input_tokens_seen": 35303048,
      "step": 60815
    },
    {
      "epoch": 9.058683348227584,
      "grad_norm": 0.033203125,
      "learning_rate": 0.019986035882127898,
      "loss": 0.7914,
      "num_input_tokens_seen": 35306216,
      "step": 60820
    },
    {
      "epoch": 9.059428060768543,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.01998419704896341,
      "loss": 0.811,
      "num_input_tokens_seen": 35308904,
      "step": 60825
    },
    {
      "epoch": 9.060172773309503,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.019982358131595972,
      "loss": 0.7971,
      "num_input_tokens_seen": 35311880,
      "step": 60830
    },
    {
      "epoch": 9.060917485850462,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.01998051913005666,
      "loss": 0.8008,
      "num_input_tokens_seen": 35314984,
      "step": 60835
    },
    {
      "epoch": 9.06166219839142,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.01997868004437654,
      "loss": 0.7849,
      "num_input_tokens_seen": 35317832,
      "step": 60840
    },
    {
      "epoch": 9.06240691093238,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.019976840874586677,
      "loss": 0.8059,
      "num_input_tokens_seen": 35320776,
      "step": 60845
    },
    {
      "epoch": 9.06315162347334,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.01997500162071815,
      "loss": 0.8168,
      "num_input_tokens_seen": 35323592,
      "step": 60850
    },
    {
      "epoch": 9.063896336014299,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.019973162282802027,
      "loss": 0.7908,
      "num_input_tokens_seen": 35326856,
      "step": 60855
    },
    {
      "epoch": 9.064641048555258,
      "grad_norm": 0.0234375,
      "learning_rate": 0.019971322860869382,
      "loss": 0.8126,
      "num_input_tokens_seen": 35329544,
      "step": 60860
    },
    {
      "epoch": 9.065385761096216,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.01996948335495129,
      "loss": 0.8021,
      "num_input_tokens_seen": 35333000,
      "step": 60865
    },
    {
      "epoch": 9.066130473637177,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.019967643765078823,
      "loss": 0.7884,
      "num_input_tokens_seen": 35335816,
      "step": 60870
    },
    {
      "epoch": 9.066875186178136,
      "grad_norm": 0.037109375,
      "learning_rate": 0.019965804091283065,
      "loss": 0.8148,
      "num_input_tokens_seen": 35339112,
      "step": 60875
    },
    {
      "epoch": 9.067619898719094,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.019963964333595093,
      "loss": 0.8085,
      "num_input_tokens_seen": 35342344,
      "step": 60880
    },
    {
      "epoch": 9.068364611260053,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.019962124492045995,
      "loss": 0.8074,
      "num_input_tokens_seen": 35345352,
      "step": 60885
    },
    {
      "epoch": 9.069109323801014,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.019960284566666844,
      "loss": 0.7851,
      "num_input_tokens_seen": 35348232,
      "step": 60890
    },
    {
      "epoch": 9.069854036341972,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.019958444557488724,
      "loss": 0.8094,
      "num_input_tokens_seen": 35351400,
      "step": 60895
    },
    {
      "epoch": 9.070598748882931,
      "grad_norm": 0.02734375,
      "learning_rate": 0.019956604464542727,
      "loss": 0.8075,
      "num_input_tokens_seen": 35354184,
      "step": 60900
    },
    {
      "epoch": 9.07134346142389,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.019954764287859927,
      "loss": 0.7935,
      "num_input_tokens_seen": 35357192,
      "step": 60905
    },
    {
      "epoch": 9.07208817396485,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.019952924027471425,
      "loss": 0.7864,
      "num_input_tokens_seen": 35360072,
      "step": 60910
    },
    {
      "epoch": 9.07283288650581,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.019951083683408307,
      "loss": 0.7849,
      "num_input_tokens_seen": 35362728,
      "step": 60915
    },
    {
      "epoch": 9.073577599046768,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.019949243255701663,
      "loss": 0.8052,
      "num_input_tokens_seen": 35365512,
      "step": 60920
    },
    {
      "epoch": 9.074322311587727,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.019947402744382583,
      "loss": 0.7972,
      "num_input_tokens_seen": 35368296,
      "step": 60925
    },
    {
      "epoch": 9.075067024128685,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.019945562149482166,
      "loss": 0.8055,
      "num_input_tokens_seen": 35371496,
      "step": 60930
    },
    {
      "epoch": 9.075811736669646,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.019943721471031496,
      "loss": 0.7887,
      "num_input_tokens_seen": 35374184,
      "step": 60935
    },
    {
      "epoch": 9.076556449210605,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.019941880709061676,
      "loss": 0.7998,
      "num_input_tokens_seen": 35377064,
      "step": 60940
    },
    {
      "epoch": 9.077301161751564,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.019940039863603806,
      "loss": 0.8286,
      "num_input_tokens_seen": 35380040,
      "step": 60945
    },
    {
      "epoch": 9.078045874292522,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.019938198934688983,
      "loss": 0.8042,
      "num_input_tokens_seen": 35382920,
      "step": 60950
    },
    {
      "epoch": 9.078790586833483,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.019936357922348314,
      "loss": 0.8022,
      "num_input_tokens_seen": 35385672,
      "step": 60955
    },
    {
      "epoch": 9.079535299374442,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.01993451682661289,
      "loss": 0.8011,
      "num_input_tokens_seen": 35388712,
      "step": 60960
    },
    {
      "epoch": 9.0802800119154,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01993267564751382,
      "loss": 0.8178,
      "num_input_tokens_seen": 35391912,
      "step": 60965
    },
    {
      "epoch": 9.081024724456359,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.01993083438508221,
      "loss": 0.802,
      "num_input_tokens_seen": 35394600,
      "step": 60970
    },
    {
      "epoch": 9.08176943699732,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.019928993039349162,
      "loss": 0.811,
      "num_input_tokens_seen": 35397352,
      "step": 60975
    },
    {
      "epoch": 9.082514149538278,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.019927151610345785,
      "loss": 0.8032,
      "num_input_tokens_seen": 35399912,
      "step": 60980
    },
    {
      "epoch": 9.083258862079237,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.01992531009810319,
      "loss": 0.8097,
      "num_input_tokens_seen": 35402600,
      "step": 60985
    },
    {
      "epoch": 9.084003574620196,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01992346850265249,
      "loss": 0.7956,
      "num_input_tokens_seen": 35405512,
      "step": 60990
    },
    {
      "epoch": 9.084748287161156,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.019921626824024787,
      "loss": 0.8042,
      "num_input_tokens_seen": 35408200,
      "step": 60995
    },
    {
      "epoch": 9.085492999702115,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.019919785062251207,
      "loss": 0.81,
      "num_input_tokens_seen": 35411272,
      "step": 61000
    },
    {
      "epoch": 9.086237712243074,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.019917943217362857,
      "loss": 0.8011,
      "num_input_tokens_seen": 35414152,
      "step": 61005
    },
    {
      "epoch": 9.086982424784033,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.019916101289390857,
      "loss": 0.7945,
      "num_input_tokens_seen": 35416776,
      "step": 61010
    },
    {
      "epoch": 9.087727137324993,
      "grad_norm": 0.041015625,
      "learning_rate": 0.019914259278366316,
      "loss": 0.7999,
      "num_input_tokens_seen": 35419624,
      "step": 61015
    },
    {
      "epoch": 9.088471849865952,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.019912417184320366,
      "loss": 0.8113,
      "num_input_tokens_seen": 35422760,
      "step": 61020
    },
    {
      "epoch": 9.08921656240691,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.019910575007284118,
      "loss": 0.8017,
      "num_input_tokens_seen": 35425736,
      "step": 61025
    },
    {
      "epoch": 9.08996127494787,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.019908732747288698,
      "loss": 0.7989,
      "num_input_tokens_seen": 35428552,
      "step": 61030
    },
    {
      "epoch": 9.09070598748883,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.019906890404365222,
      "loss": 0.7841,
      "num_input_tokens_seen": 35431496,
      "step": 61035
    },
    {
      "epoch": 9.091450700029789,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.019905047978544824,
      "loss": 0.8095,
      "num_input_tokens_seen": 35434248,
      "step": 61040
    },
    {
      "epoch": 9.092195412570748,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.01990320546985862,
      "loss": 0.8155,
      "num_input_tokens_seen": 35437032,
      "step": 61045
    },
    {
      "epoch": 9.092940125111706,
      "grad_norm": 0.0234375,
      "learning_rate": 0.019901362878337752,
      "loss": 0.7998,
      "num_input_tokens_seen": 35439976,
      "step": 61050
    },
    {
      "epoch": 9.093684837652667,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.019899520204013334,
      "loss": 0.7793,
      "num_input_tokens_seen": 35442824,
      "step": 61055
    },
    {
      "epoch": 9.094429550193626,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.019897677446916503,
      "loss": 0.8039,
      "num_input_tokens_seen": 35445736,
      "step": 61060
    },
    {
      "epoch": 9.095174262734584,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.01989583460707839,
      "loss": 0.7938,
      "num_input_tokens_seen": 35448520,
      "step": 61065
    },
    {
      "epoch": 9.095918975275543,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.019893991684530123,
      "loss": 0.8051,
      "num_input_tokens_seen": 35451432,
      "step": 61070
    },
    {
      "epoch": 9.096663687816504,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.01989214867930284,
      "loss": 0.8127,
      "num_input_tokens_seen": 35454504,
      "step": 61075
    },
    {
      "epoch": 9.097408400357462,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01989030559142768,
      "loss": 0.8087,
      "num_input_tokens_seen": 35457640,
      "step": 61080
    },
    {
      "epoch": 9.098153112898421,
      "grad_norm": 0.041015625,
      "learning_rate": 0.019888462420935782,
      "loss": 0.8038,
      "num_input_tokens_seen": 35460872,
      "step": 61085
    },
    {
      "epoch": 9.09889782543938,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.01988661916785828,
      "loss": 0.7933,
      "num_input_tokens_seen": 35463688,
      "step": 61090
    },
    {
      "epoch": 9.099642537980339,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.01988477583222631,
      "loss": 0.7852,
      "num_input_tokens_seen": 35466440,
      "step": 61095
    },
    {
      "epoch": 9.1003872505213,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.019882932414071016,
      "loss": 0.8338,
      "num_input_tokens_seen": 35469448,
      "step": 61100
    },
    {
      "epoch": 9.101131963062258,
      "grad_norm": 0.033203125,
      "learning_rate": 0.019881088913423543,
      "loss": 0.8003,
      "num_input_tokens_seen": 35472552,
      "step": 61105
    },
    {
      "epoch": 9.101876675603217,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.019879245330315033,
      "loss": 0.8161,
      "num_input_tokens_seen": 35475944,
      "step": 61110
    },
    {
      "epoch": 9.102621388144176,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.019877401664776632,
      "loss": 0.8119,
      "num_input_tokens_seen": 35479336,
      "step": 61115
    },
    {
      "epoch": 9.103366100685136,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.01987555791683949,
      "loss": 0.8004,
      "num_input_tokens_seen": 35482088,
      "step": 61120
    },
    {
      "epoch": 9.104110813226095,
      "grad_norm": 0.03125,
      "learning_rate": 0.019873714086534745,
      "loss": 0.8077,
      "num_input_tokens_seen": 35485000,
      "step": 61125
    },
    {
      "epoch": 9.104855525767054,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.019871870173893563,
      "loss": 0.8032,
      "num_input_tokens_seen": 35487944,
      "step": 61130
    },
    {
      "epoch": 9.105600238308012,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01987002617894708,
      "loss": 0.7903,
      "num_input_tokens_seen": 35490760,
      "step": 61135
    },
    {
      "epoch": 9.106344950848973,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.019868182101726455,
      "loss": 0.789,
      "num_input_tokens_seen": 35493768,
      "step": 61140
    },
    {
      "epoch": 9.107089663389932,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.019866337942262842,
      "loss": 0.7993,
      "num_input_tokens_seen": 35496584,
      "step": 61145
    },
    {
      "epoch": 9.10783437593089,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.019864493700587394,
      "loss": 0.7829,
      "num_input_tokens_seen": 35499144,
      "step": 61150
    },
    {
      "epoch": 9.10857908847185,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.019862649376731274,
      "loss": 0.8127,
      "num_input_tokens_seen": 35501864,
      "step": 61155
    },
    {
      "epoch": 9.10932380101281,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.01986080497072563,
      "loss": 0.8004,
      "num_input_tokens_seen": 35504808,
      "step": 61160
    },
    {
      "epoch": 9.110068513553768,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.019858960482601627,
      "loss": 0.7889,
      "num_input_tokens_seen": 35507816,
      "step": 61165
    },
    {
      "epoch": 9.110813226094727,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01985711591239043,
      "loss": 0.7985,
      "num_input_tokens_seen": 35510824,
      "step": 61170
    },
    {
      "epoch": 9.111557938635686,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.01985527126012319,
      "loss": 0.7963,
      "num_input_tokens_seen": 35513800,
      "step": 61175
    },
    {
      "epoch": 9.112302651176647,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.019853426525831077,
      "loss": 0.7986,
      "num_input_tokens_seen": 35516520,
      "step": 61180
    },
    {
      "epoch": 9.113047363717605,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.01985158170954526,
      "loss": 0.813,
      "num_input_tokens_seen": 35519336,
      "step": 61185
    },
    {
      "epoch": 9.113792076258564,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.0198497368112969,
      "loss": 0.7891,
      "num_input_tokens_seen": 35522216,
      "step": 61190
    },
    {
      "epoch": 9.114536788799523,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.01984789183111716,
      "loss": 0.8084,
      "num_input_tokens_seen": 35524904,
      "step": 61195
    },
    {
      "epoch": 9.115281501340483,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.019846046769037222,
      "loss": 0.8089,
      "num_input_tokens_seen": 35527720,
      "step": 61200
    },
    {
      "epoch": 9.116026213881442,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.019844201625088244,
      "loss": 0.797,
      "num_input_tokens_seen": 35530856,
      "step": 61205
    },
    {
      "epoch": 9.1167709264224,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.019842356399301406,
      "loss": 0.8123,
      "num_input_tokens_seen": 35533832,
      "step": 61210
    },
    {
      "epoch": 9.11751563896336,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01984051109170788,
      "loss": 0.8397,
      "num_input_tokens_seen": 35536552,
      "step": 61215
    },
    {
      "epoch": 9.11826035150432,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.019838665702338835,
      "loss": 0.7967,
      "num_input_tokens_seen": 35539336,
      "step": 61220
    },
    {
      "epoch": 9.119005064045279,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.01983682023122545,
      "loss": 0.8167,
      "num_input_tokens_seen": 35542184,
      "step": 61225
    },
    {
      "epoch": 9.119749776586238,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.019834974678398905,
      "loss": 0.8066,
      "num_input_tokens_seen": 35545032,
      "step": 61230
    },
    {
      "epoch": 9.120494489127196,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.019833129043890375,
      "loss": 0.7882,
      "num_input_tokens_seen": 35548008,
      "step": 61235
    },
    {
      "epoch": 9.121239201668157,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.019831283327731042,
      "loss": 0.7867,
      "num_input_tokens_seen": 35550728,
      "step": 61240
    },
    {
      "epoch": 9.121983914209116,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.019829437529952088,
      "loss": 0.7989,
      "num_input_tokens_seen": 35553544,
      "step": 61245
    },
    {
      "epoch": 9.122728626750074,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0198275916505847,
      "loss": 0.8055,
      "num_input_tokens_seen": 35556712,
      "step": 61250
    },
    {
      "epoch": 9.123473339291033,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.019825745689660057,
      "loss": 0.7952,
      "num_input_tokens_seen": 35559432,
      "step": 61255
    },
    {
      "epoch": 9.124218051831992,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01982389964720934,
      "loss": 0.8028,
      "num_input_tokens_seen": 35562248,
      "step": 61260
    },
    {
      "epoch": 9.124962764372953,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.01982205352326374,
      "loss": 0.8009,
      "num_input_tokens_seen": 35565256,
      "step": 61265
    },
    {
      "epoch": 9.125707476913911,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.01982020731785445,
      "loss": 0.782,
      "num_input_tokens_seen": 35568200,
      "step": 61270
    },
    {
      "epoch": 9.12645218945487,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.01981836103101266,
      "loss": 0.7952,
      "num_input_tokens_seen": 35570984,
      "step": 61275
    },
    {
      "epoch": 9.127196901995829,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.01981651466276955,
      "loss": 0.8182,
      "num_input_tokens_seen": 35573992,
      "step": 61280
    },
    {
      "epoch": 9.12794161453679,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.01981466821315633,
      "loss": 0.7993,
      "num_input_tokens_seen": 35576872,
      "step": 61285
    },
    {
      "epoch": 9.128686327077748,
      "grad_norm": 0.02734375,
      "learning_rate": 0.01981282168220418,
      "loss": 0.7842,
      "num_input_tokens_seen": 35579944,
      "step": 61290
    },
    {
      "epoch": 9.129431039618707,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.019810975069944295,
      "loss": 0.7821,
      "num_input_tokens_seen": 35582568,
      "step": 61295
    },
    {
      "epoch": 9.130175752159666,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.019809128376407876,
      "loss": 0.7777,
      "num_input_tokens_seen": 35585544,
      "step": 61300
    },
    {
      "epoch": 9.130920464700626,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.019807281601626123,
      "loss": 0.8177,
      "num_input_tokens_seen": 35588648,
      "step": 61305
    },
    {
      "epoch": 9.131665177241585,
      "grad_norm": 0.0234375,
      "learning_rate": 0.01980543474563023,
      "loss": 0.8109,
      "num_input_tokens_seen": 35591688,
      "step": 61310
    },
    {
      "epoch": 9.132409889782544,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.019803587808451407,
      "loss": 0.8109,
      "num_input_tokens_seen": 35594760,
      "step": 61315
    },
    {
      "epoch": 9.133154602323502,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.01980174079012085,
      "loss": 0.7845,
      "num_input_tokens_seen": 35597576,
      "step": 61320
    },
    {
      "epoch": 9.133899314864463,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.019799893690669756,
      "loss": 0.8088,
      "num_input_tokens_seen": 35600392,
      "step": 61325
    },
    {
      "epoch": 9.134644027405422,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.01979804651012934,
      "loss": 0.8136,
      "num_input_tokens_seen": 35603432,
      "step": 61330
    },
    {
      "epoch": 9.13538873994638,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.01979619924853081,
      "loss": 0.8136,
      "num_input_tokens_seen": 35606120,
      "step": 61335
    },
    {
      "epoch": 9.13613345248734,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.019794351905905363,
      "loss": 0.7746,
      "num_input_tokens_seen": 35609256,
      "step": 61340
    },
    {
      "epoch": 9.1368781650283,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.01979250448228421,
      "loss": 0.7907,
      "num_input_tokens_seen": 35612360,
      "step": 61345
    },
    {
      "epoch": 9.137622877569259,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.019790656977698575,
      "loss": 0.821,
      "num_input_tokens_seen": 35615304,
      "step": 61350
    },
    {
      "epoch": 9.138367590110217,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.01978880939217965,
      "loss": 0.7934,
      "num_input_tokens_seen": 35618376,
      "step": 61355
    },
    {
      "epoch": 9.139112302651176,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.019786961725758664,
      "loss": 0.7653,
      "num_input_tokens_seen": 35621256,
      "step": 61360
    },
    {
      "epoch": 9.139857015192137,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.019785113978466818,
      "loss": 0.7963,
      "num_input_tokens_seen": 35624040,
      "step": 61365
    },
    {
      "epoch": 9.140601727733095,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.019783266150335335,
      "loss": 0.7957,
      "num_input_tokens_seen": 35626856,
      "step": 61370
    },
    {
      "epoch": 9.141346440274054,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.01978141824139544,
      "loss": 0.817,
      "num_input_tokens_seen": 35630152,
      "step": 61375
    },
    {
      "epoch": 9.142091152815013,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.019779570251678338,
      "loss": 0.7741,
      "num_input_tokens_seen": 35633128,
      "step": 61380
    },
    {
      "epoch": 9.142835865355973,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.019777722181215256,
      "loss": 0.8076,
      "num_input_tokens_seen": 35636360,
      "step": 61385
    },
    {
      "epoch": 9.143580577896932,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.01977587403003741,
      "loss": 0.8083,
      "num_input_tokens_seen": 35639144,
      "step": 61390
    },
    {
      "epoch": 9.14432529043789,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.01977402579817603,
      "loss": 0.8022,
      "num_input_tokens_seen": 35641928,
      "step": 61395
    },
    {
      "epoch": 9.14507000297885,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.01977217748566233,
      "loss": 0.8172,
      "num_input_tokens_seen": 35644680,
      "step": 61400
    },
    {
      "epoch": 9.14581471551981,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.019770329092527544,
      "loss": 0.7945,
      "num_input_tokens_seen": 35647720,
      "step": 61405
    },
    {
      "epoch": 9.146559428060769,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.0197684806188029,
      "loss": 0.7905,
      "num_input_tokens_seen": 35650472,
      "step": 61410
    },
    {
      "epoch": 9.147304140601728,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.019766632064519625,
      "loss": 0.7916,
      "num_input_tokens_seen": 35653160,
      "step": 61415
    },
    {
      "epoch": 9.148048853142686,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.019764783429708943,
      "loss": 0.7817,
      "num_input_tokens_seen": 35656456,
      "step": 61420
    },
    {
      "epoch": 9.148793565683647,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.019762934714402085,
      "loss": 0.7955,
      "num_input_tokens_seen": 35659240,
      "step": 61425
    },
    {
      "epoch": 9.149538278224606,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.019761085918630288,
      "loss": 0.7838,
      "num_input_tokens_seen": 35662056,
      "step": 61430
    },
    {
      "epoch": 9.150282990765565,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01975923704242478,
      "loss": 0.8079,
      "num_input_tokens_seen": 35664968,
      "step": 61435
    },
    {
      "epoch": 9.151027703306523,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.019757388085816806,
      "loss": 0.7745,
      "num_input_tokens_seen": 35667624,
      "step": 61440
    },
    {
      "epoch": 9.151772415847482,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.01975553904883759,
      "loss": 0.7861,
      "num_input_tokens_seen": 35670504,
      "step": 61445
    },
    {
      "epoch": 9.152517128388443,
      "grad_norm": 0.01953125,
      "learning_rate": 0.019753689931518374,
      "loss": 0.8043,
      "num_input_tokens_seen": 35673448,
      "step": 61450
    },
    {
      "epoch": 9.153261840929401,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.019751840733890406,
      "loss": 0.7926,
      "num_input_tokens_seen": 35676360,
      "step": 61455
    },
    {
      "epoch": 9.15400655347036,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.01974999145598491,
      "loss": 0.802,
      "num_input_tokens_seen": 35679304,
      "step": 61460
    },
    {
      "epoch": 9.154751266011319,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01974814209783314,
      "loss": 0.7824,
      "num_input_tokens_seen": 35682440,
      "step": 61465
    },
    {
      "epoch": 9.15549597855228,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.01974629265946633,
      "loss": 0.8042,
      "num_input_tokens_seen": 35685256,
      "step": 61470
    },
    {
      "epoch": 9.156240691093238,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.019744443140915734,
      "loss": 0.8305,
      "num_input_tokens_seen": 35688264,
      "step": 61475
    },
    {
      "epoch": 9.156985403634197,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.019742593542212594,
      "loss": 0.8101,
      "num_input_tokens_seen": 35691272,
      "step": 61480
    },
    {
      "epoch": 9.157730116175156,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.019740743863388158,
      "loss": 0.8288,
      "num_input_tokens_seen": 35694344,
      "step": 61485
    },
    {
      "epoch": 9.158474828716116,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.019738894104473666,
      "loss": 0.7793,
      "num_input_tokens_seen": 35697384,
      "step": 61490
    },
    {
      "epoch": 9.159219541257075,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.01973704426550038,
      "loss": 0.8119,
      "num_input_tokens_seen": 35700296,
      "step": 61495
    },
    {
      "epoch": 9.159964253798034,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.019735194346499545,
      "loss": 0.8068,
      "num_input_tokens_seen": 35702920,
      "step": 61500
    },
    {
      "epoch": 9.160708966338992,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.01973334434750241,
      "loss": 0.7814,
      "num_input_tokens_seen": 35705672,
      "step": 61505
    },
    {
      "epoch": 9.161453678879953,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.019731494268540234,
      "loss": 0.8173,
      "num_input_tokens_seen": 35708744,
      "step": 61510
    },
    {
      "epoch": 9.162198391420912,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01972964410964428,
      "loss": 0.7806,
      "num_input_tokens_seen": 35711752,
      "step": 61515
    },
    {
      "epoch": 9.16294310396187,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.019727793870845785,
      "loss": 0.7936,
      "num_input_tokens_seen": 35714696,
      "step": 61520
    },
    {
      "epoch": 9.16368781650283,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.019725943552176023,
      "loss": 0.8381,
      "num_input_tokens_seen": 35717576,
      "step": 61525
    },
    {
      "epoch": 9.16443252904379,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.019724093153666244,
      "loss": 0.8144,
      "num_input_tokens_seen": 35720616,
      "step": 61530
    },
    {
      "epoch": 9.165177241584749,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.019722242675347716,
      "loss": 0.8123,
      "num_input_tokens_seen": 35723368,
      "step": 61535
    },
    {
      "epoch": 9.165921954125707,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.019720392117251698,
      "loss": 0.8002,
      "num_input_tokens_seen": 35726120,
      "step": 61540
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.01971854147940945,
      "loss": 0.8093,
      "num_input_tokens_seen": 35728872,
      "step": 61545
    },
    {
      "epoch": 9.167411379207627,
      "grad_norm": 0.04296875,
      "learning_rate": 0.019716690761852243,
      "loss": 0.8124,
      "num_input_tokens_seen": 35731656,
      "step": 61550
    },
    {
      "epoch": 9.168156091748585,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.019714839964611337,
      "loss": 0.8002,
      "num_input_tokens_seen": 35734504,
      "step": 61555
    },
    {
      "epoch": 9.168900804289544,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.019712989087718003,
      "loss": 0.8026,
      "num_input_tokens_seen": 35737320,
      "step": 61560
    },
    {
      "epoch": 9.169645516830503,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.019711138131203502,
      "loss": 0.799,
      "num_input_tokens_seen": 35740392,
      "step": 61565
    },
    {
      "epoch": 9.170390229371463,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.019709287095099115,
      "loss": 0.8032,
      "num_input_tokens_seen": 35743272,
      "step": 61570
    },
    {
      "epoch": 9.171134941912422,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.019707435979436114,
      "loss": 0.8,
      "num_input_tokens_seen": 35746216,
      "step": 61575
    },
    {
      "epoch": 9.171879654453381,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.01970558478424576,
      "loss": 0.8046,
      "num_input_tokens_seen": 35749032,
      "step": 61580
    },
    {
      "epoch": 9.17262436699434,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.01970373350955934,
      "loss": 0.8051,
      "num_input_tokens_seen": 35752200,
      "step": 61585
    },
    {
      "epoch": 9.1733690795353,
      "grad_norm": 0.03515625,
      "learning_rate": 0.019701882155408117,
      "loss": 0.8026,
      "num_input_tokens_seen": 35754984,
      "step": 61590
    },
    {
      "epoch": 9.174113792076259,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.019700030721823372,
      "loss": 0.8013,
      "num_input_tokens_seen": 35757768,
      "step": 61595
    },
    {
      "epoch": 9.174858504617218,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01969817920883639,
      "loss": 0.8036,
      "num_input_tokens_seen": 35760360,
      "step": 61600
    },
    {
      "epoch": 9.175603217158177,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.01969632761647844,
      "loss": 0.7962,
      "num_input_tokens_seen": 35762824,
      "step": 61605
    },
    {
      "epoch": 9.176347929699135,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.019694475944780812,
      "loss": 0.8074,
      "num_input_tokens_seen": 35765736,
      "step": 61610
    },
    {
      "epoch": 9.177092642240096,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.019692624193774787,
      "loss": 0.7936,
      "num_input_tokens_seen": 35768968,
      "step": 61615
    },
    {
      "epoch": 9.177837354781055,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.019690772363491645,
      "loss": 0.8014,
      "num_input_tokens_seen": 35771880,
      "step": 61620
    },
    {
      "epoch": 9.178582067322013,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.019688920453962668,
      "loss": 0.8103,
      "num_input_tokens_seen": 35774760,
      "step": 61625
    },
    {
      "epoch": 9.179326779862972,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.019687068465219146,
      "loss": 0.8038,
      "num_input_tokens_seen": 35777960,
      "step": 61630
    },
    {
      "epoch": 9.180071492403933,
      "grad_norm": 0.01953125,
      "learning_rate": 0.019685216397292364,
      "loss": 0.8045,
      "num_input_tokens_seen": 35780936,
      "step": 61635
    },
    {
      "epoch": 9.180816204944891,
      "grad_norm": 0.037109375,
      "learning_rate": 0.019683364250213616,
      "loss": 0.7929,
      "num_input_tokens_seen": 35783688,
      "step": 61640
    },
    {
      "epoch": 9.18156091748585,
      "grad_norm": 0.0234375,
      "learning_rate": 0.019681512024014188,
      "loss": 0.8092,
      "num_input_tokens_seen": 35786632,
      "step": 61645
    },
    {
      "epoch": 9.182305630026809,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.019679659718725373,
      "loss": 0.8254,
      "num_input_tokens_seen": 35789640,
      "step": 61650
    },
    {
      "epoch": 9.18305034256777,
      "grad_norm": 0.017578125,
      "learning_rate": 0.019677807334378468,
      "loss": 0.7956,
      "num_input_tokens_seen": 35792392,
      "step": 61655
    },
    {
      "epoch": 9.183795055108728,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.019675954871004757,
      "loss": 0.795,
      "num_input_tokens_seen": 35795496,
      "step": 61660
    },
    {
      "epoch": 9.184539767649687,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.019674102328635534,
      "loss": 0.8099,
      "num_input_tokens_seen": 35798472,
      "step": 61665
    },
    {
      "epoch": 9.185284480190646,
      "grad_norm": 0.037109375,
      "learning_rate": 0.019672249707302112,
      "loss": 0.7846,
      "num_input_tokens_seen": 35801160,
      "step": 61670
    },
    {
      "epoch": 9.186029192731606,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.019670397007035777,
      "loss": 0.8074,
      "num_input_tokens_seen": 35804552,
      "step": 61675
    },
    {
      "epoch": 9.186773905272565,
      "grad_norm": 0.021484375,
      "learning_rate": 0.019668544227867835,
      "loss": 0.8105,
      "num_input_tokens_seen": 35807688,
      "step": 61680
    },
    {
      "epoch": 9.187518617813524,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.019666691369829577,
      "loss": 0.8087,
      "num_input_tokens_seen": 35810568,
      "step": 61685
    },
    {
      "epoch": 9.188263330354483,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.019664838432952317,
      "loss": 0.7839,
      "num_input_tokens_seen": 35813640,
      "step": 61690
    },
    {
      "epoch": 9.189008042895443,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.01966298541726735,
      "loss": 0.7883,
      "num_input_tokens_seen": 35816424,
      "step": 61695
    },
    {
      "epoch": 9.189752755436402,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.019661132322805985,
      "loss": 0.7883,
      "num_input_tokens_seen": 35819208,
      "step": 61700
    },
    {
      "epoch": 9.19049746797736,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.019659279149599525,
      "loss": 0.8077,
      "num_input_tokens_seen": 35822120,
      "step": 61705
    },
    {
      "epoch": 9.19124218051832,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.01965742589767928,
      "loss": 0.8215,
      "num_input_tokens_seen": 35825320,
      "step": 61710
    },
    {
      "epoch": 9.19198689305928,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.01965557256707656,
      "loss": 0.8124,
      "num_input_tokens_seen": 35828232,
      "step": 61715
    },
    {
      "epoch": 9.192731605600239,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.019653719157822668,
      "loss": 0.8002,
      "num_input_tokens_seen": 35830856,
      "step": 61720
    },
    {
      "epoch": 9.193476318141197,
      "grad_norm": 0.025390625,
      "learning_rate": 0.019651865669948924,
      "loss": 0.7794,
      "num_input_tokens_seen": 35833672,
      "step": 61725
    },
    {
      "epoch": 9.194221030682156,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.019650012103486637,
      "loss": 0.7801,
      "num_input_tokens_seen": 35836776,
      "step": 61730
    },
    {
      "epoch": 9.194965743223117,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.019648158458467124,
      "loss": 0.8123,
      "num_input_tokens_seen": 35839592,
      "step": 61735
    },
    {
      "epoch": 9.195710455764075,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.019646304734921696,
      "loss": 0.8141,
      "num_input_tokens_seen": 35842280,
      "step": 61740
    },
    {
      "epoch": 9.196455168305034,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.01964445093288167,
      "loss": 0.7972,
      "num_input_tokens_seen": 35845288,
      "step": 61745
    },
    {
      "epoch": 9.197199880845993,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.01964259705237837,
      "loss": 0.7974,
      "num_input_tokens_seen": 35848008,
      "step": 61750
    },
    {
      "epoch": 9.197944593386953,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.019640743093443106,
      "loss": 0.7987,
      "num_input_tokens_seen": 35850952,
      "step": 61755
    },
    {
      "epoch": 9.198689305927912,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.01963888905610721,
      "loss": 0.7987,
      "num_input_tokens_seen": 35853736,
      "step": 61760
    },
    {
      "epoch": 9.199434018468871,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.019637034940401994,
      "loss": 0.8314,
      "num_input_tokens_seen": 35856872,
      "step": 61765
    },
    {
      "epoch": 9.20017873100983,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.019635180746358786,
      "loss": 0.8019,
      "num_input_tokens_seen": 35859880,
      "step": 61770
    },
    {
      "epoch": 9.200923443550789,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.01963332647400891,
      "loss": 0.8099,
      "num_input_tokens_seen": 35862792,
      "step": 61775
    },
    {
      "epoch": 9.201668156091749,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.019631472123383693,
      "loss": 0.8031,
      "num_input_tokens_seen": 35865992,
      "step": 61780
    },
    {
      "epoch": 9.202412868632708,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.019629617694514465,
      "loss": 0.7935,
      "num_input_tokens_seen": 35869032,
      "step": 61785
    },
    {
      "epoch": 9.203157581173667,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.019627763187432548,
      "loss": 0.8159,
      "num_input_tokens_seen": 35872104,
      "step": 61790
    },
    {
      "epoch": 9.203902293714625,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.019625908602169277,
      "loss": 0.7916,
      "num_input_tokens_seen": 35874856,
      "step": 61795
    },
    {
      "epoch": 9.204647006255586,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.019624053938755978,
      "loss": 0.8094,
      "num_input_tokens_seen": 35877896,
      "step": 61800
    },
    {
      "epoch": 9.205391718796545,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01962219919722399,
      "loss": 0.8101,
      "num_input_tokens_seen": 35880936,
      "step": 61805
    },
    {
      "epoch": 9.206136431337503,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.019620344377604643,
      "loss": 0.8024,
      "num_input_tokens_seen": 35883688,
      "step": 61810
    },
    {
      "epoch": 9.206881143878462,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.019618489479929276,
      "loss": 0.8016,
      "num_input_tokens_seen": 35886472,
      "step": 61815
    },
    {
      "epoch": 9.207625856419423,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.019616634504229225,
      "loss": 0.799,
      "num_input_tokens_seen": 35889288,
      "step": 61820
    },
    {
      "epoch": 9.208370568960381,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.019614779450535823,
      "loss": 0.7819,
      "num_input_tokens_seen": 35892168,
      "step": 61825
    },
    {
      "epoch": 9.20911528150134,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.019612924318880412,
      "loss": 0.8344,
      "num_input_tokens_seen": 35895208,
      "step": 61830
    },
    {
      "epoch": 9.209859994042299,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.019611069109294332,
      "loss": 0.7934,
      "num_input_tokens_seen": 35897864,
      "step": 61835
    },
    {
      "epoch": 9.21060470658326,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.019609213821808927,
      "loss": 0.7936,
      "num_input_tokens_seen": 35900712,
      "step": 61840
    },
    {
      "epoch": 9.211349419124218,
      "grad_norm": 0.021484375,
      "learning_rate": 0.019607358456455542,
      "loss": 0.7954,
      "num_input_tokens_seen": 35903272,
      "step": 61845
    },
    {
      "epoch": 9.212094131665177,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.019605503013265517,
      "loss": 0.806,
      "num_input_tokens_seen": 35906056,
      "step": 61850
    },
    {
      "epoch": 9.212838844206136,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0196036474922702,
      "loss": 0.8176,
      "num_input_tokens_seen": 35908968,
      "step": 61855
    },
    {
      "epoch": 9.213583556747096,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.019601791893500937,
      "loss": 0.7855,
      "num_input_tokens_seen": 35911784,
      "step": 61860
    },
    {
      "epoch": 9.214328269288055,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.019599936216989074,
      "loss": 0.8224,
      "num_input_tokens_seen": 35914568,
      "step": 61865
    },
    {
      "epoch": 9.215072981829014,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.019598080462765965,
      "loss": 0.8126,
      "num_input_tokens_seen": 35917512,
      "step": 61870
    },
    {
      "epoch": 9.215817694369973,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.01959622463086296,
      "loss": 0.7921,
      "num_input_tokens_seen": 35920200,
      "step": 61875
    },
    {
      "epoch": 9.216562406910933,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.019594368721311418,
      "loss": 0.8336,
      "num_input_tokens_seen": 35923368,
      "step": 61880
    },
    {
      "epoch": 9.217307119451892,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.019592512734142677,
      "loss": 0.7898,
      "num_input_tokens_seen": 35926120,
      "step": 61885
    },
    {
      "epoch": 9.21805183199285,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.019590656669388102,
      "loss": 0.7898,
      "num_input_tokens_seen": 35929160,
      "step": 61890
    },
    {
      "epoch": 9.21879654453381,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.019588800527079054,
      "loss": 0.7934,
      "num_input_tokens_seen": 35931848,
      "step": 61895
    },
    {
      "epoch": 9.21954125707477,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.019586944307246883,
      "loss": 0.806,
      "num_input_tokens_seen": 35934792,
      "step": 61900
    },
    {
      "epoch": 9.220285969615729,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.019585088009922945,
      "loss": 0.7994,
      "num_input_tokens_seen": 35937896,
      "step": 61905
    },
    {
      "epoch": 9.221030682156687,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.019583231635138608,
      "loss": 0.7874,
      "num_input_tokens_seen": 35940776,
      "step": 61910
    },
    {
      "epoch": 9.221775394697646,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.019581375182925232,
      "loss": 0.8013,
      "num_input_tokens_seen": 35943592,
      "step": 61915
    },
    {
      "epoch": 9.222520107238607,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.019579518653314178,
      "loss": 0.7769,
      "num_input_tokens_seen": 35946280,
      "step": 61920
    },
    {
      "epoch": 9.223264819779565,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.019577662046336812,
      "loss": 0.8345,
      "num_input_tokens_seen": 35949096,
      "step": 61925
    },
    {
      "epoch": 9.224009532320524,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.019575805362024492,
      "loss": 0.7758,
      "num_input_tokens_seen": 35952040,
      "step": 61930
    },
    {
      "epoch": 9.224754244861483,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.019573948600408595,
      "loss": 0.786,
      "num_input_tokens_seen": 35954760,
      "step": 61935
    },
    {
      "epoch": 9.225498957402444,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.019572091761520492,
      "loss": 0.7888,
      "num_input_tokens_seen": 35957864,
      "step": 61940
    },
    {
      "epoch": 9.226243669943402,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.019570234845391537,
      "loss": 0.8095,
      "num_input_tokens_seen": 35960712,
      "step": 61945
    },
    {
      "epoch": 9.226988382484361,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.019568377852053117,
      "loss": 0.8003,
      "num_input_tokens_seen": 35963656,
      "step": 61950
    },
    {
      "epoch": 9.22773309502532,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.019566520781536586,
      "loss": 0.784,
      "num_input_tokens_seen": 35966472,
      "step": 61955
    },
    {
      "epoch": 9.228477807566279,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.019564663633873332,
      "loss": 0.797,
      "num_input_tokens_seen": 35969384,
      "step": 61960
    },
    {
      "epoch": 9.229222520107239,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.019562806409094725,
      "loss": 0.8061,
      "num_input_tokens_seen": 35972520,
      "step": 61965
    },
    {
      "epoch": 9.229967232648198,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.019560949107232144,
      "loss": 0.8254,
      "num_input_tokens_seen": 35975880,
      "step": 61970
    },
    {
      "epoch": 9.230711945189157,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01955909172831696,
      "loss": 0.7936,
      "num_input_tokens_seen": 35978632,
      "step": 61975
    },
    {
      "epoch": 9.231456657730115,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.01955723427238056,
      "loss": 0.8052,
      "num_input_tokens_seen": 35981416,
      "step": 61980
    },
    {
      "epoch": 9.232201370271076,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.019555376739454313,
      "loss": 0.7726,
      "num_input_tokens_seen": 35984328,
      "step": 61985
    },
    {
      "epoch": 9.232946082812035,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.01955351912956961,
      "loss": 0.8194,
      "num_input_tokens_seen": 35987016,
      "step": 61990
    },
    {
      "epoch": 9.233690795352993,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.019551661442757825,
      "loss": 0.8286,
      "num_input_tokens_seen": 35989800,
      "step": 61995
    },
    {
      "epoch": 9.234435507893952,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.019549803679050348,
      "loss": 0.8034,
      "num_input_tokens_seen": 35992712,
      "step": 62000
    },
    {
      "epoch": 9.235180220434913,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.01954794583847856,
      "loss": 0.7827,
      "num_input_tokens_seen": 35995816,
      "step": 62005
    },
    {
      "epoch": 9.235924932975871,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.019546087921073853,
      "loss": 0.8069,
      "num_input_tokens_seen": 35998600,
      "step": 62010
    },
    {
      "epoch": 9.23666964551683,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.01954422992686761,
      "loss": 0.838,
      "num_input_tokens_seen": 36001352,
      "step": 62015
    },
    {
      "epoch": 9.237414358057789,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.019542371855891225,
      "loss": 0.812,
      "num_input_tokens_seen": 36004456,
      "step": 62020
    },
    {
      "epoch": 9.23815907059875,
      "grad_norm": 0.021484375,
      "learning_rate": 0.019540513708176076,
      "loss": 0.799,
      "num_input_tokens_seen": 36007240,
      "step": 62025
    },
    {
      "epoch": 9.238903783139708,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.019538655483753567,
      "loss": 0.788,
      "num_input_tokens_seen": 36009992,
      "step": 62030
    },
    {
      "epoch": 9.239648495680667,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.019536797182655084,
      "loss": 0.8282,
      "num_input_tokens_seen": 36013000,
      "step": 62035
    },
    {
      "epoch": 9.240393208221626,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.019534938804912024,
      "loss": 0.831,
      "num_input_tokens_seen": 36016200,
      "step": 62040
    },
    {
      "epoch": 9.241137920762586,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.019533080350555787,
      "loss": 0.8077,
      "num_input_tokens_seen": 36018920,
      "step": 62045
    },
    {
      "epoch": 9.241882633303545,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.01953122181961776,
      "loss": 0.8293,
      "num_input_tokens_seen": 36021576,
      "step": 62050
    },
    {
      "epoch": 9.242627345844504,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.019529363212129346,
      "loss": 0.8149,
      "num_input_tokens_seen": 36024776,
      "step": 62055
    },
    {
      "epoch": 9.243372058385463,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.019527504528121947,
      "loss": 0.799,
      "num_input_tokens_seen": 36027432,
      "step": 62060
    },
    {
      "epoch": 9.244116770926423,
      "grad_norm": 0.0234375,
      "learning_rate": 0.01952564576762696,
      "loss": 0.8013,
      "num_input_tokens_seen": 36030344,
      "step": 62065
    },
    {
      "epoch": 9.244861483467382,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.019523786930675784,
      "loss": 0.8067,
      "num_input_tokens_seen": 36033672,
      "step": 62070
    },
    {
      "epoch": 9.24560619600834,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.019521928017299826,
      "loss": 0.7882,
      "num_input_tokens_seen": 36036616,
      "step": 62075
    },
    {
      "epoch": 9.2463509085493,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.019520069027530493,
      "loss": 0.7959,
      "num_input_tokens_seen": 36039368,
      "step": 62080
    },
    {
      "epoch": 9.24709562109026,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.019518209961399188,
      "loss": 0.8021,
      "num_input_tokens_seen": 36042152,
      "step": 62085
    },
    {
      "epoch": 9.247840333631219,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.019516350818937316,
      "loss": 0.7798,
      "num_input_tokens_seen": 36045096,
      "step": 62090
    },
    {
      "epoch": 9.248585046172177,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.019514491600176286,
      "loss": 0.8054,
      "num_input_tokens_seen": 36047976,
      "step": 62095
    },
    {
      "epoch": 9.249329758713136,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.01951263230514751,
      "loss": 0.7784,
      "num_input_tokens_seen": 36050888,
      "step": 62100
    },
    {
      "epoch": 9.250074471254097,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.019510772933882398,
      "loss": 0.7928,
      "num_input_tokens_seen": 36053768,
      "step": 62105
    },
    {
      "epoch": 9.250819183795056,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.019508913486412358,
      "loss": 0.7981,
      "num_input_tokens_seen": 36056680,
      "step": 62110
    },
    {
      "epoch": 9.251563896336014,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.019507053962768816,
      "loss": 0.7996,
      "num_input_tokens_seen": 36059560,
      "step": 62115
    },
    {
      "epoch": 9.252308608876973,
      "grad_norm": 0.017578125,
      "learning_rate": 0.01950519436298317,
      "loss": 0.8184,
      "num_input_tokens_seen": 36062312,
      "step": 62120
    },
    {
      "epoch": 9.253053321417934,
      "grad_norm": 0.01446533203125,
      "learning_rate": 0.019503334687086846,
      "loss": 0.8074,
      "num_input_tokens_seen": 36065128,
      "step": 62125
    },
    {
      "epoch": 9.253798033958892,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.01950147493511126,
      "loss": 0.8195,
      "num_input_tokens_seen": 36067912,
      "step": 62130
    },
    {
      "epoch": 9.254542746499851,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.01949961510708783,
      "loss": 0.7991,
      "num_input_tokens_seen": 36070632,
      "step": 62135
    },
    {
      "epoch": 9.25528745904081,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.019497755203047976,
      "loss": 0.8108,
      "num_input_tokens_seen": 36073512,
      "step": 62140
    },
    {
      "epoch": 9.256032171581769,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.019495895223023123,
      "loss": 0.7858,
      "num_input_tokens_seen": 36076328,
      "step": 62145
    },
    {
      "epoch": 9.25677688412273,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.019494035167044687,
      "loss": 0.8027,
      "num_input_tokens_seen": 36079432,
      "step": 62150
    },
    {
      "epoch": 9.257521596663688,
      "grad_norm": 0.021484375,
      "learning_rate": 0.019492175035144093,
      "loss": 0.8345,
      "num_input_tokens_seen": 36082312,
      "step": 62155
    },
    {
      "epoch": 9.258266309204647,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.019490314827352764,
      "loss": 0.8142,
      "num_input_tokens_seen": 36084904,
      "step": 62160
    },
    {
      "epoch": 9.259011021745605,
      "grad_norm": 0.0234375,
      "learning_rate": 0.019488454543702134,
      "loss": 0.8198,
      "num_input_tokens_seen": 36087784,
      "step": 62165
    },
    {
      "epoch": 9.259755734286566,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.019486594184223625,
      "loss": 0.7971,
      "num_input_tokens_seen": 36090600,
      "step": 62170
    },
    {
      "epoch": 9.260500446827525,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01948473374894867,
      "loss": 0.8176,
      "num_input_tokens_seen": 36093448,
      "step": 62175
    },
    {
      "epoch": 9.261245159368483,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.019482873237908694,
      "loss": 0.811,
      "num_input_tokens_seen": 36096392,
      "step": 62180
    },
    {
      "epoch": 9.261989871909442,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.019481012651135134,
      "loss": 0.8103,
      "num_input_tokens_seen": 36099368,
      "step": 62185
    },
    {
      "epoch": 9.262734584450403,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.01947915198865942,
      "loss": 0.8244,
      "num_input_tokens_seen": 36102376,
      "step": 62190
    },
    {
      "epoch": 9.263479296991362,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.01947729125051298,
      "loss": 0.7913,
      "num_input_tokens_seen": 36105128,
      "step": 62195
    },
    {
      "epoch": 9.26422400953232,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.019475430436727253,
      "loss": 0.8008,
      "num_input_tokens_seen": 36108296,
      "step": 62200
    },
    {
      "epoch": 9.264968722073279,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.019473569547333685,
      "loss": 0.803,
      "num_input_tokens_seen": 36111656,
      "step": 62205
    },
    {
      "epoch": 9.26571343461424,
      "grad_norm": 0.02734375,
      "learning_rate": 0.019471708582363703,
      "loss": 0.7973,
      "num_input_tokens_seen": 36114536,
      "step": 62210
    },
    {
      "epoch": 9.266458147155198,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.01946984754184875,
      "loss": 0.8109,
      "num_input_tokens_seen": 36117672,
      "step": 62215
    },
    {
      "epoch": 9.267202859696157,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.019467986425820268,
      "loss": 0.7894,
      "num_input_tokens_seen": 36120360,
      "step": 62220
    },
    {
      "epoch": 9.267947572237116,
      "grad_norm": 0.02734375,
      "learning_rate": 0.019466125234309694,
      "loss": 0.8172,
      "num_input_tokens_seen": 36123208,
      "step": 62225
    },
    {
      "epoch": 9.268692284778076,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.01946426396734847,
      "loss": 0.7859,
      "num_input_tokens_seen": 36125992,
      "step": 62230
    },
    {
      "epoch": 9.269436997319035,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.019462402624968046,
      "loss": 0.7869,
      "num_input_tokens_seen": 36128936,
      "step": 62235
    },
    {
      "epoch": 9.270181709859994,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.019460541207199863,
      "loss": 0.8035,
      "num_input_tokens_seen": 36132072,
      "step": 62240
    },
    {
      "epoch": 9.270926422400953,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.019458679714075375,
      "loss": 0.799,
      "num_input_tokens_seen": 36135304,
      "step": 62245
    },
    {
      "epoch": 9.271671134941913,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.019456818145626022,
      "loss": 0.8246,
      "num_input_tokens_seen": 36138312,
      "step": 62250
    },
    {
      "epoch": 9.272415847482872,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.019454956501883256,
      "loss": 0.7887,
      "num_input_tokens_seen": 36141192,
      "step": 62255
    },
    {
      "epoch": 9.27316056002383,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.019453094782878527,
      "loss": 0.8093,
      "num_input_tokens_seen": 36144232,
      "step": 62260
    },
    {
      "epoch": 9.27390527256479,
      "grad_norm": 0.01953125,
      "learning_rate": 0.01945123298864329,
      "loss": 0.8103,
      "num_input_tokens_seen": 36147016,
      "step": 62265
    },
    {
      "epoch": 9.27464998510575,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.019449371119208993,
      "loss": 0.7932,
      "num_input_tokens_seen": 36149832,
      "step": 62270
    },
    {
      "epoch": 9.275394697646709,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.019447509174607094,
      "loss": 0.807,
      "num_input_tokens_seen": 36153000,
      "step": 62275
    },
    {
      "epoch": 9.276139410187668,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.019445647154869048,
      "loss": 0.8005,
      "num_input_tokens_seen": 36156040,
      "step": 62280
    },
    {
      "epoch": 9.276884122728626,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.019443785060026312,
      "loss": 0.7872,
      "num_input_tokens_seen": 36158568,
      "step": 62285
    },
    {
      "epoch": 9.277628835269585,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.019441922890110343,
      "loss": 0.7889,
      "num_input_tokens_seen": 36161512,
      "step": 62290
    },
    {
      "epoch": 9.278373547810546,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.019440060645152604,
      "loss": 0.8124,
      "num_input_tokens_seen": 36164328,
      "step": 62295
    },
    {
      "epoch": 9.279118260351504,
      "grad_norm": 0.025390625,
      "learning_rate": 0.019438198325184553,
      "loss": 0.8014,
      "num_input_tokens_seen": 36167272,
      "step": 62300
    },
    {
      "epoch": 9.279862972892463,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.019436335930237646,
      "loss": 0.7899,
      "num_input_tokens_seen": 36170120,
      "step": 62305
    },
    {
      "epoch": 9.280607685433422,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.019434473460343354,
      "loss": 0.7821,
      "num_input_tokens_seen": 36173160,
      "step": 62310
    },
    {
      "epoch": 9.281352397974382,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.019432610915533145,
      "loss": 0.8096,
      "num_input_tokens_seen": 36175816,
      "step": 62315
    },
    {
      "epoch": 9.282097110515341,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.019430748295838476,
      "loss": 0.8089,
      "num_input_tokens_seen": 36179304,
      "step": 62320
    },
    {
      "epoch": 9.2828418230563,
      "grad_norm": 0.015625,
      "learning_rate": 0.01942888560129082,
      "loss": 0.8089,
      "num_input_tokens_seen": 36182024,
      "step": 62325
    },
    {
      "epoch": 9.283586535597259,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.01942702283192164,
      "loss": 0.8057,
      "num_input_tokens_seen": 36184904,
      "step": 62330
    },
    {
      "epoch": 9.28433124813822,
      "grad_norm": 0.025390625,
      "learning_rate": 0.01942515998776241,
      "loss": 0.8041,
      "num_input_tokens_seen": 36188040,
      "step": 62335
    },
    {
      "epoch": 9.285075960679178,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.019423297068844603,
      "loss": 0.7907,
      "num_input_tokens_seen": 36191048,
      "step": 62340
    },
    {
      "epoch": 9.285820673220137,
      "grad_norm": 0.0166015625,
      "learning_rate": 0.019421434075199685,
      "loss": 0.7895,
      "num_input_tokens_seen": 36193768,
      "step": 62345
    },
    {
      "epoch": 9.286565385761095,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.019419571006859135,
      "loss": 0.78,
      "num_input_tokens_seen": 36196584,
      "step": 62350
    },
    {
      "epoch": 9.287310098302056,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.019417707863854418,
      "loss": 0.8052,
      "num_input_tokens_seen": 36199432,
      "step": 62355
    },
    {
      "epoch": 9.288054810843015,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.019415844646217022,
      "loss": 0.7966,
      "num_input_tokens_seen": 36202280,
      "step": 62360
    },
    {
      "epoch": 9.288799523383974,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.019413981353978413,
      "loss": 0.8148,
      "num_input_tokens_seen": 36205064,
      "step": 62365
    },
    {
      "epoch": 9.289544235924932,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01941211798717008,
      "loss": 0.8158,
      "num_input_tokens_seen": 36207880,
      "step": 62370
    },
    {
      "epoch": 9.290288948465893,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.0194102545458235,
      "loss": 0.8013,
      "num_input_tokens_seen": 36210664,
      "step": 62375
    },
    {
      "epoch": 9.291033661006852,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.019408391029970152,
      "loss": 0.7741,
      "num_input_tokens_seen": 36213192,
      "step": 62380
    },
    {
      "epoch": 9.29177837354781,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.01940652743964151,
      "loss": 0.8126,
      "num_input_tokens_seen": 36215976,
      "step": 62385
    },
    {
      "epoch": 9.292523086088769,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.01940466377486907,
      "loss": 0.7956,
      "num_input_tokens_seen": 36218888,
      "step": 62390
    },
    {
      "epoch": 9.29326779862973,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.01940280003568431,
      "loss": 0.8245,
      "num_input_tokens_seen": 36221640,
      "step": 62395
    },
    {
      "epoch": 9.294012511170688,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.01940093622211872,
      "loss": 0.7985,
      "num_input_tokens_seen": 36224360,
      "step": 62400
    },
    {
      "epoch": 9.294757223711647,
      "grad_norm": 0.021484375,
      "learning_rate": 0.01939907233420379,
      "loss": 0.7952,
      "num_input_tokens_seen": 36227432,
      "step": 62405
    },
    {
      "epoch": 9.295501936252606,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.019397208371970998,
      "loss": 0.798,
      "num_input_tokens_seen": 36230344,
      "step": 62410
    },
    {
      "epoch": 9.296246648793566,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.019395344335451842,
      "loss": 0.7968,
      "num_input_tokens_seen": 36233160,
      "step": 62415
    },
    {
      "epoch": 9.296991361334525,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.019393480224677812,
      "loss": 0.815,
      "num_input_tokens_seen": 36236008,
      "step": 62420
    },
    {
      "epoch": 9.297736073875484,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.019391616039680395,
      "loss": 0.8005,
      "num_input_tokens_seen": 36239208,
      "step": 62425
    },
    {
      "epoch": 9.298480786416443,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.019389751780491092,
      "loss": 0.7861,
      "num_input_tokens_seen": 36242216,
      "step": 62430
    },
    {
      "epoch": 9.299225498957403,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.019387887447141393,
      "loss": 0.7867,
      "num_input_tokens_seen": 36245320,
      "step": 62435
    },
    {
      "epoch": 9.299970211498362,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.019386023039662793,
      "loss": 0.7974,
      "num_input_tokens_seen": 36248104,
      "step": 62440
    },
    {
      "epoch": 9.30071492403932,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.019384158558086798,
      "loss": 0.7738,
      "num_input_tokens_seen": 36250696,
      "step": 62445
    },
    {
      "epoch": 9.30145963658028,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.019382294002444893,
      "loss": 0.8007,
      "num_input_tokens_seen": 36253448,
      "step": 62450
    },
    {
      "epoch": 9.30220434912124,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.019380429372768588,
      "loss": 0.8036,
      "num_input_tokens_seen": 36256072,
      "step": 62455
    },
    {
      "epoch": 9.302949061662199,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.019378564669089382,
      "loss": 0.8024,
      "num_input_tokens_seen": 36258952,
      "step": 62460
    },
    {
      "epoch": 9.303693774203158,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.019376699891438774,
      "loss": 0.7807,
      "num_input_tokens_seen": 36261832,
      "step": 62465
    },
    {
      "epoch": 9.304438486744116,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.019374835039848267,
      "loss": 0.8099,
      "num_input_tokens_seen": 36264680,
      "step": 62470
    },
    {
      "epoch": 9.305183199285075,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.019372970114349373,
      "loss": 0.7778,
      "num_input_tokens_seen": 36267656,
      "step": 62475
    },
    {
      "epoch": 9.305927911826036,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.019371105114973593,
      "loss": 0.8309,
      "num_input_tokens_seen": 36270408,
      "step": 62480
    },
    {
      "epoch": 9.306672624366994,
      "grad_norm": 0.0234375,
      "learning_rate": 0.019369240041752434,
      "loss": 0.7924,
      "num_input_tokens_seen": 36273224,
      "step": 62485
    },
    {
      "epoch": 9.307417336907953,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0193673748947174,
      "loss": 0.8065,
      "num_input_tokens_seen": 36276232,
      "step": 62490
    },
    {
      "epoch": 9.308162049448912,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.01936550967390001,
      "loss": 0.8126,
      "num_input_tokens_seen": 36279272,
      "step": 62495
    },
    {
      "epoch": 9.308906761989872,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.019363644379331777,
      "loss": 0.7836,
      "num_input_tokens_seen": 36281928,
      "step": 62500
    },
    {
      "epoch": 9.309651474530831,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.0193617790110442,
      "loss": 0.8314,
      "num_input_tokens_seen": 36284744,
      "step": 62505
    },
    {
      "epoch": 9.31039618707179,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.0193599135690688,
      "loss": 0.8043,
      "num_input_tokens_seen": 36287752,
      "step": 62510
    },
    {
      "epoch": 9.311140899612749,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.019358048053437094,
      "loss": 0.8001,
      "num_input_tokens_seen": 36290504,
      "step": 62515
    },
    {
      "epoch": 9.31188561215371,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.019356182464180596,
      "loss": 0.7777,
      "num_input_tokens_seen": 36293384,
      "step": 62520
    },
    {
      "epoch": 9.312630324694668,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.01935431680133082,
      "loss": 0.8026,
      "num_input_tokens_seen": 36296328,
      "step": 62525
    },
    {
      "epoch": 9.313375037235627,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01935245106491929,
      "loss": 0.8011,
      "num_input_tokens_seen": 36299240,
      "step": 62530
    },
    {
      "epoch": 9.314119749776586,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.019350585254977522,
      "loss": 0.7947,
      "num_input_tokens_seen": 36302024,
      "step": 62535
    },
    {
      "epoch": 9.314864462317546,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.019348719371537034,
      "loss": 0.8181,
      "num_input_tokens_seen": 36305288,
      "step": 62540
    },
    {
      "epoch": 9.315609174858505,
      "grad_norm": 0.0152587890625,
      "learning_rate": 0.01934685341462936,
      "loss": 0.8103,
      "num_input_tokens_seen": 36308008,
      "step": 62545
    },
    {
      "epoch": 9.316353887399464,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.01934498738428601,
      "loss": 0.7947,
      "num_input_tokens_seen": 36310984,
      "step": 62550
    },
    {
      "epoch": 9.317098599940422,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.019343121280538514,
      "loss": 0.7806,
      "num_input_tokens_seen": 36313960,
      "step": 62555
    },
    {
      "epoch": 9.317843312481383,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.019341255103418397,
      "loss": 0.8173,
      "num_input_tokens_seen": 36316744,
      "step": 62560
    },
    {
      "epoch": 9.318588025022342,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.019339388852957187,
      "loss": 0.7888,
      "num_input_tokens_seen": 36319496,
      "step": 62565
    },
    {
      "epoch": 9.3193327375633,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.019337522529186413,
      "loss": 0.8063,
      "num_input_tokens_seen": 36322248,
      "step": 62570
    },
    {
      "epoch": 9.32007745010426,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.01933565613213761,
      "loss": 0.8324,
      "num_input_tokens_seen": 36325224,
      "step": 62575
    },
    {
      "epoch": 9.32082216264522,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.0193337896618423,
      "loss": 0.804,
      "num_input_tokens_seen": 36327976,
      "step": 62580
    },
    {
      "epoch": 9.321566875186178,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01933192311833202,
      "loss": 0.8075,
      "num_input_tokens_seen": 36330792,
      "step": 62585
    },
    {
      "epoch": 9.322311587727137,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.0193300565016383,
      "loss": 0.7997,
      "num_input_tokens_seen": 36333576,
      "step": 62590
    },
    {
      "epoch": 9.323056300268096,
      "grad_norm": 0.015625,
      "learning_rate": 0.019328189811792674,
      "loss": 0.8262,
      "num_input_tokens_seen": 36336424,
      "step": 62595
    },
    {
      "epoch": 9.323801012809056,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.01932632304882668,
      "loss": 0.8189,
      "num_input_tokens_seen": 36339304,
      "step": 62600
    },
    {
      "epoch": 9.324545725350015,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.019324456212771856,
      "loss": 0.821,
      "num_input_tokens_seen": 36341992,
      "step": 62605
    },
    {
      "epoch": 9.325290437890974,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.019322589303659744,
      "loss": 0.7991,
      "num_input_tokens_seen": 36344968,
      "step": 62610
    },
    {
      "epoch": 9.326035150431933,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.019320722321521874,
      "loss": 0.8014,
      "num_input_tokens_seen": 36348136,
      "step": 62615
    },
    {
      "epoch": 9.326779862972893,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.01931885526638979,
      "loss": 0.7946,
      "num_input_tokens_seen": 36350824,
      "step": 62620
    },
    {
      "epoch": 9.327524575513852,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.019316988138295046,
      "loss": 0.7801,
      "num_input_tokens_seen": 36353800,
      "step": 62625
    },
    {
      "epoch": 9.32826928805481,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.019315120937269167,
      "loss": 0.8206,
      "num_input_tokens_seen": 36356584,
      "step": 62630
    },
    {
      "epoch": 9.32901400059577,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.019313253663343705,
      "loss": 0.8112,
      "num_input_tokens_seen": 36359368,
      "step": 62635
    },
    {
      "epoch": 9.32975871313673,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.019311386316550206,
      "loss": 0.7849,
      "num_input_tokens_seen": 36362248,
      "step": 62640
    },
    {
      "epoch": 9.330503425677689,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.019309518896920223,
      "loss": 0.8051,
      "num_input_tokens_seen": 36364936,
      "step": 62645
    },
    {
      "epoch": 9.331248138218648,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.01930765140448529,
      "loss": 0.8106,
      "num_input_tokens_seen": 36368328,
      "step": 62650
    },
    {
      "epoch": 9.331992850759606,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.01930578383927697,
      "loss": 0.804,
      "num_input_tokens_seen": 36371144,
      "step": 62655
    },
    {
      "epoch": 9.332737563300565,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.019303916201326803,
      "loss": 0.805,
      "num_input_tokens_seen": 36373896,
      "step": 62660
    },
    {
      "epoch": 9.333482275841526,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.019302048490666354,
      "loss": 0.7896,
      "num_input_tokens_seen": 36376840,
      "step": 62665
    },
    {
      "epoch": 9.334226988382484,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.019300180707327166,
      "loss": 0.8133,
      "num_input_tokens_seen": 36379656,
      "step": 62670
    },
    {
      "epoch": 9.334971700923443,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.019298312851340788,
      "loss": 0.7925,
      "num_input_tokens_seen": 36382472,
      "step": 62675
    },
    {
      "epoch": 9.335716413464402,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.019296444922738794,
      "loss": 0.7932,
      "num_input_tokens_seen": 36385448,
      "step": 62680
    },
    {
      "epoch": 9.336461126005362,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.019294576921552724,
      "loss": 0.8085,
      "num_input_tokens_seen": 36388520,
      "step": 62685
    },
    {
      "epoch": 9.337205838546321,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.01929270884781414,
      "loss": 0.8076,
      "num_input_tokens_seen": 36391272,
      "step": 62690
    },
    {
      "epoch": 9.33795055108728,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.019290840701554606,
      "loss": 0.7808,
      "num_input_tokens_seen": 36394152,
      "step": 62695
    },
    {
      "epoch": 9.338695263628239,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.019288972482805675,
      "loss": 0.8184,
      "num_input_tokens_seen": 36397192,
      "step": 62700
    },
    {
      "epoch": 9.3394399761692,
      "grad_norm": 0.025390625,
      "learning_rate": 0.01928710419159892,
      "loss": 0.7837,
      "num_input_tokens_seen": 36400168,
      "step": 62705
    },
    {
      "epoch": 9.340184688710158,
      "grad_norm": 0.02734375,
      "learning_rate": 0.019285235827965895,
      "loss": 0.7927,
      "num_input_tokens_seen": 36402920,
      "step": 62710
    },
    {
      "epoch": 9.340929401251117,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.01928336739193816,
      "loss": 0.7917,
      "num_input_tokens_seen": 36406088,
      "step": 62715
    },
    {
      "epoch": 9.341674113792076,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.019281498883547295,
      "loss": 0.7831,
      "num_input_tokens_seen": 36409256,
      "step": 62720
    },
    {
      "epoch": 9.342418826333036,
      "grad_norm": 0.0234375,
      "learning_rate": 0.01927963030282485,
      "loss": 0.8001,
      "num_input_tokens_seen": 36411880,
      "step": 62725
    },
    {
      "epoch": 9.343163538873995,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.019277761649802404,
      "loss": 0.8091,
      "num_input_tokens_seen": 36414760,
      "step": 62730
    },
    {
      "epoch": 9.343908251414954,
      "grad_norm": 0.0172119140625,
      "learning_rate": 0.019275892924511522,
      "loss": 0.7794,
      "num_input_tokens_seen": 36417608,
      "step": 62735
    },
    {
      "epoch": 9.344652963955912,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.019274024126983775,
      "loss": 0.8217,
      "num_input_tokens_seen": 36420648,
      "step": 62740
    },
    {
      "epoch": 9.345397676496873,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.019272155257250737,
      "loss": 0.8135,
      "num_input_tokens_seen": 36423432,
      "step": 62745
    },
    {
      "epoch": 9.346142389037832,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.019270286315343974,
      "loss": 0.8241,
      "num_input_tokens_seen": 36426312,
      "step": 62750
    },
    {
      "epoch": 9.34688710157879,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.019268417301295068,
      "loss": 0.7762,
      "num_input_tokens_seen": 36429256,
      "step": 62755
    },
    {
      "epoch": 9.34763181411975,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.019266548215135588,
      "loss": 0.7964,
      "num_input_tokens_seen": 36432360,
      "step": 62760
    },
    {
      "epoch": 9.34837652666071,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.01926467905689711,
      "loss": 0.7887,
      "num_input_tokens_seen": 36435400,
      "step": 62765
    },
    {
      "epoch": 9.349121239201668,
      "grad_norm": 0.033203125,
      "learning_rate": 0.019262809826611212,
      "loss": 0.7966,
      "num_input_tokens_seen": 36438312,
      "step": 62770
    },
    {
      "epoch": 9.349865951742627,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.019260940524309485,
      "loss": 0.783,
      "num_input_tokens_seen": 36441256,
      "step": 62775
    },
    {
      "epoch": 9.350610664283586,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.01925907115002349,
      "loss": 0.8173,
      "num_input_tokens_seen": 36444168,
      "step": 62780
    },
    {
      "epoch": 9.351355376824547,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.019257201703784817,
      "loss": 0.7706,
      "num_input_tokens_seen": 36446760,
      "step": 62785
    },
    {
      "epoch": 9.352100089365505,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.01925533218562505,
      "loss": 0.8087,
      "num_input_tokens_seen": 36449512,
      "step": 62790
    },
    {
      "epoch": 9.352844801906464,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.019253462595575768,
      "loss": 0.7933,
      "num_input_tokens_seen": 36452648,
      "step": 62795
    },
    {
      "epoch": 9.353589514447423,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01925159293366856,
      "loss": 0.7851,
      "num_input_tokens_seen": 36455432,
      "step": 62800
    },
    {
      "epoch": 9.354334226988382,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.019249723199935013,
      "loss": 0.7953,
      "num_input_tokens_seen": 36458600,
      "step": 62805
    },
    {
      "epoch": 9.355078939529342,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.019247853394406707,
      "loss": 0.8014,
      "num_input_tokens_seen": 36461544,
      "step": 62810
    },
    {
      "epoch": 9.3558236520703,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.01924598351711524,
      "loss": 0.7819,
      "num_input_tokens_seen": 36464392,
      "step": 62815
    },
    {
      "epoch": 9.35656836461126,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.019244113568092195,
      "loss": 0.8081,
      "num_input_tokens_seen": 36467080,
      "step": 62820
    },
    {
      "epoch": 9.357313077152218,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.019242243547369163,
      "loss": 0.8091,
      "num_input_tokens_seen": 36469896,
      "step": 62825
    },
    {
      "epoch": 9.358057789693179,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.01924037345497774,
      "loss": 0.7906,
      "num_input_tokens_seen": 36472744,
      "step": 62830
    },
    {
      "epoch": 9.358802502234138,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.019238503290949515,
      "loss": 0.8221,
      "num_input_tokens_seen": 36475368,
      "step": 62835
    },
    {
      "epoch": 9.359547214775096,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.019236633055316088,
      "loss": 0.7916,
      "num_input_tokens_seen": 36478152,
      "step": 62840
    },
    {
      "epoch": 9.360291927316055,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.019234762748109053,
      "loss": 0.7951,
      "num_input_tokens_seen": 36480904,
      "step": 62845
    },
    {
      "epoch": 9.361036639857016,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.01923289236936,
      "loss": 0.7914,
      "num_input_tokens_seen": 36483784,
      "step": 62850
    },
    {
      "epoch": 9.361781352397974,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.019231021919100533,
      "loss": 0.8292,
      "num_input_tokens_seen": 36486792,
      "step": 62855
    },
    {
      "epoch": 9.362526064938933,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.019229151397362254,
      "loss": 0.795,
      "num_input_tokens_seen": 36489832,
      "step": 62860
    },
    {
      "epoch": 9.363270777479892,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.019227280804176763,
      "loss": 0.8055,
      "num_input_tokens_seen": 36492744,
      "step": 62865
    },
    {
      "epoch": 9.364015490020853,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.01922541013957565,
      "loss": 0.7908,
      "num_input_tokens_seen": 36495432,
      "step": 62870
    },
    {
      "epoch": 9.364760202561811,
      "grad_norm": 0.0234375,
      "learning_rate": 0.019223539403590537,
      "loss": 0.8103,
      "num_input_tokens_seen": 36497992,
      "step": 62875
    },
    {
      "epoch": 9.36550491510277,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.019221668596253013,
      "loss": 0.787,
      "num_input_tokens_seen": 36500648,
      "step": 62880
    },
    {
      "epoch": 9.366249627643729,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.01921979771759469,
      "loss": 0.7995,
      "num_input_tokens_seen": 36503784,
      "step": 62885
    },
    {
      "epoch": 9.36699434018469,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.01921792676764717,
      "loss": 0.8071,
      "num_input_tokens_seen": 36506376,
      "step": 62890
    },
    {
      "epoch": 9.367739052725648,
      "grad_norm": 0.0546875,
      "learning_rate": 0.019216055746442067,
      "loss": 0.8026,
      "num_input_tokens_seen": 36509480,
      "step": 62895
    },
    {
      "epoch": 9.368483765266607,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.019214184654010986,
      "loss": 0.8176,
      "num_input_tokens_seen": 36512264,
      "step": 62900
    },
    {
      "epoch": 9.369228477807566,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.019212313490385543,
      "loss": 0.7757,
      "num_input_tokens_seen": 36515080,
      "step": 62905
    },
    {
      "epoch": 9.369973190348526,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.019210442255597347,
      "loss": 0.811,
      "num_input_tokens_seen": 36517896,
      "step": 62910
    },
    {
      "epoch": 9.370717902889485,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.019208570949677997,
      "loss": 0.785,
      "num_input_tokens_seen": 36520552,
      "step": 62915
    },
    {
      "epoch": 9.371462615430444,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.019206699572659126,
      "loss": 0.798,
      "num_input_tokens_seen": 36523464,
      "step": 62920
    },
    {
      "epoch": 9.372207327971402,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01920482812457234,
      "loss": 0.7972,
      "num_input_tokens_seen": 36526216,
      "step": 62925
    },
    {
      "epoch": 9.372952040512363,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.019202956605449253,
      "loss": 0.817,
      "num_input_tokens_seen": 36529032,
      "step": 62930
    },
    {
      "epoch": 9.373696753053322,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01920108501532149,
      "loss": 0.7951,
      "num_input_tokens_seen": 36531592,
      "step": 62935
    },
    {
      "epoch": 9.37444146559428,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.019199213354220663,
      "loss": 0.8098,
      "num_input_tokens_seen": 36534760,
      "step": 62940
    },
    {
      "epoch": 9.37518617813524,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0191973416221784,
      "loss": 0.8209,
      "num_input_tokens_seen": 36537736,
      "step": 62945
    },
    {
      "epoch": 9.3759308906762,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.01919546981922631,
      "loss": 0.8069,
      "num_input_tokens_seen": 36540680,
      "step": 62950
    },
    {
      "epoch": 9.376675603217159,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.01919359794539602,
      "loss": 0.7973,
      "num_input_tokens_seen": 36543144,
      "step": 62955
    },
    {
      "epoch": 9.377420315758117,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.01919172600071916,
      "loss": 0.8053,
      "num_input_tokens_seen": 36546088,
      "step": 62960
    },
    {
      "epoch": 9.378165028299076,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.019189853985227344,
      "loss": 0.7897,
      "num_input_tokens_seen": 36549384,
      "step": 62965
    },
    {
      "epoch": 9.378909740840037,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.019187981898952208,
      "loss": 0.7885,
      "num_input_tokens_seen": 36552424,
      "step": 62970
    },
    {
      "epoch": 9.379654453380995,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.019186109741925375,
      "loss": 0.8078,
      "num_input_tokens_seen": 36555432,
      "step": 62975
    },
    {
      "epoch": 9.380399165921954,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.01918423751417847,
      "loss": 0.8036,
      "num_input_tokens_seen": 36558632,
      "step": 62980
    },
    {
      "epoch": 9.381143878462913,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.019182365215743128,
      "loss": 0.8151,
      "num_input_tokens_seen": 36561768,
      "step": 62985
    },
    {
      "epoch": 9.381888591003872,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.019180492846650972,
      "loss": 0.8049,
      "num_input_tokens_seen": 36564488,
      "step": 62990
    },
    {
      "epoch": 9.382633303544832,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.01917862040693364,
      "loss": 0.8171,
      "num_input_tokens_seen": 36567240,
      "step": 62995
    },
    {
      "epoch": 9.383378016085791,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.019176747896622767,
      "loss": 0.809,
      "num_input_tokens_seen": 36570248,
      "step": 63000
    },
    {
      "epoch": 9.38412272862675,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.019174875315749976,
      "loss": 0.8063,
      "num_input_tokens_seen": 36573000,
      "step": 63005
    },
    {
      "epoch": 9.384867441167708,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.019173002664346916,
      "loss": 0.8122,
      "num_input_tokens_seen": 36576008,
      "step": 63010
    },
    {
      "epoch": 9.385612153708669,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.019171129942445214,
      "loss": 0.817,
      "num_input_tokens_seen": 36578696,
      "step": 63015
    },
    {
      "epoch": 9.386356866249628,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.01916925715007651,
      "loss": 0.7889,
      "num_input_tokens_seen": 36581448,
      "step": 63020
    },
    {
      "epoch": 9.387101578790586,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.019167384287272448,
      "loss": 0.8033,
      "num_input_tokens_seen": 36584456,
      "step": 63025
    },
    {
      "epoch": 9.387846291331545,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.01916551135406466,
      "loss": 0.7882,
      "num_input_tokens_seen": 36587528,
      "step": 63030
    },
    {
      "epoch": 9.388591003872506,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.01916363835048479,
      "loss": 0.7949,
      "num_input_tokens_seen": 36590408,
      "step": 63035
    },
    {
      "epoch": 9.389335716413465,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.019161765276564487,
      "loss": 0.8034,
      "num_input_tokens_seen": 36593160,
      "step": 63040
    },
    {
      "epoch": 9.390080428954423,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.019159892132335387,
      "loss": 0.8213,
      "num_input_tokens_seen": 36596008,
      "step": 63045
    },
    {
      "epoch": 9.390825141495382,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.019158018917829138,
      "loss": 0.7948,
      "num_input_tokens_seen": 36598696,
      "step": 63050
    },
    {
      "epoch": 9.391569854036343,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.019156145633077385,
      "loss": 0.808,
      "num_input_tokens_seen": 36601640,
      "step": 63055
    },
    {
      "epoch": 9.392314566577301,
      "grad_norm": 0.0234375,
      "learning_rate": 0.019154272278111773,
      "loss": 0.807,
      "num_input_tokens_seen": 36604424,
      "step": 63060
    },
    {
      "epoch": 9.39305927911826,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.019152398852963957,
      "loss": 0.8032,
      "num_input_tokens_seen": 36607272,
      "step": 63065
    },
    {
      "epoch": 9.393803991659219,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.019150525357665583,
      "loss": 0.7987,
      "num_input_tokens_seen": 36610120,
      "step": 63070
    },
    {
      "epoch": 9.39454870420018,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.0191486517922483,
      "loss": 0.7908,
      "num_input_tokens_seen": 36612904,
      "step": 63075
    },
    {
      "epoch": 9.395293416741138,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.01914677815674376,
      "loss": 0.7974,
      "num_input_tokens_seen": 36615752,
      "step": 63080
    },
    {
      "epoch": 9.396038129282097,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.01914490445118362,
      "loss": 0.8066,
      "num_input_tokens_seen": 36618888,
      "step": 63085
    },
    {
      "epoch": 9.396782841823056,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.01914303067559953,
      "loss": 0.802,
      "num_input_tokens_seen": 36621896,
      "step": 63090
    },
    {
      "epoch": 9.397527554364016,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.019141156830023147,
      "loss": 0.8249,
      "num_input_tokens_seen": 36624712,
      "step": 63095
    },
    {
      "epoch": 9.398272266904975,
      "grad_norm": 0.01409912109375,
      "learning_rate": 0.01913928291448613,
      "loss": 0.7892,
      "num_input_tokens_seen": 36627400,
      "step": 63100
    },
    {
      "epoch": 9.399016979445934,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.019137408929020137,
      "loss": 0.8036,
      "num_input_tokens_seen": 36630312,
      "step": 63105
    },
    {
      "epoch": 9.399761691986892,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.019135534873656823,
      "loss": 0.8017,
      "num_input_tokens_seen": 36633192,
      "step": 63110
    },
    {
      "epoch": 9.400506404527853,
      "grad_norm": 0.033203125,
      "learning_rate": 0.019133660748427848,
      "loss": 0.8096,
      "num_input_tokens_seen": 36636008,
      "step": 63115
    },
    {
      "epoch": 9.401251117068812,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.01913178655336488,
      "loss": 0.7897,
      "num_input_tokens_seen": 36638984,
      "step": 63120
    },
    {
      "epoch": 9.40199582960977,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.019129912288499574,
      "loss": 0.7907,
      "num_input_tokens_seen": 36641960,
      "step": 63125
    },
    {
      "epoch": 9.40274054215073,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.0191280379538636,
      "loss": 0.8021,
      "num_input_tokens_seen": 36644776,
      "step": 63130
    },
    {
      "epoch": 9.40348525469169,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.019126163549488616,
      "loss": 0.8053,
      "num_input_tokens_seen": 36647560,
      "step": 63135
    },
    {
      "epoch": 9.404229967232649,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.019124289075406302,
      "loss": 0.7993,
      "num_input_tokens_seen": 36650536,
      "step": 63140
    },
    {
      "epoch": 9.404974679773607,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.019122414531648306,
      "loss": 0.8071,
      "num_input_tokens_seen": 36653288,
      "step": 63145
    },
    {
      "epoch": 9.405719392314566,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.019120539918246315,
      "loss": 0.7905,
      "num_input_tokens_seen": 36656424,
      "step": 63150
    },
    {
      "epoch": 9.406464104855527,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.019118665235231983,
      "loss": 0.7852,
      "num_input_tokens_seen": 36659368,
      "step": 63155
    },
    {
      "epoch": 9.407208817396485,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.01911679048263699,
      "loss": 0.7966,
      "num_input_tokens_seen": 36662408,
      "step": 63160
    },
    {
      "epoch": 9.407953529937444,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.019114915660493004,
      "loss": 0.8062,
      "num_input_tokens_seen": 36665576,
      "step": 63165
    },
    {
      "epoch": 9.408698242478403,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.019113040768831704,
      "loss": 0.7734,
      "num_input_tokens_seen": 36668360,
      "step": 63170
    },
    {
      "epoch": 9.409442955019362,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.01911116580768476,
      "loss": 0.8047,
      "num_input_tokens_seen": 36671400,
      "step": 63175
    },
    {
      "epoch": 9.410187667560322,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.019109290777083843,
      "loss": 0.8116,
      "num_input_tokens_seen": 36674344,
      "step": 63180
    },
    {
      "epoch": 9.410932380101281,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.01910741567706064,
      "loss": 0.8021,
      "num_input_tokens_seen": 36677544,
      "step": 63185
    },
    {
      "epoch": 9.41167709264224,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.019105540507646825,
      "loss": 0.8212,
      "num_input_tokens_seen": 36680744,
      "step": 63190
    },
    {
      "epoch": 9.412421805183198,
      "grad_norm": 0.021484375,
      "learning_rate": 0.01910366526887407,
      "loss": 0.7904,
      "num_input_tokens_seen": 36683624,
      "step": 63195
    },
    {
      "epoch": 9.413166517724159,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.019101789960774065,
      "loss": 0.8394,
      "num_input_tokens_seen": 36686376,
      "step": 63200
    },
    {
      "epoch": 9.413911230265118,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.019099914583378487,
      "loss": 0.822,
      "num_input_tokens_seen": 36689544,
      "step": 63205
    },
    {
      "epoch": 9.414655942806077,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.01909803913671902,
      "loss": 0.8194,
      "num_input_tokens_seen": 36692200,
      "step": 63210
    },
    {
      "epoch": 9.415400655347035,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.019096163620827344,
      "loss": 0.813,
      "num_input_tokens_seen": 36695400,
      "step": 63215
    },
    {
      "epoch": 9.416145367887996,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.01909428803573515,
      "loss": 0.7967,
      "num_input_tokens_seen": 36698216,
      "step": 63220
    },
    {
      "epoch": 9.416890080428955,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.019092412381474116,
      "loss": 0.7931,
      "num_input_tokens_seen": 36701288,
      "step": 63225
    },
    {
      "epoch": 9.417634792969913,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01909053665807594,
      "loss": 0.8013,
      "num_input_tokens_seen": 36704072,
      "step": 63230
    },
    {
      "epoch": 9.418379505510872,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.0190886608655723,
      "loss": 0.7851,
      "num_input_tokens_seen": 36707080,
      "step": 63235
    },
    {
      "epoch": 9.419124218051833,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.019086785003994897,
      "loss": 0.8048,
      "num_input_tokens_seen": 36709800,
      "step": 63240
    },
    {
      "epoch": 9.419868930592791,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.019084909073375407,
      "loss": 0.7994,
      "num_input_tokens_seen": 36712328,
      "step": 63245
    },
    {
      "epoch": 9.42061364313375,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.019083033073745533,
      "loss": 0.8101,
      "num_input_tokens_seen": 36715560,
      "step": 63250
    },
    {
      "epoch": 9.421358355674709,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.019081157005136965,
      "loss": 0.8114,
      "num_input_tokens_seen": 36718248,
      "step": 63255
    },
    {
      "epoch": 9.42210306821567,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.019079280867581396,
      "loss": 0.8042,
      "num_input_tokens_seen": 36720936,
      "step": 63260
    },
    {
      "epoch": 9.422847780756628,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.019077404661110523,
      "loss": 0.8081,
      "num_input_tokens_seen": 36723688,
      "step": 63265
    },
    {
      "epoch": 9.423592493297587,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.019075528385756046,
      "loss": 0.7806,
      "num_input_tokens_seen": 36726664,
      "step": 63270
    },
    {
      "epoch": 9.424337205838546,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.019073652041549656,
      "loss": 0.7995,
      "num_input_tokens_seen": 36729448,
      "step": 63275
    },
    {
      "epoch": 9.425081918379506,
      "grad_norm": 0.0159912109375,
      "learning_rate": 0.019071775628523056,
      "loss": 0.7892,
      "num_input_tokens_seen": 36732104,
      "step": 63280
    },
    {
      "epoch": 9.425826630920465,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.019069899146707946,
      "loss": 0.7985,
      "num_input_tokens_seen": 36735080,
      "step": 63285
    },
    {
      "epoch": 9.426571343461424,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.019068022596136023,
      "loss": 0.8101,
      "num_input_tokens_seen": 36738024,
      "step": 63290
    },
    {
      "epoch": 9.427316056002383,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.019066145976838995,
      "loss": 0.7909,
      "num_input_tokens_seen": 36740904,
      "step": 63295
    },
    {
      "epoch": 9.428060768543343,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.01906426928884856,
      "loss": 0.8019,
      "num_input_tokens_seen": 36744040,
      "step": 63300
    },
    {
      "epoch": 9.428805481084302,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.019062392532196434,
      "loss": 0.7893,
      "num_input_tokens_seen": 36746888,
      "step": 63305
    },
    {
      "epoch": 9.42955019362526,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.019060515706914312,
      "loss": 0.8133,
      "num_input_tokens_seen": 36749736,
      "step": 63310
    },
    {
      "epoch": 9.43029490616622,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.019058638813033903,
      "loss": 0.7901,
      "num_input_tokens_seen": 36752584,
      "step": 63315
    },
    {
      "epoch": 9.43103961870718,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.019056761850586915,
      "loss": 0.7936,
      "num_input_tokens_seen": 36755784,
      "step": 63320
    },
    {
      "epoch": 9.431784331248139,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.01905488481960506,
      "loss": 0.8057,
      "num_input_tokens_seen": 36758600,
      "step": 63325
    },
    {
      "epoch": 9.432529043789097,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.019053007720120044,
      "loss": 0.807,
      "num_input_tokens_seen": 36761672,
      "step": 63330
    },
    {
      "epoch": 9.433273756330056,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.01905113055216358,
      "loss": 0.8022,
      "num_input_tokens_seen": 36764680,
      "step": 63335
    },
    {
      "epoch": 9.434018468871017,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.019049253315767394,
      "loss": 0.7789,
      "num_input_tokens_seen": 36767304,
      "step": 63340
    },
    {
      "epoch": 9.434763181411975,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01904737601096318,
      "loss": 0.8077,
      "num_input_tokens_seen": 36770216,
      "step": 63345
    },
    {
      "epoch": 9.435507893952934,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.019045498637782664,
      "loss": 0.8004,
      "num_input_tokens_seen": 36773096,
      "step": 63350
    },
    {
      "epoch": 9.436252606493893,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01904362119625756,
      "loss": 0.8048,
      "num_input_tokens_seen": 36775976,
      "step": 63355
    },
    {
      "epoch": 9.436997319034852,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.019041743686419584,
      "loss": 0.8038,
      "num_input_tokens_seen": 36779016,
      "step": 63360
    },
    {
      "epoch": 9.437742031575812,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.019039866108300458,
      "loss": 0.8182,
      "num_input_tokens_seen": 36781768,
      "step": 63365
    },
    {
      "epoch": 9.438486744116771,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.0190379884619319,
      "loss": 0.8152,
      "num_input_tokens_seen": 36784840,
      "step": 63370
    },
    {
      "epoch": 9.43923145665773,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.019036110747345632,
      "loss": 0.795,
      "num_input_tokens_seen": 36787816,
      "step": 63375
    },
    {
      "epoch": 9.439976169198689,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.019034232964573374,
      "loss": 0.8097,
      "num_input_tokens_seen": 36790696,
      "step": 63380
    },
    {
      "epoch": 9.440720881739649,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.019032355113646846,
      "loss": 0.8076,
      "num_input_tokens_seen": 36793768,
      "step": 63385
    },
    {
      "epoch": 9.441465594280608,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.019030477194597786,
      "loss": 0.8029,
      "num_input_tokens_seen": 36796680,
      "step": 63390
    },
    {
      "epoch": 9.442210306821567,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.019028599207457902,
      "loss": 0.7792,
      "num_input_tokens_seen": 36799656,
      "step": 63395
    },
    {
      "epoch": 9.442955019362525,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01902672115225893,
      "loss": 0.7934,
      "num_input_tokens_seen": 36802376,
      "step": 63400
    },
    {
      "epoch": 9.443699731903486,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.019024843029032602,
      "loss": 0.8352,
      "num_input_tokens_seen": 36804968,
      "step": 63405
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.019022964837810637,
      "loss": 0.8004,
      "num_input_tokens_seen": 36808136,
      "step": 63410
    },
    {
      "epoch": 9.445189156985403,
      "grad_norm": 0.025390625,
      "learning_rate": 0.019021086578624773,
      "loss": 0.8089,
      "num_input_tokens_seen": 36810920,
      "step": 63415
    },
    {
      "epoch": 9.445933869526362,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01901920825150673,
      "loss": 0.7938,
      "num_input_tokens_seen": 36813992,
      "step": 63420
    },
    {
      "epoch": 9.446678582067323,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.01901732985648826,
      "loss": 0.7788,
      "num_input_tokens_seen": 36816744,
      "step": 63425
    },
    {
      "epoch": 9.447423294608281,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01901545139360108,
      "loss": 0.8253,
      "num_input_tokens_seen": 36819880,
      "step": 63430
    },
    {
      "epoch": 9.44816800714924,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.019013572862876926,
      "loss": 0.814,
      "num_input_tokens_seen": 36822920,
      "step": 63435
    },
    {
      "epoch": 9.448912719690199,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.019011694264347546,
      "loss": 0.8166,
      "num_input_tokens_seen": 36825960,
      "step": 63440
    },
    {
      "epoch": 9.44965743223116,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.019009815598044662,
      "loss": 0.7882,
      "num_input_tokens_seen": 36829064,
      "step": 63445
    },
    {
      "epoch": 9.450402144772118,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01900793686400002,
      "loss": 0.8238,
      "num_input_tokens_seen": 36832040,
      "step": 63450
    },
    {
      "epoch": 9.451146857313077,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.01900605806224536,
      "loss": 0.8262,
      "num_input_tokens_seen": 36834952,
      "step": 63455
    },
    {
      "epoch": 9.451891569854036,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.019004179192812415,
      "loss": 0.8066,
      "num_input_tokens_seen": 36837832,
      "step": 63460
    },
    {
      "epoch": 9.452636282394996,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.019002300255732938,
      "loss": 0.7961,
      "num_input_tokens_seen": 36840904,
      "step": 63465
    },
    {
      "epoch": 9.453380994935955,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.019000421251038666,
      "loss": 0.7878,
      "num_input_tokens_seen": 36843688,
      "step": 63470
    },
    {
      "epoch": 9.454125707476914,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.018998542178761346,
      "loss": 0.8049,
      "num_input_tokens_seen": 36846440,
      "step": 63475
    },
    {
      "epoch": 9.454870420017873,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.01899666303893271,
      "loss": 0.819,
      "num_input_tokens_seen": 36849416,
      "step": 63480
    },
    {
      "epoch": 9.455615132558833,
      "grad_norm": 0.03125,
      "learning_rate": 0.01899478383158452,
      "loss": 0.8117,
      "num_input_tokens_seen": 36852456,
      "step": 63485
    },
    {
      "epoch": 9.456359845099792,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.018992904556748517,
      "loss": 0.8061,
      "num_input_tokens_seen": 36855368,
      "step": 63490
    },
    {
      "epoch": 9.45710455764075,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.018991025214456447,
      "loss": 0.8045,
      "num_input_tokens_seen": 36858088,
      "step": 63495
    },
    {
      "epoch": 9.45784927018171,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.01898914580474007,
      "loss": 0.7927,
      "num_input_tokens_seen": 36860808,
      "step": 63500
    },
    {
      "epoch": 9.458593982722668,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.018987266327631123,
      "loss": 0.7944,
      "num_input_tokens_seen": 36863624,
      "step": 63505
    },
    {
      "epoch": 9.459338695263629,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.018985386783161366,
      "loss": 0.7962,
      "num_input_tokens_seen": 36866248,
      "step": 63510
    },
    {
      "epoch": 9.460083407804587,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.018983507171362546,
      "loss": 0.8012,
      "num_input_tokens_seen": 36868936,
      "step": 63515
    },
    {
      "epoch": 9.460828120345546,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.018981627492266423,
      "loss": 0.7975,
      "num_input_tokens_seen": 36871784,
      "step": 63520
    },
    {
      "epoch": 9.461572832886505,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.018979747745904753,
      "loss": 0.8131,
      "num_input_tokens_seen": 36874984,
      "step": 63525
    },
    {
      "epoch": 9.462317545427466,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.018977867932309284,
      "loss": 0.7894,
      "num_input_tokens_seen": 36877736,
      "step": 63530
    },
    {
      "epoch": 9.463062257968424,
      "grad_norm": 0.04296875,
      "learning_rate": 0.018975988051511787,
      "loss": 0.8109,
      "num_input_tokens_seen": 36880360,
      "step": 63535
    },
    {
      "epoch": 9.463806970509383,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.018974108103544007,
      "loss": 0.8047,
      "num_input_tokens_seen": 36883208,
      "step": 63540
    },
    {
      "epoch": 9.464551683050342,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.018972228088437713,
      "loss": 0.787,
      "num_input_tokens_seen": 36886184,
      "step": 63545
    },
    {
      "epoch": 9.465296395591302,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.018970348006224658,
      "loss": 0.8085,
      "num_input_tokens_seen": 36889160,
      "step": 63550
    },
    {
      "epoch": 9.466041108132261,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.018968467856936615,
      "loss": 0.8125,
      "num_input_tokens_seen": 36892040,
      "step": 63555
    },
    {
      "epoch": 9.46678582067322,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.018966587640605336,
      "loss": 0.8137,
      "num_input_tokens_seen": 36894856,
      "step": 63560
    },
    {
      "epoch": 9.467530533214179,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.01896470735726259,
      "loss": 0.7997,
      "num_input_tokens_seen": 36897768,
      "step": 63565
    },
    {
      "epoch": 9.46827524575514,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.018962827006940144,
      "loss": 0.7947,
      "num_input_tokens_seen": 36900616,
      "step": 63570
    },
    {
      "epoch": 9.469019958296098,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.018960946589669764,
      "loss": 0.8133,
      "num_input_tokens_seen": 36903592,
      "step": 63575
    },
    {
      "epoch": 9.469764670837057,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.018959066105483216,
      "loss": 0.8119,
      "num_input_tokens_seen": 36906472,
      "step": 63580
    },
    {
      "epoch": 9.470509383378015,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.01895718555441227,
      "loss": 0.7932,
      "num_input_tokens_seen": 36909896,
      "step": 63585
    },
    {
      "epoch": 9.471254095918976,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.0189553049364887,
      "loss": 0.7968,
      "num_input_tokens_seen": 36912840,
      "step": 63590
    },
    {
      "epoch": 9.471998808459935,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.018953424251744268,
      "loss": 0.8125,
      "num_input_tokens_seen": 36916040,
      "step": 63595
    },
    {
      "epoch": 9.472743521000893,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.018951543500210753,
      "loss": 0.7882,
      "num_input_tokens_seen": 36918824,
      "step": 63600
    },
    {
      "epoch": 9.473488233541852,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01894966268191993,
      "loss": 0.8044,
      "num_input_tokens_seen": 36921864,
      "step": 63605
    },
    {
      "epoch": 9.474232946082813,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.018947781796903564,
      "loss": 0.7806,
      "num_input_tokens_seen": 36924456,
      "step": 63610
    },
    {
      "epoch": 9.474977658623772,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.01894590084519344,
      "loss": 0.784,
      "num_input_tokens_seen": 36927368,
      "step": 63615
    },
    {
      "epoch": 9.47572237116473,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.018944019826821333,
      "loss": 0.8142,
      "num_input_tokens_seen": 36930760,
      "step": 63620
    },
    {
      "epoch": 9.476467083705689,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.018942138741819017,
      "loss": 0.8101,
      "num_input_tokens_seen": 36933448,
      "step": 63625
    },
    {
      "epoch": 9.47721179624665,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.018940257590218276,
      "loss": 0.7943,
      "num_input_tokens_seen": 36936232,
      "step": 63630
    },
    {
      "epoch": 9.477956508787608,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.018938376372050886,
      "loss": 0.799,
      "num_input_tokens_seen": 36939080,
      "step": 63635
    },
    {
      "epoch": 9.478701221328567,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.018936495087348633,
      "loss": 0.8113,
      "num_input_tokens_seen": 36942184,
      "step": 63640
    },
    {
      "epoch": 9.479445933869526,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.01893461373614329,
      "loss": 0.817,
      "num_input_tokens_seen": 36945192,
      "step": 63645
    },
    {
      "epoch": 9.480190646410486,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.018932732318466653,
      "loss": 0.8108,
      "num_input_tokens_seen": 36948264,
      "step": 63650
    },
    {
      "epoch": 9.480935358951445,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.0189308508343505,
      "loss": 0.8125,
      "num_input_tokens_seen": 36951240,
      "step": 63655
    },
    {
      "epoch": 9.481680071492404,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.018928969283826615,
      "loss": 0.7913,
      "num_input_tokens_seen": 36954152,
      "step": 63660
    },
    {
      "epoch": 9.482424784033363,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.018927087666926783,
      "loss": 0.8021,
      "num_input_tokens_seen": 36957096,
      "step": 63665
    },
    {
      "epoch": 9.483169496574323,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.018925205983682802,
      "loss": 0.8108,
      "num_input_tokens_seen": 36959976,
      "step": 63670
    },
    {
      "epoch": 9.483914209115282,
      "grad_norm": 0.0625,
      "learning_rate": 0.01892332423412646,
      "loss": 0.8303,
      "num_input_tokens_seen": 36962664,
      "step": 63675
    },
    {
      "epoch": 9.48465892165624,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.01892144241828954,
      "loss": 0.8192,
      "num_input_tokens_seen": 36965416,
      "step": 63680
    },
    {
      "epoch": 9.4854036341972,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.018919560536203826,
      "loss": 0.7959,
      "num_input_tokens_seen": 36968232,
      "step": 63685
    },
    {
      "epoch": 9.486148346738158,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.018917678587901127,
      "loss": 0.8068,
      "num_input_tokens_seen": 36970952,
      "step": 63690
    },
    {
      "epoch": 9.486893059279119,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.018915796573413232,
      "loss": 0.7919,
      "num_input_tokens_seen": 36973768,
      "step": 63695
    },
    {
      "epoch": 9.487637771820078,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.018913914492771935,
      "loss": 0.7779,
      "num_input_tokens_seen": 36976712,
      "step": 63700
    },
    {
      "epoch": 9.488382484361036,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.018912032346009026,
      "loss": 0.8135,
      "num_input_tokens_seen": 36979592,
      "step": 63705
    },
    {
      "epoch": 9.489127196901995,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.01891015013315631,
      "loss": 0.7979,
      "num_input_tokens_seen": 36982184,
      "step": 63710
    },
    {
      "epoch": 9.489871909442956,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01890826785424558,
      "loss": 0.8054,
      "num_input_tokens_seen": 36984840,
      "step": 63715
    },
    {
      "epoch": 9.490616621983914,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.018906385509308635,
      "loss": 0.8011,
      "num_input_tokens_seen": 36987944,
      "step": 63720
    },
    {
      "epoch": 9.491361334524873,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.018904503098377278,
      "loss": 0.7925,
      "num_input_tokens_seen": 36990728,
      "step": 63725
    },
    {
      "epoch": 9.492106047065832,
      "grad_norm": 0.03515625,
      "learning_rate": 0.018902620621483308,
      "loss": 0.8078,
      "num_input_tokens_seen": 36993480,
      "step": 63730
    },
    {
      "epoch": 9.492850759606792,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.018900738078658535,
      "loss": 0.8123,
      "num_input_tokens_seen": 36996104,
      "step": 63735
    },
    {
      "epoch": 9.493595472147751,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.01889885546993475,
      "loss": 0.8062,
      "num_input_tokens_seen": 36998856,
      "step": 63740
    },
    {
      "epoch": 9.49434018468871,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.018896972795343768,
      "loss": 0.8041,
      "num_input_tokens_seen": 37001448,
      "step": 63745
    },
    {
      "epoch": 9.495084897229669,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.018895090054917386,
      "loss": 0.8004,
      "num_input_tokens_seen": 37004360,
      "step": 63750
    },
    {
      "epoch": 9.49582960977063,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.018893207248687423,
      "loss": 0.8096,
      "num_input_tokens_seen": 37006984,
      "step": 63755
    },
    {
      "epoch": 9.496574322311588,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.01889132437668568,
      "loss": 0.7892,
      "num_input_tokens_seen": 37009960,
      "step": 63760
    },
    {
      "epoch": 9.497319034852547,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.018889441438943964,
      "loss": 0.807,
      "num_input_tokens_seen": 37013000,
      "step": 63765
    },
    {
      "epoch": 9.498063747393505,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.01888755843549409,
      "loss": 0.7894,
      "num_input_tokens_seen": 37015784,
      "step": 63770
    },
    {
      "epoch": 9.498808459934466,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.018885675366367868,
      "loss": 0.8195,
      "num_input_tokens_seen": 37018344,
      "step": 63775
    },
    {
      "epoch": 9.499553172475425,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.018883792231597106,
      "loss": 0.7783,
      "num_input_tokens_seen": 37021416,
      "step": 63780
    },
    {
      "epoch": 9.500297885016384,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01888190903121362,
      "loss": 0.7974,
      "num_input_tokens_seen": 37024552,
      "step": 63785
    },
    {
      "epoch": 9.501042597557342,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.018880025765249234,
      "loss": 0.8019,
      "num_input_tokens_seen": 37027464,
      "step": 63790
    },
    {
      "epoch": 9.501787310098303,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.018878142433735753,
      "loss": 0.7913,
      "num_input_tokens_seen": 37030312,
      "step": 63795
    },
    {
      "epoch": 9.502532022639262,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.018876259036704996,
      "loss": 0.7975,
      "num_input_tokens_seen": 37033128,
      "step": 63800
    },
    {
      "epoch": 9.50327673518022,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.018874375574188786,
      "loss": 0.794,
      "num_input_tokens_seen": 37036040,
      "step": 63805
    },
    {
      "epoch": 9.504021447721179,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.018872492046218933,
      "loss": 0.8022,
      "num_input_tokens_seen": 37038952,
      "step": 63810
    },
    {
      "epoch": 9.50476616026214,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.018870608452827262,
      "loss": 0.7898,
      "num_input_tokens_seen": 37041928,
      "step": 63815
    },
    {
      "epoch": 9.505510872803098,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.018868724794045597,
      "loss": 0.7914,
      "num_input_tokens_seen": 37044680,
      "step": 63820
    },
    {
      "epoch": 9.506255585344057,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01886684106990576,
      "loss": 0.8005,
      "num_input_tokens_seen": 37047784,
      "step": 63825
    },
    {
      "epoch": 9.507000297885016,
      "grad_norm": 0.025390625,
      "learning_rate": 0.018864957280439573,
      "loss": 0.8032,
      "num_input_tokens_seen": 37050664,
      "step": 63830
    },
    {
      "epoch": 9.507745010425975,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.01886307342567886,
      "loss": 0.7894,
      "num_input_tokens_seen": 37053640,
      "step": 63835
    },
    {
      "epoch": 9.508489722966935,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.018861189505655452,
      "loss": 0.8149,
      "num_input_tokens_seen": 37056648,
      "step": 63840
    },
    {
      "epoch": 9.509234435507894,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.018859305520401166,
      "loss": 0.827,
      "num_input_tokens_seen": 37059560,
      "step": 63845
    },
    {
      "epoch": 9.509979148048853,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.018857421469947837,
      "loss": 0.8016,
      "num_input_tokens_seen": 37062728,
      "step": 63850
    },
    {
      "epoch": 9.510723860589813,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.018855537354327292,
      "loss": 0.8101,
      "num_input_tokens_seen": 37065672,
      "step": 63855
    },
    {
      "epoch": 9.511468573130772,
      "grad_norm": 0.025390625,
      "learning_rate": 0.018853653173571364,
      "loss": 0.8072,
      "num_input_tokens_seen": 37068488,
      "step": 63860
    },
    {
      "epoch": 9.51221328567173,
      "grad_norm": 0.0234375,
      "learning_rate": 0.018851768927711884,
      "loss": 0.8175,
      "num_input_tokens_seen": 37071240,
      "step": 63865
    },
    {
      "epoch": 9.51295799821269,
      "grad_norm": 0.021484375,
      "learning_rate": 0.01884988461678068,
      "loss": 0.7892,
      "num_input_tokens_seen": 37074472,
      "step": 63870
    },
    {
      "epoch": 9.513702710753648,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01884800024080959,
      "loss": 0.7786,
      "num_input_tokens_seen": 37077128,
      "step": 63875
    },
    {
      "epoch": 9.514447423294609,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.018846115799830445,
      "loss": 0.799,
      "num_input_tokens_seen": 37080072,
      "step": 63880
    },
    {
      "epoch": 9.515192135835568,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.01884423129387508,
      "loss": 0.8072,
      "num_input_tokens_seen": 37082856,
      "step": 63885
    },
    {
      "epoch": 9.515936848376526,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.01884234672297534,
      "loss": 0.7868,
      "num_input_tokens_seen": 37085992,
      "step": 63890
    },
    {
      "epoch": 9.516681560917485,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.018840462087163053,
      "loss": 0.8159,
      "num_input_tokens_seen": 37088968,
      "step": 63895
    },
    {
      "epoch": 9.517426273458446,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.018838577386470066,
      "loss": 0.7895,
      "num_input_tokens_seen": 37091976,
      "step": 63900
    },
    {
      "epoch": 9.518170985999404,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.01883669262092821,
      "loss": 0.8212,
      "num_input_tokens_seen": 37095240,
      "step": 63905
    },
    {
      "epoch": 9.518915698540363,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.018834807790569338,
      "loss": 0.8292,
      "num_input_tokens_seen": 37098376,
      "step": 63910
    },
    {
      "epoch": 9.519660411081322,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.018832922895425284,
      "loss": 0.8455,
      "num_input_tokens_seen": 37101352,
      "step": 63915
    },
    {
      "epoch": 9.520405123622282,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.01883103793552789,
      "loss": 0.8092,
      "num_input_tokens_seen": 37104104,
      "step": 63920
    },
    {
      "epoch": 9.521149836163241,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.018829152910909006,
      "loss": 0.805,
      "num_input_tokens_seen": 37106984,
      "step": 63925
    },
    {
      "epoch": 9.5218945487042,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.018827267821600473,
      "loss": 0.8154,
      "num_input_tokens_seen": 37110024,
      "step": 63930
    },
    {
      "epoch": 9.522639261245159,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.018825382667634144,
      "loss": 0.7989,
      "num_input_tokens_seen": 37112872,
      "step": 63935
    },
    {
      "epoch": 9.52338397378612,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.018823497449041864,
      "loss": 0.8114,
      "num_input_tokens_seen": 37115656,
      "step": 63940
    },
    {
      "epoch": 9.524128686327078,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.018821612165855477,
      "loss": 0.8035,
      "num_input_tokens_seen": 37118568,
      "step": 63945
    },
    {
      "epoch": 9.524873398868037,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.018819726818106834,
      "loss": 0.8101,
      "num_input_tokens_seen": 37121416,
      "step": 63950
    },
    {
      "epoch": 9.525618111408996,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.018817841405827797,
      "loss": 0.8223,
      "num_input_tokens_seen": 37124488,
      "step": 63955
    },
    {
      "epoch": 9.526362823949956,
      "grad_norm": 0.02734375,
      "learning_rate": 0.018815955929050204,
      "loss": 0.7969,
      "num_input_tokens_seen": 37127400,
      "step": 63960
    },
    {
      "epoch": 9.527107536490915,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.01881407038780591,
      "loss": 0.8093,
      "num_input_tokens_seen": 37130152,
      "step": 63965
    },
    {
      "epoch": 9.527852249031874,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.018812184782126783,
      "loss": 0.7845,
      "num_input_tokens_seen": 37133096,
      "step": 63970
    },
    {
      "epoch": 9.528596961572832,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.01881029911204466,
      "loss": 0.8179,
      "num_input_tokens_seen": 37136072,
      "step": 63975
    },
    {
      "epoch": 9.529341674113793,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.01880841337759141,
      "loss": 0.7788,
      "num_input_tokens_seen": 37139240,
      "step": 63980
    },
    {
      "epoch": 9.530086386654752,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.018806527578798887,
      "loss": 0.7858,
      "num_input_tokens_seen": 37142152,
      "step": 63985
    },
    {
      "epoch": 9.53083109919571,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.018804641715698945,
      "loss": 0.809,
      "num_input_tokens_seen": 37145000,
      "step": 63990
    },
    {
      "epoch": 9.53157581173667,
      "grad_norm": 0.029296875,
      "learning_rate": 0.018802755788323456,
      "loss": 0.7812,
      "num_input_tokens_seen": 37148008,
      "step": 63995
    },
    {
      "epoch": 9.53232052427763,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.01880086979670427,
      "loss": 0.822,
      "num_input_tokens_seen": 37150664,
      "step": 64000
    },
    {
      "epoch": 9.533065236818588,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.01879898374087325,
      "loss": 0.7929,
      "num_input_tokens_seen": 37153768,
      "step": 64005
    },
    {
      "epoch": 9.533809949359547,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.01879709762086226,
      "loss": 0.7854,
      "num_input_tokens_seen": 37157000,
      "step": 64010
    },
    {
      "epoch": 9.534554661900506,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.018795211436703166,
      "loss": 0.7873,
      "num_input_tokens_seen": 37160072,
      "step": 64015
    },
    {
      "epoch": 9.535299374441465,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.018793325188427832,
      "loss": 0.8199,
      "num_input_tokens_seen": 37162856,
      "step": 64020
    },
    {
      "epoch": 9.536044086982425,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01879143887606812,
      "loss": 0.7688,
      "num_input_tokens_seen": 37166056,
      "step": 64025
    },
    {
      "epoch": 9.536788799523384,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.01878955249965591,
      "loss": 0.7716,
      "num_input_tokens_seen": 37169128,
      "step": 64030
    },
    {
      "epoch": 9.537533512064343,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.01878766605922306,
      "loss": 0.7912,
      "num_input_tokens_seen": 37172136,
      "step": 64035
    },
    {
      "epoch": 9.538278224605303,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.01878577955480144,
      "loss": 0.7916,
      "num_input_tokens_seen": 37175272,
      "step": 64040
    },
    {
      "epoch": 9.539022937146262,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.018783892986422922,
      "loss": 0.8044,
      "num_input_tokens_seen": 37178216,
      "step": 64045
    },
    {
      "epoch": 9.53976764968722,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.018782006354119377,
      "loss": 0.7863,
      "num_input_tokens_seen": 37181064,
      "step": 64050
    },
    {
      "epoch": 9.54051236222818,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.01878011965792268,
      "loss": 0.8003,
      "num_input_tokens_seen": 37184104,
      "step": 64055
    },
    {
      "epoch": 9.541257074769138,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0187782328978647,
      "loss": 0.8412,
      "num_input_tokens_seen": 37186792,
      "step": 64060
    },
    {
      "epoch": 9.542001787310099,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.01877634607397732,
      "loss": 0.8121,
      "num_input_tokens_seen": 37189704,
      "step": 64065
    },
    {
      "epoch": 9.542746499851058,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.018774459186292405,
      "loss": 0.8051,
      "num_input_tokens_seen": 37192712,
      "step": 64070
    },
    {
      "epoch": 9.543491212392016,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01877257223484185,
      "loss": 0.8014,
      "num_input_tokens_seen": 37195656,
      "step": 64075
    },
    {
      "epoch": 9.544235924932975,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.018770685219657513,
      "loss": 0.7714,
      "num_input_tokens_seen": 37198376,
      "step": 64080
    },
    {
      "epoch": 9.544980637473936,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.01876879814077128,
      "loss": 0.7958,
      "num_input_tokens_seen": 37201448,
      "step": 64085
    },
    {
      "epoch": 9.545725350014894,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.018766910998215033,
      "loss": 0.7883,
      "num_input_tokens_seen": 37204488,
      "step": 64090
    },
    {
      "epoch": 9.546470062555853,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.018765023792020656,
      "loss": 0.8056,
      "num_input_tokens_seen": 37207624,
      "step": 64095
    },
    {
      "epoch": 9.547214775096812,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.018763136522220027,
      "loss": 0.7871,
      "num_input_tokens_seen": 37210664,
      "step": 64100
    },
    {
      "epoch": 9.547959487637772,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.01876124918884503,
      "loss": 0.8118,
      "num_input_tokens_seen": 37213608,
      "step": 64105
    },
    {
      "epoch": 9.548704200178731,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.018759361791927553,
      "loss": 0.7905,
      "num_input_tokens_seen": 37216584,
      "step": 64110
    },
    {
      "epoch": 9.54944891271969,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.018757474331499484,
      "loss": 0.7977,
      "num_input_tokens_seen": 37219368,
      "step": 64115
    },
    {
      "epoch": 9.550193625260649,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.018755586807592695,
      "loss": 0.8285,
      "num_input_tokens_seen": 37222280,
      "step": 64120
    },
    {
      "epoch": 9.55093833780161,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.01875369922023909,
      "loss": 0.7824,
      "num_input_tokens_seen": 37224968,
      "step": 64125
    },
    {
      "epoch": 9.551683050342568,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.01875181156947055,
      "loss": 0.8037,
      "num_input_tokens_seen": 37227848,
      "step": 64130
    },
    {
      "epoch": 9.552427762883527,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01874992385531897,
      "loss": 0.8171,
      "num_input_tokens_seen": 37230568,
      "step": 64135
    },
    {
      "epoch": 9.553172475424486,
      "grad_norm": 0.025390625,
      "learning_rate": 0.018748036077816235,
      "loss": 0.8047,
      "num_input_tokens_seen": 37233512,
      "step": 64140
    },
    {
      "epoch": 9.553917187965446,
      "grad_norm": 0.03125,
      "learning_rate": 0.01874614823699424,
      "loss": 0.7851,
      "num_input_tokens_seen": 37236296,
      "step": 64145
    },
    {
      "epoch": 9.554661900506405,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.018744260332884875,
      "loss": 0.8113,
      "num_input_tokens_seen": 37239336,
      "step": 64150
    },
    {
      "epoch": 9.555406613047364,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.01874237236552004,
      "loss": 0.8368,
      "num_input_tokens_seen": 37242344,
      "step": 64155
    },
    {
      "epoch": 9.556151325588322,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01874048433493163,
      "loss": 0.8119,
      "num_input_tokens_seen": 37244904,
      "step": 64160
    },
    {
      "epoch": 9.556896038129283,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.018738596241151536,
      "loss": 0.7813,
      "num_input_tokens_seen": 37247816,
      "step": 64165
    },
    {
      "epoch": 9.557640750670242,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.01873670808421166,
      "loss": 0.8065,
      "num_input_tokens_seen": 37250568,
      "step": 64170
    },
    {
      "epoch": 9.5583854632112,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.018734819864143895,
      "loss": 0.7844,
      "num_input_tokens_seen": 37253512,
      "step": 64175
    },
    {
      "epoch": 9.55913017575216,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.018732931580980144,
      "loss": 0.819,
      "num_input_tokens_seen": 37256616,
      "step": 64180
    },
    {
      "epoch": 9.55987488829312,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.018731043234752314,
      "loss": 0.8021,
      "num_input_tokens_seen": 37259752,
      "step": 64185
    },
    {
      "epoch": 9.560619600834078,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.018729154825492294,
      "loss": 0.8023,
      "num_input_tokens_seen": 37262824,
      "step": 64190
    },
    {
      "epoch": 9.561364313375037,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.018727266353232,
      "loss": 0.8079,
      "num_input_tokens_seen": 37265544,
      "step": 64195
    },
    {
      "epoch": 9.562109025915996,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.018725377818003326,
      "loss": 0.8146,
      "num_input_tokens_seen": 37268520,
      "step": 64200
    },
    {
      "epoch": 9.562853738456955,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.01872348921983818,
      "loss": 0.7986,
      "num_input_tokens_seen": 37271368,
      "step": 64205
    },
    {
      "epoch": 9.563598450997915,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.018721600558768467,
      "loss": 0.7772,
      "num_input_tokens_seen": 37274216,
      "step": 64210
    },
    {
      "epoch": 9.564343163538874,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.018719711834826094,
      "loss": 0.8076,
      "num_input_tokens_seen": 37277128,
      "step": 64215
    },
    {
      "epoch": 9.565087876079833,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.018717823048042974,
      "loss": 0.8075,
      "num_input_tokens_seen": 37279912,
      "step": 64220
    },
    {
      "epoch": 9.565832588620792,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.018715934198451006,
      "loss": 0.7941,
      "num_input_tokens_seen": 37282632,
      "step": 64225
    },
    {
      "epoch": 9.566577301161752,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.018714045286082115,
      "loss": 0.7915,
      "num_input_tokens_seen": 37285416,
      "step": 64230
    },
    {
      "epoch": 9.56732201370271,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0187121563109682,
      "loss": 0.789,
      "num_input_tokens_seen": 37288200,
      "step": 64235
    },
    {
      "epoch": 9.56806672624367,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.018710267273141176,
      "loss": 0.795,
      "num_input_tokens_seen": 37291048,
      "step": 64240
    },
    {
      "epoch": 9.568811438784628,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.018708378172632958,
      "loss": 0.7997,
      "num_input_tokens_seen": 37294152,
      "step": 64245
    },
    {
      "epoch": 9.569556151325589,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.01870648900947546,
      "loss": 0.8082,
      "num_input_tokens_seen": 37297320,
      "step": 64250
    },
    {
      "epoch": 9.570300863866548,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.018704599783700593,
      "loss": 0.7856,
      "num_input_tokens_seen": 37300424,
      "step": 64255
    },
    {
      "epoch": 9.571045576407506,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.01870271049534028,
      "loss": 0.7998,
      "num_input_tokens_seen": 37303144,
      "step": 64260
    },
    {
      "epoch": 9.571790288948465,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.01870082114442644,
      "loss": 0.7815,
      "num_input_tokens_seen": 37305928,
      "step": 64265
    },
    {
      "epoch": 9.572535001489426,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.018698931730990985,
      "loss": 0.7885,
      "num_input_tokens_seen": 37308904,
      "step": 64270
    },
    {
      "epoch": 9.573279714030384,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01869704225506584,
      "loss": 0.7821,
      "num_input_tokens_seen": 37311688,
      "step": 64275
    },
    {
      "epoch": 9.574024426571343,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.018695152716682924,
      "loss": 0.7922,
      "num_input_tokens_seen": 37314984,
      "step": 64280
    },
    {
      "epoch": 9.574769139112302,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.018693263115874156,
      "loss": 0.8043,
      "num_input_tokens_seen": 37317864,
      "step": 64285
    },
    {
      "epoch": 9.575513851653263,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.018691373452671464,
      "loss": 0.8257,
      "num_input_tokens_seen": 37321096,
      "step": 64290
    },
    {
      "epoch": 9.576258564194221,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.018689483727106763,
      "loss": 0.82,
      "num_input_tokens_seen": 37323784,
      "step": 64295
    },
    {
      "epoch": 9.57700327673518,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.018687593939211992,
      "loss": 0.7789,
      "num_input_tokens_seen": 37326632,
      "step": 64300
    },
    {
      "epoch": 9.577747989276139,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.018685704089019067,
      "loss": 0.8236,
      "num_input_tokens_seen": 37329768,
      "step": 64305
    },
    {
      "epoch": 9.5784927018171,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.018683814176559916,
      "loss": 0.8048,
      "num_input_tokens_seen": 37332488,
      "step": 64310
    },
    {
      "epoch": 9.579237414358058,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.018681924201866467,
      "loss": 0.7919,
      "num_input_tokens_seen": 37335304,
      "step": 64315
    },
    {
      "epoch": 9.579982126899017,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.018680034164970657,
      "loss": 0.7808,
      "num_input_tokens_seen": 37338312,
      "step": 64320
    },
    {
      "epoch": 9.580726839439976,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.018678144065904404,
      "loss": 0.8008,
      "num_input_tokens_seen": 37341256,
      "step": 64325
    },
    {
      "epoch": 9.581471551980936,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.018676253904699645,
      "loss": 0.8056,
      "num_input_tokens_seen": 37343816,
      "step": 64330
    },
    {
      "epoch": 9.582216264521895,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.01867436368138832,
      "loss": 0.7933,
      "num_input_tokens_seen": 37346888,
      "step": 64335
    },
    {
      "epoch": 9.582960977062854,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.018672473396002347,
      "loss": 0.8127,
      "num_input_tokens_seen": 37349704,
      "step": 64340
    },
    {
      "epoch": 9.583705689603812,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.01867058304857367,
      "loss": 0.7907,
      "num_input_tokens_seen": 37353000,
      "step": 64345
    },
    {
      "epoch": 9.584450402144771,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.018668692639134222,
      "loss": 0.7895,
      "num_input_tokens_seen": 37355912,
      "step": 64350
    },
    {
      "epoch": 9.585195114685732,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.018666802167715945,
      "loss": 0.8086,
      "num_input_tokens_seen": 37359016,
      "step": 64355
    },
    {
      "epoch": 9.58593982722669,
      "grad_norm": 0.017578125,
      "learning_rate": 0.01866491163435077,
      "loss": 0.8214,
      "num_input_tokens_seen": 37361640,
      "step": 64360
    },
    {
      "epoch": 9.58668453976765,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.018663021039070637,
      "loss": 0.7995,
      "num_input_tokens_seen": 37364840,
      "step": 64365
    },
    {
      "epoch": 9.58742925230861,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.01866113038190749,
      "loss": 0.7741,
      "num_input_tokens_seen": 37367656,
      "step": 64370
    },
    {
      "epoch": 9.588173964849569,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.018659239662893262,
      "loss": 0.8259,
      "num_input_tokens_seen": 37370824,
      "step": 64375
    },
    {
      "epoch": 9.588918677390527,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.0186573488820599,
      "loss": 0.8215,
      "num_input_tokens_seen": 37373960,
      "step": 64380
    },
    {
      "epoch": 9.589663389931486,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.018655458039439347,
      "loss": 0.8121,
      "num_input_tokens_seen": 37376584,
      "step": 64385
    },
    {
      "epoch": 9.590408102472445,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.018653567135063543,
      "loss": 0.7943,
      "num_input_tokens_seen": 37379304,
      "step": 64390
    },
    {
      "epoch": 9.591152815013405,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.01865167616896444,
      "loss": 0.798,
      "num_input_tokens_seen": 37382280,
      "step": 64395
    },
    {
      "epoch": 9.591897527554364,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.01864978514117398,
      "loss": 0.8132,
      "num_input_tokens_seen": 37384904,
      "step": 64400
    },
    {
      "epoch": 9.592642240095323,
      "grad_norm": 0.0390625,
      "learning_rate": 0.018647894051724106,
      "loss": 0.8266,
      "num_input_tokens_seen": 37387816,
      "step": 64405
    },
    {
      "epoch": 9.593386952636282,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.018646002900646767,
      "loss": 0.7936,
      "num_input_tokens_seen": 37390696,
      "step": 64410
    },
    {
      "epoch": 9.594131665177242,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.018644111687973915,
      "loss": 0.8146,
      "num_input_tokens_seen": 37393576,
      "step": 64415
    },
    {
      "epoch": 9.594876377718201,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.018642220413737507,
      "loss": 0.7933,
      "num_input_tokens_seen": 37396360,
      "step": 64420
    },
    {
      "epoch": 9.59562109025916,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.01864032907796948,
      "loss": 0.797,
      "num_input_tokens_seen": 37399240,
      "step": 64425
    },
    {
      "epoch": 9.596365802800118,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.018638437680701798,
      "loss": 0.7989,
      "num_input_tokens_seen": 37402312,
      "step": 64430
    },
    {
      "epoch": 9.597110515341079,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.018636546221966407,
      "loss": 0.7912,
      "num_input_tokens_seen": 37405032,
      "step": 64435
    },
    {
      "epoch": 9.597855227882038,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.01863465470179526,
      "loss": 0.789,
      "num_input_tokens_seen": 37407912,
      "step": 64440
    },
    {
      "epoch": 9.598599940422996,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.01863276312022032,
      "loss": 0.8061,
      "num_input_tokens_seen": 37410920,
      "step": 64445
    },
    {
      "epoch": 9.599344652963955,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.01863087147727354,
      "loss": 0.7769,
      "num_input_tokens_seen": 37413704,
      "step": 64450
    },
    {
      "epoch": 9.600089365504916,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.018628979772986876,
      "loss": 0.7906,
      "num_input_tokens_seen": 37416808,
      "step": 64455
    },
    {
      "epoch": 9.600834078045875,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.018627088007392285,
      "loss": 0.7962,
      "num_input_tokens_seen": 37419656,
      "step": 64460
    },
    {
      "epoch": 9.601578790586833,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.01862519618052173,
      "loss": 0.7875,
      "num_input_tokens_seen": 37422728,
      "step": 64465
    },
    {
      "epoch": 9.602323503127792,
      "grad_norm": 0.01556396484375,
      "learning_rate": 0.018623304292407168,
      "loss": 0.7942,
      "num_input_tokens_seen": 37425384,
      "step": 64470
    },
    {
      "epoch": 9.603068215668753,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.018621412343080564,
      "loss": 0.7774,
      "num_input_tokens_seen": 37428232,
      "step": 64475
    },
    {
      "epoch": 9.603812928209711,
      "grad_norm": 0.021484375,
      "learning_rate": 0.018619520332573882,
      "loss": 0.7993,
      "num_input_tokens_seen": 37430952,
      "step": 64480
    },
    {
      "epoch": 9.60455764075067,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.01861762826091908,
      "loss": 0.8082,
      "num_input_tokens_seen": 37433768,
      "step": 64485
    },
    {
      "epoch": 9.605302353291629,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.018615736128148125,
      "loss": 0.8159,
      "num_input_tokens_seen": 37436520,
      "step": 64490
    },
    {
      "epoch": 9.60604706583259,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.018613843934292985,
      "loss": 0.8057,
      "num_input_tokens_seen": 37439208,
      "step": 64495
    },
    {
      "epoch": 9.606791778373548,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.018611951679385626,
      "loss": 0.8055,
      "num_input_tokens_seen": 37441960,
      "step": 64500
    },
    {
      "epoch": 9.607536490914507,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.01861005936345801,
      "loss": 0.7984,
      "num_input_tokens_seen": 37445096,
      "step": 64505
    },
    {
      "epoch": 9.608281203455466,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.018608166986542107,
      "loss": 0.7952,
      "num_input_tokens_seen": 37448008,
      "step": 64510
    },
    {
      "epoch": 9.609025915996426,
      "grad_norm": 0.03515625,
      "learning_rate": 0.018606274548669893,
      "loss": 0.763,
      "num_input_tokens_seen": 37451016,
      "step": 64515
    },
    {
      "epoch": 9.609770628537385,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.018604382049873344,
      "loss": 0.794,
      "num_input_tokens_seen": 37453864,
      "step": 64520
    },
    {
      "epoch": 9.610515341078344,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.018602489490184414,
      "loss": 0.7974,
      "num_input_tokens_seen": 37456456,
      "step": 64525
    },
    {
      "epoch": 9.611260053619302,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.01860059686963509,
      "loss": 0.7625,
      "num_input_tokens_seen": 37459208,
      "step": 64530
    },
    {
      "epoch": 9.612004766160261,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.01859870418825734,
      "loss": 0.785,
      "num_input_tokens_seen": 37462024,
      "step": 64535
    },
    {
      "epoch": 9.612749478701222,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.01859681144608314,
      "loss": 0.8078,
      "num_input_tokens_seen": 37464936,
      "step": 64540
    },
    {
      "epoch": 9.61349419124218,
      "grad_norm": 0.0167236328125,
      "learning_rate": 0.018594918643144463,
      "loss": 0.7941,
      "num_input_tokens_seen": 37467816,
      "step": 64545
    },
    {
      "epoch": 9.61423890378314,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.018593025779473294,
      "loss": 0.7817,
      "num_input_tokens_seen": 37470792,
      "step": 64550
    },
    {
      "epoch": 9.6149836163241,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.018591132855101604,
      "loss": 0.7918,
      "num_input_tokens_seen": 37473704,
      "step": 64555
    },
    {
      "epoch": 9.615728328865059,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.018589239870061378,
      "loss": 0.8524,
      "num_input_tokens_seen": 37476392,
      "step": 64560
    },
    {
      "epoch": 9.616473041406017,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.018587346824384593,
      "loss": 0.7956,
      "num_input_tokens_seen": 37479496,
      "step": 64565
    },
    {
      "epoch": 9.617217753946976,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.01858545371810322,
      "loss": 0.7987,
      "num_input_tokens_seen": 37482248,
      "step": 64570
    },
    {
      "epoch": 9.617962466487935,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.01858356055124926,
      "loss": 0.8102,
      "num_input_tokens_seen": 37485064,
      "step": 64575
    },
    {
      "epoch": 9.618707179028895,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.01858166732385468,
      "loss": 0.8583,
      "num_input_tokens_seen": 37488136,
      "step": 64580
    },
    {
      "epoch": 9.619451891569854,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01857977403595147,
      "loss": 0.8012,
      "num_input_tokens_seen": 37491144,
      "step": 64585
    },
    {
      "epoch": 9.620196604110813,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.018577880687571623,
      "loss": 0.7928,
      "num_input_tokens_seen": 37493864,
      "step": 64590
    },
    {
      "epoch": 9.620941316651772,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.018575987278747118,
      "loss": 0.7874,
      "num_input_tokens_seen": 37496648,
      "step": 64595
    },
    {
      "epoch": 9.621686029192732,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01857409380950994,
      "loss": 0.8062,
      "num_input_tokens_seen": 37499528,
      "step": 64600
    },
    {
      "epoch": 9.622430741733691,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.01857220027989208,
      "loss": 0.7785,
      "num_input_tokens_seen": 37502152,
      "step": 64605
    },
    {
      "epoch": 9.62317545427465,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.018570306689925525,
      "loss": 0.8107,
      "num_input_tokens_seen": 37504968,
      "step": 64610
    },
    {
      "epoch": 9.623920166815608,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.018568413039642268,
      "loss": 0.7989,
      "num_input_tokens_seen": 37507848,
      "step": 64615
    },
    {
      "epoch": 9.624664879356569,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.018566519329074296,
      "loss": 0.801,
      "num_input_tokens_seen": 37510888,
      "step": 64620
    },
    {
      "epoch": 9.625409591897528,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.01856462555825361,
      "loss": 0.7977,
      "num_input_tokens_seen": 37513768,
      "step": 64625
    },
    {
      "epoch": 9.626154304438487,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.0185627317272122,
      "loss": 0.7898,
      "num_input_tokens_seen": 37516552,
      "step": 64630
    },
    {
      "epoch": 9.626899016979445,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.018560837835982055,
      "loss": 0.8058,
      "num_input_tokens_seen": 37519528,
      "step": 64635
    },
    {
      "epoch": 9.627643729520406,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.018558943884595173,
      "loss": 0.7997,
      "num_input_tokens_seen": 37522440,
      "step": 64640
    },
    {
      "epoch": 9.628388442061365,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.01855704987308355,
      "loss": 0.7896,
      "num_input_tokens_seen": 37525128,
      "step": 64645
    },
    {
      "epoch": 9.629133154602323,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.018555155801479185,
      "loss": 0.8105,
      "num_input_tokens_seen": 37527848,
      "step": 64650
    },
    {
      "epoch": 9.629877867143282,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.018553261669814077,
      "loss": 0.7924,
      "num_input_tokens_seen": 37531112,
      "step": 64655
    },
    {
      "epoch": 9.630622579684243,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.018551367478120223,
      "loss": 0.8044,
      "num_input_tokens_seen": 37533768,
      "step": 64660
    },
    {
      "epoch": 9.631367292225201,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.018549473226429632,
      "loss": 0.8032,
      "num_input_tokens_seen": 37536584,
      "step": 64665
    },
    {
      "epoch": 9.63211200476616,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.018547578914774293,
      "loss": 0.8175,
      "num_input_tokens_seen": 37539592,
      "step": 64670
    },
    {
      "epoch": 9.632856717307119,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.018545684543186215,
      "loss": 0.8304,
      "num_input_tokens_seen": 37542440,
      "step": 64675
    },
    {
      "epoch": 9.63360142984808,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.018543790111697402,
      "loss": 0.8017,
      "num_input_tokens_seen": 37545000,
      "step": 64680
    },
    {
      "epoch": 9.634346142389038,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.01854189562033985,
      "loss": 0.8147,
      "num_input_tokens_seen": 37547752,
      "step": 64685
    },
    {
      "epoch": 9.635090854929997,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.01854000106914558,
      "loss": 0.8133,
      "num_input_tokens_seen": 37550504,
      "step": 64690
    },
    {
      "epoch": 9.635835567470956,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.018538106458146582,
      "loss": 0.7991,
      "num_input_tokens_seen": 37553576,
      "step": 64695
    },
    {
      "epoch": 9.636580280011916,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.018536211787374877,
      "loss": 0.8098,
      "num_input_tokens_seen": 37556584,
      "step": 64700
    },
    {
      "epoch": 9.637324992552875,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.018534317056862464,
      "loss": 0.8054,
      "num_input_tokens_seen": 37559496,
      "step": 64705
    },
    {
      "epoch": 9.638069705093834,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.018532422266641358,
      "loss": 0.8247,
      "num_input_tokens_seen": 37562696,
      "step": 64710
    },
    {
      "epoch": 9.638814417634793,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.018530527416743568,
      "loss": 0.8244,
      "num_input_tokens_seen": 37565640,
      "step": 64715
    },
    {
      "epoch": 9.639559130175751,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.018528632507201107,
      "loss": 0.7845,
      "num_input_tokens_seen": 37568552,
      "step": 64720
    },
    {
      "epoch": 9.640303842716712,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.018526737538045984,
      "loss": 0.795,
      "num_input_tokens_seen": 37571688,
      "step": 64725
    },
    {
      "epoch": 9.64104855525767,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.01852484250931022,
      "loss": 0.7965,
      "num_input_tokens_seen": 37574408,
      "step": 64730
    },
    {
      "epoch": 9.64179326779863,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.018522947421025816,
      "loss": 0.7769,
      "num_input_tokens_seen": 37577384,
      "step": 64735
    },
    {
      "epoch": 9.642537980339588,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.0185210522732248,
      "loss": 0.8083,
      "num_input_tokens_seen": 37580360,
      "step": 64740
    },
    {
      "epoch": 9.643282692880549,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.018519157065939183,
      "loss": 0.8078,
      "num_input_tokens_seen": 37583080,
      "step": 64745
    },
    {
      "epoch": 9.644027405421507,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.018517261799200987,
      "loss": 0.7793,
      "num_input_tokens_seen": 37585896,
      "step": 64750
    },
    {
      "epoch": 9.644772117962466,
      "grad_norm": 0.0546875,
      "learning_rate": 0.018515366473042225,
      "loss": 0.8101,
      "num_input_tokens_seen": 37588712,
      "step": 64755
    },
    {
      "epoch": 9.645516830503425,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.01851347108749492,
      "loss": 0.8058,
      "num_input_tokens_seen": 37591752,
      "step": 64760
    },
    {
      "epoch": 9.646261543044385,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.018511575642591097,
      "loss": 0.7899,
      "num_input_tokens_seen": 37594440,
      "step": 64765
    },
    {
      "epoch": 9.647006255585344,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.018509680138362766,
      "loss": 0.7999,
      "num_input_tokens_seen": 37597576,
      "step": 64770
    },
    {
      "epoch": 9.647750968126303,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.018507784574841958,
      "loss": 0.8126,
      "num_input_tokens_seen": 37600392,
      "step": 64775
    },
    {
      "epoch": 9.648495680667262,
      "grad_norm": 0.0234375,
      "learning_rate": 0.018505888952060692,
      "loss": 0.7931,
      "num_input_tokens_seen": 37603528,
      "step": 64780
    },
    {
      "epoch": 9.649240393208222,
      "grad_norm": 0.02734375,
      "learning_rate": 0.018503993270051,
      "loss": 0.8163,
      "num_input_tokens_seen": 37606280,
      "step": 64785
    },
    {
      "epoch": 9.649985105749181,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.018502097528844897,
      "loss": 0.792,
      "num_input_tokens_seen": 37608904,
      "step": 64790
    },
    {
      "epoch": 9.65072981829014,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.01850020172847442,
      "loss": 0.7933,
      "num_input_tokens_seen": 37611944,
      "step": 64795
    },
    {
      "epoch": 9.651474530831099,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.0184983058689716,
      "loss": 0.7879,
      "num_input_tokens_seen": 37614792,
      "step": 64800
    },
    {
      "epoch": 9.652219243372059,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.018496409950368446,
      "loss": 0.7982,
      "num_input_tokens_seen": 37617512,
      "step": 64805
    },
    {
      "epoch": 9.652963955913018,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.018494513972697003,
      "loss": 0.8043,
      "num_input_tokens_seen": 37620200,
      "step": 64810
    },
    {
      "epoch": 9.653708668453977,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.0184926179359893,
      "loss": 0.799,
      "num_input_tokens_seen": 37622984,
      "step": 64815
    },
    {
      "epoch": 9.654453380994935,
      "grad_norm": 0.0146484375,
      "learning_rate": 0.018490721840277363,
      "loss": 0.8004,
      "num_input_tokens_seen": 37625608,
      "step": 64820
    },
    {
      "epoch": 9.655198093535896,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01848882568559323,
      "loss": 0.8211,
      "num_input_tokens_seen": 37628328,
      "step": 64825
    },
    {
      "epoch": 9.655942806076855,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.018486929471968932,
      "loss": 0.7966,
      "num_input_tokens_seen": 37631176,
      "step": 64830
    },
    {
      "epoch": 9.656687518617813,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.018485033199436507,
      "loss": 0.7943,
      "num_input_tokens_seen": 37633864,
      "step": 64835
    },
    {
      "epoch": 9.657432231158772,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.01848313686802799,
      "loss": 0.8072,
      "num_input_tokens_seen": 37636584,
      "step": 64840
    },
    {
      "epoch": 9.658176943699733,
      "grad_norm": 0.046875,
      "learning_rate": 0.018481240477775412,
      "loss": 0.7795,
      "num_input_tokens_seen": 37640840,
      "step": 64845
    },
    {
      "epoch": 9.658921656240691,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.018479344028710813,
      "loss": 0.8129,
      "num_input_tokens_seen": 37643848,
      "step": 64850
    },
    {
      "epoch": 9.65966636878165,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.018477447520866233,
      "loss": 0.7758,
      "num_input_tokens_seen": 37646600,
      "step": 64855
    },
    {
      "epoch": 9.660411081322609,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.018475550954273712,
      "loss": 0.7952,
      "num_input_tokens_seen": 37649256,
      "step": 64860
    },
    {
      "epoch": 9.66115579386357,
      "grad_norm": 0.0390625,
      "learning_rate": 0.018473654328965295,
      "loss": 0.8289,
      "num_input_tokens_seen": 37652264,
      "step": 64865
    },
    {
      "epoch": 9.661900506404528,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.018471757644973018,
      "loss": 0.8143,
      "num_input_tokens_seen": 37655400,
      "step": 64870
    },
    {
      "epoch": 9.662645218945487,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.018469860902328914,
      "loss": 0.8,
      "num_input_tokens_seen": 37658216,
      "step": 64875
    },
    {
      "epoch": 9.663389931486446,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.01846796410106505,
      "loss": 0.8107,
      "num_input_tokens_seen": 37661352,
      "step": 64880
    },
    {
      "epoch": 9.664134644027406,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.018466067241213448,
      "loss": 0.7968,
      "num_input_tokens_seen": 37664200,
      "step": 64885
    },
    {
      "epoch": 9.664879356568365,
      "grad_norm": 0.03515625,
      "learning_rate": 0.018464170322806166,
      "loss": 0.8084,
      "num_input_tokens_seen": 37667176,
      "step": 64890
    },
    {
      "epoch": 9.665624069109324,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.01846227334587525,
      "loss": 0.7883,
      "num_input_tokens_seen": 37669800,
      "step": 64895
    },
    {
      "epoch": 9.666368781650283,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.01846037631045274,
      "loss": 0.8088,
      "num_input_tokens_seen": 37673128,
      "step": 64900
    },
    {
      "epoch": 9.667113494191241,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.018458479216570692,
      "loss": 0.7959,
      "num_input_tokens_seen": 37676136,
      "step": 64905
    },
    {
      "epoch": 9.667858206732202,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.018456582064261148,
      "loss": 0.8126,
      "num_input_tokens_seen": 37679016,
      "step": 64910
    },
    {
      "epoch": 9.66860291927316,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.018454684853556166,
      "loss": 0.806,
      "num_input_tokens_seen": 37681992,
      "step": 64915
    },
    {
      "epoch": 9.66934763181412,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.018452787584487793,
      "loss": 0.7762,
      "num_input_tokens_seen": 37685192,
      "step": 64920
    },
    {
      "epoch": 9.670092344355078,
      "grad_norm": 0.033203125,
      "learning_rate": 0.018450890257088088,
      "loss": 0.8142,
      "num_input_tokens_seen": 37688040,
      "step": 64925
    },
    {
      "epoch": 9.670837056896039,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0184489928713891,
      "loss": 0.7708,
      "num_input_tokens_seen": 37691016,
      "step": 64930
    },
    {
      "epoch": 9.671581769436997,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.018447095427422875,
      "loss": 0.8145,
      "num_input_tokens_seen": 37693736,
      "step": 64935
    },
    {
      "epoch": 9.672326481977956,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.018445197925221483,
      "loss": 0.8049,
      "num_input_tokens_seen": 37696808,
      "step": 64940
    },
    {
      "epoch": 9.673071194518915,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.01844330036481697,
      "loss": 0.8069,
      "num_input_tokens_seen": 37699944,
      "step": 64945
    },
    {
      "epoch": 9.673815907059875,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.018441402746241398,
      "loss": 0.7988,
      "num_input_tokens_seen": 37702760,
      "step": 64950
    },
    {
      "epoch": 9.674560619600834,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.01843950506952682,
      "loss": 0.8013,
      "num_input_tokens_seen": 37705640,
      "step": 64955
    },
    {
      "epoch": 9.675305332141793,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.01843760733470531,
      "loss": 0.8033,
      "num_input_tokens_seen": 37708648,
      "step": 64960
    },
    {
      "epoch": 9.676050044682752,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.018435709541808915,
      "loss": 0.8002,
      "num_input_tokens_seen": 37711656,
      "step": 64965
    },
    {
      "epoch": 9.676794757223712,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.018433811690869696,
      "loss": 0.8138,
      "num_input_tokens_seen": 37714408,
      "step": 64970
    },
    {
      "epoch": 9.677539469764671,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.018431913781919714,
      "loss": 0.7925,
      "num_input_tokens_seen": 37717224,
      "step": 64975
    },
    {
      "epoch": 9.67828418230563,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.018430015814991044,
      "loss": 0.7875,
      "num_input_tokens_seen": 37720040,
      "step": 64980
    },
    {
      "epoch": 9.679028894846589,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01842811779011574,
      "loss": 0.8184,
      "num_input_tokens_seen": 37723176,
      "step": 64985
    },
    {
      "epoch": 9.679773607387549,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01842621970732587,
      "loss": 0.7892,
      "num_input_tokens_seen": 37725928,
      "step": 64990
    },
    {
      "epoch": 9.680518319928508,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.0184243215666535,
      "loss": 0.7926,
      "num_input_tokens_seen": 37728680,
      "step": 64995
    },
    {
      "epoch": 9.681263032469467,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0184224233681307,
      "loss": 0.8033,
      "num_input_tokens_seen": 37731912,
      "step": 65000
    },
    {
      "epoch": 9.682007745010425,
      "grad_norm": 0.03125,
      "learning_rate": 0.018420525111789533,
      "loss": 0.8156,
      "num_input_tokens_seen": 37734728,
      "step": 65005
    },
    {
      "epoch": 9.682752457551386,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.018418626797662072,
      "loss": 0.8195,
      "num_input_tokens_seen": 37737576,
      "step": 65010
    },
    {
      "epoch": 9.683497170092345,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.018416728425780383,
      "loss": 0.7936,
      "num_input_tokens_seen": 37740136,
      "step": 65015
    },
    {
      "epoch": 9.684241882633303,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.018414829996176545,
      "loss": 0.8048,
      "num_input_tokens_seen": 37743048,
      "step": 65020
    },
    {
      "epoch": 9.684986595174262,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01841293150888262,
      "loss": 0.8098,
      "num_input_tokens_seen": 37745864,
      "step": 65025
    },
    {
      "epoch": 9.685731307715223,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.01841103296393069,
      "loss": 0.7996,
      "num_input_tokens_seen": 37748712,
      "step": 65030
    },
    {
      "epoch": 9.686476020256181,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01840913436135282,
      "loss": 0.7742,
      "num_input_tokens_seen": 37751720,
      "step": 65035
    },
    {
      "epoch": 9.68722073279714,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.018407235701181093,
      "loss": 0.7874,
      "num_input_tokens_seen": 37754472,
      "step": 65040
    },
    {
      "epoch": 9.687965445338099,
      "grad_norm": 0.05078125,
      "learning_rate": 0.018405336983447585,
      "loss": 0.8207,
      "num_input_tokens_seen": 37757288,
      "step": 65045
    },
    {
      "epoch": 9.688710157879058,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.018403438208184363,
      "loss": 0.7962,
      "num_input_tokens_seen": 37760456,
      "step": 65050
    },
    {
      "epoch": 9.689454870420018,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.018401539375423517,
      "loss": 0.8305,
      "num_input_tokens_seen": 37763208,
      "step": 65055
    },
    {
      "epoch": 9.690199582960977,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01839964048519712,
      "loss": 0.804,
      "num_input_tokens_seen": 37765896,
      "step": 65060
    },
    {
      "epoch": 9.690944295501936,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01839774153753725,
      "loss": 0.8139,
      "num_input_tokens_seen": 37768840,
      "step": 65065
    },
    {
      "epoch": 9.691689008042896,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.01839584253247599,
      "loss": 0.7828,
      "num_input_tokens_seen": 37771912,
      "step": 65070
    },
    {
      "epoch": 9.692433720583855,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.01839394347004542,
      "loss": 0.8174,
      "num_input_tokens_seen": 37774696,
      "step": 65075
    },
    {
      "epoch": 9.693178433124814,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.018392044350277626,
      "loss": 0.8145,
      "num_input_tokens_seen": 37778088,
      "step": 65080
    },
    {
      "epoch": 9.693923145665773,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.018390145173204694,
      "loss": 0.7971,
      "num_input_tokens_seen": 37780872,
      "step": 65085
    },
    {
      "epoch": 9.694667858206731,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.018388245938858697,
      "loss": 0.7806,
      "num_input_tokens_seen": 37784008,
      "step": 65090
    },
    {
      "epoch": 9.695412570747692,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.018386346647271738,
      "loss": 0.7765,
      "num_input_tokens_seen": 37787048,
      "step": 65095
    },
    {
      "epoch": 9.69615728328865,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.018384447298475888,
      "loss": 0.8146,
      "num_input_tokens_seen": 37790024,
      "step": 65100
    },
    {
      "epoch": 9.69690199582961,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.018382547892503244,
      "loss": 0.7751,
      "num_input_tokens_seen": 37793224,
      "step": 65105
    },
    {
      "epoch": 9.697646708370568,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.018380648429385884,
      "loss": 0.7869,
      "num_input_tokens_seen": 37796168,
      "step": 65110
    },
    {
      "epoch": 9.698391420911529,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.01837874890915591,
      "loss": 0.8051,
      "num_input_tokens_seen": 37799080,
      "step": 65115
    },
    {
      "epoch": 9.699136133452487,
      "grad_norm": 0.029296875,
      "learning_rate": 0.01837684933184541,
      "loss": 0.8034,
      "num_input_tokens_seen": 37801704,
      "step": 65120
    },
    {
      "epoch": 9.699880845993446,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.01837494969748647,
      "loss": 0.8047,
      "num_input_tokens_seen": 37804968,
      "step": 65125
    },
    {
      "epoch": 9.700625558534405,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.018373050006111186,
      "loss": 0.7949,
      "num_input_tokens_seen": 37807880,
      "step": 65130
    },
    {
      "epoch": 9.701370271075366,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.01837115025775165,
      "loss": 0.8214,
      "num_input_tokens_seen": 37811240,
      "step": 65135
    },
    {
      "epoch": 9.702114983616324,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.018369250452439954,
      "loss": 0.8006,
      "num_input_tokens_seen": 37814120,
      "step": 65140
    },
    {
      "epoch": 9.702859696157283,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.0183673505902082,
      "loss": 0.7939,
      "num_input_tokens_seen": 37816968,
      "step": 65145
    },
    {
      "epoch": 9.703604408698242,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.018365450671088475,
      "loss": 0.8243,
      "num_input_tokens_seen": 37819528,
      "step": 65150
    },
    {
      "epoch": 9.704349121239202,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.018363550695112887,
      "loss": 0.8226,
      "num_input_tokens_seen": 37822056,
      "step": 65155
    },
    {
      "epoch": 9.705093833780161,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.018361650662313527,
      "loss": 0.7958,
      "num_input_tokens_seen": 37825128,
      "step": 65160
    },
    {
      "epoch": 9.70583854632112,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.018359750572722495,
      "loss": 0.788,
      "num_input_tokens_seen": 37828200,
      "step": 65165
    },
    {
      "epoch": 9.706583258862079,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.01835785042637189,
      "loss": 0.7601,
      "num_input_tokens_seen": 37830728,
      "step": 65170
    },
    {
      "epoch": 9.70732797140304,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01835595022329382,
      "loss": 0.7975,
      "num_input_tokens_seen": 37833864,
      "step": 65175
    },
    {
      "epoch": 9.708072683943998,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.018354049963520375,
      "loss": 0.7979,
      "num_input_tokens_seen": 37836776,
      "step": 65180
    },
    {
      "epoch": 9.708817396484957,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.01835214964708367,
      "loss": 0.792,
      "num_input_tokens_seen": 37839560,
      "step": 65185
    },
    {
      "epoch": 9.709562109025915,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.0183502492740158,
      "loss": 0.8065,
      "num_input_tokens_seen": 37842600,
      "step": 65190
    },
    {
      "epoch": 9.710306821566876,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.018348348844348876,
      "loss": 0.8009,
      "num_input_tokens_seen": 37845288,
      "step": 65195
    },
    {
      "epoch": 9.711051534107835,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.018346448358114997,
      "loss": 0.7703,
      "num_input_tokens_seen": 37848232,
      "step": 65200
    },
    {
      "epoch": 9.711796246648793,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.018344547815346283,
      "loss": 0.8104,
      "num_input_tokens_seen": 37851048,
      "step": 65205
    },
    {
      "epoch": 9.712540959189752,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.018342647216074825,
      "loss": 0.7992,
      "num_input_tokens_seen": 37853928,
      "step": 65210
    },
    {
      "epoch": 9.713285671730713,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.01834074656033274,
      "loss": 0.7964,
      "num_input_tokens_seen": 37856840,
      "step": 65215
    },
    {
      "epoch": 9.714030384271672,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01833884584815214,
      "loss": 0.797,
      "num_input_tokens_seen": 37859752,
      "step": 65220
    },
    {
      "epoch": 9.71477509681263,
      "grad_norm": 0.03125,
      "learning_rate": 0.018336945079565136,
      "loss": 0.8137,
      "num_input_tokens_seen": 37862440,
      "step": 65225
    },
    {
      "epoch": 9.715519809353589,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.018335044254603833,
      "loss": 0.7908,
      "num_input_tokens_seen": 37865384,
      "step": 65230
    },
    {
      "epoch": 9.716264521894548,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.018333143373300345,
      "loss": 0.7903,
      "num_input_tokens_seen": 37868168,
      "step": 65235
    },
    {
      "epoch": 9.717009234435508,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.018331242435686786,
      "loss": 0.786,
      "num_input_tokens_seen": 37871112,
      "step": 65240
    },
    {
      "epoch": 9.717753946976467,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.01832934144179528,
      "loss": 0.7904,
      "num_input_tokens_seen": 37874024,
      "step": 65245
    },
    {
      "epoch": 9.718498659517426,
      "grad_norm": 0.03125,
      "learning_rate": 0.018327440391657925,
      "loss": 0.7889,
      "num_input_tokens_seen": 37877096,
      "step": 65250
    },
    {
      "epoch": 9.719243372058386,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.01832553928530685,
      "loss": 0.8122,
      "num_input_tokens_seen": 37880104,
      "step": 65255
    },
    {
      "epoch": 9.719988084599345,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.018323638122774176,
      "loss": 0.8051,
      "num_input_tokens_seen": 37883144,
      "step": 65260
    },
    {
      "epoch": 9.720732797140304,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.018321736904092004,
      "loss": 0.8081,
      "num_input_tokens_seen": 37886056,
      "step": 65265
    },
    {
      "epoch": 9.721477509681263,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.018319835629292465,
      "loss": 0.8179,
      "num_input_tokens_seen": 37889096,
      "step": 65270
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.018317934298407678,
      "loss": 0.8208,
      "num_input_tokens_seen": 37891912,
      "step": 65275
    },
    {
      "epoch": 9.722966934763182,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.01831603291146976,
      "loss": 0.7987,
      "num_input_tokens_seen": 37894504,
      "step": 65280
    },
    {
      "epoch": 9.72371164730414,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.018314131468510842,
      "loss": 0.8151,
      "num_input_tokens_seen": 37897416,
      "step": 65285
    },
    {
      "epoch": 9.7244563598451,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01831222996956304,
      "loss": 0.8375,
      "num_input_tokens_seen": 37900552,
      "step": 65290
    },
    {
      "epoch": 9.725201072386058,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.01831032841465848,
      "loss": 0.8091,
      "num_input_tokens_seen": 37903624,
      "step": 65295
    },
    {
      "epoch": 9.725945784927019,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.018308426803829284,
      "loss": 0.7973,
      "num_input_tokens_seen": 37906664,
      "step": 65300
    },
    {
      "epoch": 9.726690497467978,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.018306525137107576,
      "loss": 0.8199,
      "num_input_tokens_seen": 37909384,
      "step": 65305
    },
    {
      "epoch": 9.727435210008936,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.01830462341452549,
      "loss": 0.7987,
      "num_input_tokens_seen": 37912616,
      "step": 65310
    },
    {
      "epoch": 9.728179922549895,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.01830272163611515,
      "loss": 0.8036,
      "num_input_tokens_seen": 37915432,
      "step": 65315
    },
    {
      "epoch": 9.728924635090856,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.018300819801908685,
      "loss": 0.797,
      "num_input_tokens_seen": 37918312,
      "step": 65320
    },
    {
      "epoch": 9.729669347631814,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.018298917911938224,
      "loss": 0.7982,
      "num_input_tokens_seen": 37921224,
      "step": 65325
    },
    {
      "epoch": 9.730414060172773,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0182970159662359,
      "loss": 0.8168,
      "num_input_tokens_seen": 37923944,
      "step": 65330
    },
    {
      "epoch": 9.731158772713732,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01829511396483384,
      "loss": 0.7814,
      "num_input_tokens_seen": 37926824,
      "step": 65335
    },
    {
      "epoch": 9.731903485254692,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.018293211907764177,
      "loss": 0.8214,
      "num_input_tokens_seen": 37929640,
      "step": 65340
    },
    {
      "epoch": 9.732648197795651,
      "grad_norm": 0.0179443359375,
      "learning_rate": 0.018291309795059045,
      "loss": 0.7958,
      "num_input_tokens_seen": 37932328,
      "step": 65345
    },
    {
      "epoch": 9.73339291033661,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01828940762675058,
      "loss": 0.8088,
      "num_input_tokens_seen": 37935432,
      "step": 65350
    },
    {
      "epoch": 9.734137622877569,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.018287505402870916,
      "loss": 0.8032,
      "num_input_tokens_seen": 37938312,
      "step": 65355
    },
    {
      "epoch": 9.73488233541853,
      "grad_norm": 0.0162353515625,
      "learning_rate": 0.018285603123452197,
      "loss": 0.7959,
      "num_input_tokens_seen": 37940904,
      "step": 65360
    },
    {
      "epoch": 9.735627047959488,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01828370078852655,
      "loss": 0.8068,
      "num_input_tokens_seen": 37943560,
      "step": 65365
    },
    {
      "epoch": 9.736371760500447,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.01828179839812611,
      "loss": 0.7885,
      "num_input_tokens_seen": 37946184,
      "step": 65370
    },
    {
      "epoch": 9.737116473041405,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.01827989595228302,
      "loss": 0.794,
      "num_input_tokens_seen": 37949032,
      "step": 65375
    },
    {
      "epoch": 9.737861185582366,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.01827799345102943,
      "loss": 0.8003,
      "num_input_tokens_seen": 37952072,
      "step": 65380
    },
    {
      "epoch": 9.738605898123325,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.018276090894397464,
      "loss": 0.8041,
      "num_input_tokens_seen": 37954920,
      "step": 65385
    },
    {
      "epoch": 9.739350610664284,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.01827418828241928,
      "loss": 0.7889,
      "num_input_tokens_seen": 37957768,
      "step": 65390
    },
    {
      "epoch": 9.740095323205242,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.018272285615127008,
      "loss": 0.7989,
      "num_input_tokens_seen": 37960360,
      "step": 65395
    },
    {
      "epoch": 9.740840035746203,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.018270382892552797,
      "loss": 0.8106,
      "num_input_tokens_seen": 37962984,
      "step": 65400
    },
    {
      "epoch": 9.741584748287162,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.01826848011472879,
      "loss": 0.7936,
      "num_input_tokens_seen": 37965672,
      "step": 65405
    },
    {
      "epoch": 9.74232946082812,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.01826657728168713,
      "loss": 0.8057,
      "num_input_tokens_seen": 37968584,
      "step": 65410
    },
    {
      "epoch": 9.743074173369079,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01826467439345997,
      "loss": 0.7904,
      "num_input_tokens_seen": 37971464,
      "step": 65415
    },
    {
      "epoch": 9.743818885910038,
      "grad_norm": 0.025390625,
      "learning_rate": 0.018262771450079457,
      "loss": 0.7859,
      "num_input_tokens_seen": 37974632,
      "step": 65420
    },
    {
      "epoch": 9.744563598450998,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.018260868451577737,
      "loss": 0.7993,
      "num_input_tokens_seen": 37977512,
      "step": 65425
    },
    {
      "epoch": 9.745308310991957,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.018258965397986954,
      "loss": 0.8152,
      "num_input_tokens_seen": 37980232,
      "step": 65430
    },
    {
      "epoch": 9.746053023532916,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.018257062289339265,
      "loss": 0.8003,
      "num_input_tokens_seen": 37983208,
      "step": 65435
    },
    {
      "epoch": 9.746797736073875,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.01825515912566682,
      "loss": 0.8076,
      "num_input_tokens_seen": 37986056,
      "step": 65440
    },
    {
      "epoch": 9.747542448614835,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.018253255907001772,
      "loss": 0.796,
      "num_input_tokens_seen": 37989256,
      "step": 65445
    },
    {
      "epoch": 9.748287161155794,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.018251352633376267,
      "loss": 0.7868,
      "num_input_tokens_seen": 37992168,
      "step": 65450
    },
    {
      "epoch": 9.749031873696753,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.018249449304822467,
      "loss": 0.7812,
      "num_input_tokens_seen": 37995272,
      "step": 65455
    },
    {
      "epoch": 9.749776586237711,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.018247545921372524,
      "loss": 0.7998,
      "num_input_tokens_seen": 37998184,
      "step": 65460
    },
    {
      "epoch": 9.750521298778672,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.01824564248305859,
      "loss": 0.8118,
      "num_input_tokens_seen": 38001192,
      "step": 65465
    },
    {
      "epoch": 9.75126601131963,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.018243738989912824,
      "loss": 0.7881,
      "num_input_tokens_seen": 38003912,
      "step": 65470
    },
    {
      "epoch": 9.75201072386059,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.018241835441967388,
      "loss": 0.7857,
      "num_input_tokens_seen": 38006920,
      "step": 65475
    },
    {
      "epoch": 9.752755436401548,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.018239931839254434,
      "loss": 0.808,
      "num_input_tokens_seen": 38009576,
      "step": 65480
    },
    {
      "epoch": 9.753500148942509,
      "grad_norm": 0.025390625,
      "learning_rate": 0.018238028181806133,
      "loss": 0.7996,
      "num_input_tokens_seen": 38012360,
      "step": 65485
    },
    {
      "epoch": 9.754244861483468,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.018236124469654627,
      "loss": 0.8204,
      "num_input_tokens_seen": 38015304,
      "step": 65490
    },
    {
      "epoch": 9.754989574024426,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.018234220702832094,
      "loss": 0.8142,
      "num_input_tokens_seen": 38018280,
      "step": 65495
    },
    {
      "epoch": 9.755734286565385,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.018232316881370683,
      "loss": 0.7893,
      "num_input_tokens_seen": 38021096,
      "step": 65500
    },
    {
      "epoch": 9.756478999106346,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01823041300530257,
      "loss": 0.8156,
      "num_input_tokens_seen": 38023976,
      "step": 65505
    },
    {
      "epoch": 9.757223711647304,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.018228509074659904,
      "loss": 0.7957,
      "num_input_tokens_seen": 38026952,
      "step": 65510
    },
    {
      "epoch": 9.757968424188263,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.01822660508947486,
      "loss": 0.7753,
      "num_input_tokens_seen": 38029832,
      "step": 65515
    },
    {
      "epoch": 9.758713136729222,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.018224701049779607,
      "loss": 0.8277,
      "num_input_tokens_seen": 38032840,
      "step": 65520
    },
    {
      "epoch": 9.759457849270182,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.01822279695560631,
      "loss": 0.7975,
      "num_input_tokens_seen": 38036168,
      "step": 65525
    },
    {
      "epoch": 9.760202561811141,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.018220892806987132,
      "loss": 0.8282,
      "num_input_tokens_seen": 38038920,
      "step": 65530
    },
    {
      "epoch": 9.7609472743521,
      "grad_norm": 0.029296875,
      "learning_rate": 0.01821898860395424,
      "loss": 0.7982,
      "num_input_tokens_seen": 38042152,
      "step": 65535
    },
    {
      "epoch": 9.761691986893059,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.01821708434653981,
      "loss": 0.8066,
      "num_input_tokens_seen": 38045320,
      "step": 65540
    },
    {
      "epoch": 9.76243669943402,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.018215180034776006,
      "loss": 0.8011,
      "num_input_tokens_seen": 38048072,
      "step": 65545
    },
    {
      "epoch": 9.763181411974978,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.018213275668695007,
      "loss": 0.8032,
      "num_input_tokens_seen": 38050856,
      "step": 65550
    },
    {
      "epoch": 9.763926124515937,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.01821137124832898,
      "loss": 0.7944,
      "num_input_tokens_seen": 38053608,
      "step": 65555
    },
    {
      "epoch": 9.764670837056896,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.0182094667737101,
      "loss": 0.8014,
      "num_input_tokens_seen": 38056168,
      "step": 65560
    },
    {
      "epoch": 9.765415549597854,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.01820756224487054,
      "loss": 0.8192,
      "num_input_tokens_seen": 38059048,
      "step": 65565
    },
    {
      "epoch": 9.766160262138815,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.018205657661842477,
      "loss": 0.7861,
      "num_input_tokens_seen": 38061928,
      "step": 65570
    },
    {
      "epoch": 9.766904974679774,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.018203753024658084,
      "loss": 0.8002,
      "num_input_tokens_seen": 38065032,
      "step": 65575
    },
    {
      "epoch": 9.767649687220732,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.018201848333349542,
      "loss": 0.8343,
      "num_input_tokens_seen": 38067944,
      "step": 65580
    },
    {
      "epoch": 9.768394399761693,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.018199943587949022,
      "loss": 0.8063,
      "num_input_tokens_seen": 38070536,
      "step": 65585
    },
    {
      "epoch": 9.769139112302652,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.018198038788488715,
      "loss": 0.804,
      "num_input_tokens_seen": 38073480,
      "step": 65590
    },
    {
      "epoch": 9.76988382484361,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01819613393500079,
      "loss": 0.8076,
      "num_input_tokens_seen": 38076648,
      "step": 65595
    },
    {
      "epoch": 9.77062853738457,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01819422902751743,
      "loss": 0.7995,
      "num_input_tokens_seen": 38080008,
      "step": 65600
    },
    {
      "epoch": 9.771373249925528,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.01819232406607081,
      "loss": 0.8058,
      "num_input_tokens_seen": 38082696,
      "step": 65605
    },
    {
      "epoch": 9.772117962466488,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.01819041905069313,
      "loss": 0.8025,
      "num_input_tokens_seen": 38085608,
      "step": 65610
    },
    {
      "epoch": 9.772862675007447,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01818851398141656,
      "loss": 0.816,
      "num_input_tokens_seen": 38088200,
      "step": 65615
    },
    {
      "epoch": 9.773607387548406,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.018186608858273282,
      "loss": 0.7932,
      "num_input_tokens_seen": 38090760,
      "step": 65620
    },
    {
      "epoch": 9.774352100089365,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.018184703681295496,
      "loss": 0.8024,
      "num_input_tokens_seen": 38093544,
      "step": 65625
    },
    {
      "epoch": 9.775096812630325,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01818279845051537,
      "loss": 0.7969,
      "num_input_tokens_seen": 38096392,
      "step": 65630
    },
    {
      "epoch": 9.775841525171284,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.0181808931659651,
      "loss": 0.8099,
      "num_input_tokens_seen": 38099368,
      "step": 65635
    },
    {
      "epoch": 9.776586237712243,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.018178987827676872,
      "loss": 0.8193,
      "num_input_tokens_seen": 38102792,
      "step": 65640
    },
    {
      "epoch": 9.777330950253202,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.018177082435682876,
      "loss": 0.8002,
      "num_input_tokens_seen": 38105544,
      "step": 65645
    },
    {
      "epoch": 9.778075662794162,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.018175176990015307,
      "loss": 0.8009,
      "num_input_tokens_seen": 38108680,
      "step": 65650
    },
    {
      "epoch": 9.77882037533512,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.01817327149070634,
      "loss": 0.7908,
      "num_input_tokens_seen": 38111656,
      "step": 65655
    },
    {
      "epoch": 9.77956508787608,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01817136593778819,
      "loss": 0.806,
      "num_input_tokens_seen": 38114696,
      "step": 65660
    },
    {
      "epoch": 9.780309800417038,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.018169460331293023,
      "loss": 0.8063,
      "num_input_tokens_seen": 38117640,
      "step": 65665
    },
    {
      "epoch": 9.781054512957999,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.018167554671253052,
      "loss": 0.7932,
      "num_input_tokens_seen": 38120424,
      "step": 65670
    },
    {
      "epoch": 9.781799225498958,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.018165648957700457,
      "loss": 0.8084,
      "num_input_tokens_seen": 38123176,
      "step": 65675
    },
    {
      "epoch": 9.782543938039916,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.018163743190667447,
      "loss": 0.7966,
      "num_input_tokens_seen": 38126056,
      "step": 65680
    },
    {
      "epoch": 9.783288650580875,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.018161837370186208,
      "loss": 0.8029,
      "num_input_tokens_seen": 38129064,
      "step": 65685
    },
    {
      "epoch": 9.784033363121836,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.018159931496288945,
      "loss": 0.7925,
      "num_input_tokens_seen": 38132008,
      "step": 65690
    },
    {
      "epoch": 9.784778075662794,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.01815802556900785,
      "loss": 0.8186,
      "num_input_tokens_seen": 38134760,
      "step": 65695
    },
    {
      "epoch": 9.785522788203753,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.018156119588375118,
      "loss": 0.8046,
      "num_input_tokens_seen": 38137352,
      "step": 65700
    },
    {
      "epoch": 9.786267500744712,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.018154213554422954,
      "loss": 0.8087,
      "num_input_tokens_seen": 38140200,
      "step": 65705
    },
    {
      "epoch": 9.787012213285673,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.01815230746718356,
      "loss": 0.8033,
      "num_input_tokens_seen": 38143144,
      "step": 65710
    },
    {
      "epoch": 9.787756925826631,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.018150401326689134,
      "loss": 0.7904,
      "num_input_tokens_seen": 38146216,
      "step": 65715
    },
    {
      "epoch": 9.78850163836759,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.01814849513297188,
      "loss": 0.806,
      "num_input_tokens_seen": 38149064,
      "step": 65720
    },
    {
      "epoch": 9.789246350908549,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.018146588886064,
      "loss": 0.8056,
      "num_input_tokens_seen": 38151944,
      "step": 65725
    },
    {
      "epoch": 9.78999106344951,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.018144682585997704,
      "loss": 0.8055,
      "num_input_tokens_seen": 38154568,
      "step": 65730
    },
    {
      "epoch": 9.790735775990468,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.018142776232805188,
      "loss": 0.7948,
      "num_input_tokens_seen": 38157320,
      "step": 65735
    },
    {
      "epoch": 9.791480488531427,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01814086982651866,
      "loss": 0.7956,
      "num_input_tokens_seen": 38161096,
      "step": 65740
    },
    {
      "epoch": 9.792225201072386,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.018138963367170326,
      "loss": 0.7975,
      "num_input_tokens_seen": 38163848,
      "step": 65745
    },
    {
      "epoch": 9.792969913613344,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.018137056854792402,
      "loss": 0.7906,
      "num_input_tokens_seen": 38166568,
      "step": 65750
    },
    {
      "epoch": 9.793714626154305,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.018135150289417087,
      "loss": 0.7834,
      "num_input_tokens_seen": 38169224,
      "step": 65755
    },
    {
      "epoch": 9.794459338695264,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.018133243671076597,
      "loss": 0.7757,
      "num_input_tokens_seen": 38172072,
      "step": 65760
    },
    {
      "epoch": 9.795204051236222,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.018131336999803137,
      "loss": 0.7587,
      "num_input_tokens_seen": 38174888,
      "step": 65765
    },
    {
      "epoch": 9.795948763777183,
      "grad_norm": 0.041015625,
      "learning_rate": 0.018129430275628924,
      "loss": 0.8391,
      "num_input_tokens_seen": 38177928,
      "step": 65770
    },
    {
      "epoch": 9.796693476318142,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.018127523498586165,
      "loss": 0.8133,
      "num_input_tokens_seen": 38180904,
      "step": 65775
    },
    {
      "epoch": 9.7974381888591,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.018125616668707076,
      "loss": 0.8136,
      "num_input_tokens_seen": 38183816,
      "step": 65780
    },
    {
      "epoch": 9.79818290140006,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.01812370978602387,
      "loss": 0.8163,
      "num_input_tokens_seen": 38186824,
      "step": 65785
    },
    {
      "epoch": 9.798927613941018,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.01812180285056876,
      "loss": 0.8074,
      "num_input_tokens_seen": 38189992,
      "step": 65790
    },
    {
      "epoch": 9.799672326481979,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.018119895862373968,
      "loss": 0.7978,
      "num_input_tokens_seen": 38193064,
      "step": 65795
    },
    {
      "epoch": 9.800417039022937,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0181179888214717,
      "loss": 0.8153,
      "num_input_tokens_seen": 38196168,
      "step": 65800
    },
    {
      "epoch": 9.801161751563896,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.018116081727894185,
      "loss": 0.7985,
      "num_input_tokens_seen": 38198952,
      "step": 65805
    },
    {
      "epoch": 9.801906464104855,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.018114174581673637,
      "loss": 0.8117,
      "num_input_tokens_seen": 38201768,
      "step": 65810
    },
    {
      "epoch": 9.802651176645815,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.018112267382842273,
      "loss": 0.7973,
      "num_input_tokens_seen": 38204392,
      "step": 65815
    },
    {
      "epoch": 9.803395889186774,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.018110360131432312,
      "loss": 0.7988,
      "num_input_tokens_seen": 38207208,
      "step": 65820
    },
    {
      "epoch": 9.804140601727733,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.018108452827475984,
      "loss": 0.8039,
      "num_input_tokens_seen": 38210216,
      "step": 65825
    },
    {
      "epoch": 9.804885314268692,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0181065454710055,
      "loss": 0.7969,
      "num_input_tokens_seen": 38213000,
      "step": 65830
    },
    {
      "epoch": 9.805630026809652,
      "grad_norm": 0.041015625,
      "learning_rate": 0.018104638062053088,
      "loss": 0.8156,
      "num_input_tokens_seen": 38216072,
      "step": 65835
    },
    {
      "epoch": 9.80637473935061,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.018102730600650967,
      "loss": 0.802,
      "num_input_tokens_seen": 38219080,
      "step": 65840
    },
    {
      "epoch": 9.80711945189157,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.018100823086831374,
      "loss": 0.8161,
      "num_input_tokens_seen": 38222088,
      "step": 65845
    },
    {
      "epoch": 9.807864164432528,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.018098915520626525,
      "loss": 0.7979,
      "num_input_tokens_seen": 38225544,
      "step": 65850
    },
    {
      "epoch": 9.808608876973489,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.018097007902068644,
      "loss": 0.7993,
      "num_input_tokens_seen": 38228456,
      "step": 65855
    },
    {
      "epoch": 9.809353589514448,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.018095100231189966,
      "loss": 0.8202,
      "num_input_tokens_seen": 38231592,
      "step": 65860
    },
    {
      "epoch": 9.810098302055406,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.018093192508022715,
      "loss": 0.8162,
      "num_input_tokens_seen": 38234280,
      "step": 65865
    },
    {
      "epoch": 9.810843014596365,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.018091284732599116,
      "loss": 0.7978,
      "num_input_tokens_seen": 38237128,
      "step": 65870
    },
    {
      "epoch": 9.811587727137326,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.018089376904951406,
      "loss": 0.7643,
      "num_input_tokens_seen": 38239912,
      "step": 65875
    },
    {
      "epoch": 9.812332439678285,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.018087469025111814,
      "loss": 0.8,
      "num_input_tokens_seen": 38242664,
      "step": 65880
    },
    {
      "epoch": 9.813077152219243,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.01808556109311257,
      "loss": 0.8073,
      "num_input_tokens_seen": 38245736,
      "step": 65885
    },
    {
      "epoch": 9.813821864760202,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.01808365310898591,
      "loss": 0.7934,
      "num_input_tokens_seen": 38248584,
      "step": 65890
    },
    {
      "epoch": 9.814566577301163,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.018081745072764062,
      "loss": 0.7969,
      "num_input_tokens_seen": 38251496,
      "step": 65895
    },
    {
      "epoch": 9.815311289842121,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.018079836984479264,
      "loss": 0.8086,
      "num_input_tokens_seen": 38254280,
      "step": 65900
    },
    {
      "epoch": 9.81605600238308,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01807792884416375,
      "loss": 0.7917,
      "num_input_tokens_seen": 38257128,
      "step": 65905
    },
    {
      "epoch": 9.816800714924039,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.018076020651849753,
      "loss": 0.807,
      "num_input_tokens_seen": 38260104,
      "step": 65910
    },
    {
      "epoch": 9.817545427465,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.018074112407569516,
      "loss": 0.8256,
      "num_input_tokens_seen": 38263336,
      "step": 65915
    },
    {
      "epoch": 9.818290140005958,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.018072204111355277,
      "loss": 0.7859,
      "num_input_tokens_seen": 38266280,
      "step": 65920
    },
    {
      "epoch": 9.819034852546917,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.01807029576323927,
      "loss": 0.8144,
      "num_input_tokens_seen": 38269448,
      "step": 65925
    },
    {
      "epoch": 9.819779565087876,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.01806838736325374,
      "loss": 0.8064,
      "num_input_tokens_seen": 38272200,
      "step": 65930
    },
    {
      "epoch": 9.820524277628834,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.01806647891143092,
      "loss": 0.7905,
      "num_input_tokens_seen": 38275208,
      "step": 65935
    },
    {
      "epoch": 9.821268990169795,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.018064570407803054,
      "loss": 0.8215,
      "num_input_tokens_seen": 38278088,
      "step": 65940
    },
    {
      "epoch": 9.822013702710754,
      "grad_norm": 0.02734375,
      "learning_rate": 0.018062661852402385,
      "loss": 0.8092,
      "num_input_tokens_seen": 38280616,
      "step": 65945
    },
    {
      "epoch": 9.822758415251712,
      "grad_norm": 0.01953125,
      "learning_rate": 0.01806075324526116,
      "loss": 0.787,
      "num_input_tokens_seen": 38283496,
      "step": 65950
    },
    {
      "epoch": 9.823503127792671,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.018058844586411625,
      "loss": 0.8029,
      "num_input_tokens_seen": 38286408,
      "step": 65955
    },
    {
      "epoch": 9.824247840333632,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.01805693587588601,
      "loss": 0.8186,
      "num_input_tokens_seen": 38289352,
      "step": 65960
    },
    {
      "epoch": 9.82499255287459,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.018055027113716576,
      "loss": 0.7983,
      "num_input_tokens_seen": 38292104,
      "step": 65965
    },
    {
      "epoch": 9.82573726541555,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.01805311829993556,
      "loss": 0.7934,
      "num_input_tokens_seen": 38295016,
      "step": 65970
    },
    {
      "epoch": 9.826481977956508,
      "grad_norm": 0.02734375,
      "learning_rate": 0.018051209434575215,
      "loss": 0.8151,
      "num_input_tokens_seen": 38297896,
      "step": 65975
    },
    {
      "epoch": 9.827226690497469,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.018049300517667788,
      "loss": 0.8066,
      "num_input_tokens_seen": 38300808,
      "step": 65980
    },
    {
      "epoch": 9.827971403038427,
      "grad_norm": 0.0234375,
      "learning_rate": 0.018047391549245527,
      "loss": 0.8058,
      "num_input_tokens_seen": 38303496,
      "step": 65985
    },
    {
      "epoch": 9.828716115579386,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.018045482529340687,
      "loss": 0.7946,
      "num_input_tokens_seen": 38306216,
      "step": 65990
    },
    {
      "epoch": 9.829460828120345,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.01804357345798551,
      "loss": 0.8064,
      "num_input_tokens_seen": 38308904,
      "step": 65995
    },
    {
      "epoch": 9.830205540661305,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.018041664335212253,
      "loss": 0.786,
      "num_input_tokens_seen": 38311944,
      "step": 66000
    },
    {
      "epoch": 9.830950253202264,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.018039755161053172,
      "loss": 0.7921,
      "num_input_tokens_seen": 38314664,
      "step": 66005
    },
    {
      "epoch": 9.831694965743223,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.018037845935540513,
      "loss": 0.7996,
      "num_input_tokens_seen": 38317352,
      "step": 66010
    },
    {
      "epoch": 9.832439678284182,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.018035936658706536,
      "loss": 0.7904,
      "num_input_tokens_seen": 38320296,
      "step": 66015
    },
    {
      "epoch": 9.833184390825142,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0180340273305835,
      "loss": 0.8058,
      "num_input_tokens_seen": 38323368,
      "step": 66020
    },
    {
      "epoch": 9.833929103366101,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01803211795120365,
      "loss": 0.7861,
      "num_input_tokens_seen": 38326344,
      "step": 66025
    },
    {
      "epoch": 9.83467381590706,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.01803020852059925,
      "loss": 0.7931,
      "num_input_tokens_seen": 38329352,
      "step": 66030
    },
    {
      "epoch": 9.835418528448018,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.018028299038802556,
      "loss": 0.7909,
      "num_input_tokens_seen": 38332296,
      "step": 66035
    },
    {
      "epoch": 9.836163240988979,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01802638950584583,
      "loss": 0.7866,
      "num_input_tokens_seen": 38335112,
      "step": 66040
    },
    {
      "epoch": 9.836907953529938,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.018024479921761323,
      "loss": 0.7962,
      "num_input_tokens_seen": 38338120,
      "step": 66045
    },
    {
      "epoch": 9.837652666070897,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.018022570286581312,
      "loss": 0.7753,
      "num_input_tokens_seen": 38341160,
      "step": 66050
    },
    {
      "epoch": 9.838397378611855,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.01802066060033804,
      "loss": 0.7703,
      "num_input_tokens_seen": 38344168,
      "step": 66055
    },
    {
      "epoch": 9.839142091152816,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.018018750863063784,
      "loss": 0.8308,
      "num_input_tokens_seen": 38347048,
      "step": 66060
    },
    {
      "epoch": 9.839886803693775,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.018016841074790794,
      "loss": 0.8058,
      "num_input_tokens_seen": 38350056,
      "step": 66065
    },
    {
      "epoch": 9.840631516234733,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.01801493123555134,
      "loss": 0.8281,
      "num_input_tokens_seen": 38352776,
      "step": 66070
    },
    {
      "epoch": 9.841376228775692,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.01801302134537769,
      "loss": 0.8135,
      "num_input_tokens_seen": 38355816,
      "step": 66075
    },
    {
      "epoch": 9.842120941316653,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.018011111404302105,
      "loss": 0.8058,
      "num_input_tokens_seen": 38358920,
      "step": 66080
    },
    {
      "epoch": 9.842865653857611,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.018009201412356853,
      "loss": 0.7862,
      "num_input_tokens_seen": 38361768,
      "step": 66085
    },
    {
      "epoch": 9.84361036639857,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.018007291369574206,
      "loss": 0.8313,
      "num_input_tokens_seen": 38364616,
      "step": 66090
    },
    {
      "epoch": 9.844355078939529,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.018005381275986426,
      "loss": 0.7795,
      "num_input_tokens_seen": 38367528,
      "step": 66095
    },
    {
      "epoch": 9.84509979148049,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.01800347113162578,
      "loss": 0.7993,
      "num_input_tokens_seen": 38370504,
      "step": 66100
    },
    {
      "epoch": 9.845844504021448,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.01800156093652454,
      "loss": 0.7928,
      "num_input_tokens_seen": 38373320,
      "step": 66105
    },
    {
      "epoch": 9.846589216562407,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.017999650690714984,
      "loss": 0.8274,
      "num_input_tokens_seen": 38376200,
      "step": 66110
    },
    {
      "epoch": 9.847333929103366,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.017997740394229376,
      "loss": 0.8211,
      "num_input_tokens_seen": 38379176,
      "step": 66115
    },
    {
      "epoch": 9.848078641644324,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.017995830047099995,
      "loss": 0.7771,
      "num_input_tokens_seen": 38382152,
      "step": 66120
    },
    {
      "epoch": 9.848823354185285,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.017993919649359105,
      "loss": 0.8343,
      "num_input_tokens_seen": 38385192,
      "step": 66125
    },
    {
      "epoch": 9.849568066726244,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.017992009201038987,
      "loss": 0.8,
      "num_input_tokens_seen": 38388328,
      "step": 66130
    },
    {
      "epoch": 9.850312779267203,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.017990098702171917,
      "loss": 0.7901,
      "num_input_tokens_seen": 38391048,
      "step": 66135
    },
    {
      "epoch": 9.851057491808161,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.017988188152790165,
      "loss": 0.7928,
      "num_input_tokens_seen": 38393832,
      "step": 66140
    },
    {
      "epoch": 9.851802204349122,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.017986277552926012,
      "loss": 0.7958,
      "num_input_tokens_seen": 38396840,
      "step": 66145
    },
    {
      "epoch": 9.85254691689008,
      "grad_norm": 0.029296875,
      "learning_rate": 0.017984366902611737,
      "loss": 0.7862,
      "num_input_tokens_seen": 38399784,
      "step": 66150
    },
    {
      "epoch": 9.85329162943104,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.017982456201879618,
      "loss": 0.8137,
      "num_input_tokens_seen": 38403016,
      "step": 66155
    },
    {
      "epoch": 9.854036341971998,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.017980545450761925,
      "loss": 0.8127,
      "num_input_tokens_seen": 38406024,
      "step": 66160
    },
    {
      "epoch": 9.854781054512959,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.01797863464929095,
      "loss": 0.7859,
      "num_input_tokens_seen": 38409160,
      "step": 66165
    },
    {
      "epoch": 9.855525767053917,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.01797672379749897,
      "loss": 0.8023,
      "num_input_tokens_seen": 38412136,
      "step": 66170
    },
    {
      "epoch": 9.856270479594876,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01797481289541827,
      "loss": 0.8127,
      "num_input_tokens_seen": 38414888,
      "step": 66175
    },
    {
      "epoch": 9.857015192135835,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.017972901943081123,
      "loss": 0.8074,
      "num_input_tokens_seen": 38417640,
      "step": 66180
    },
    {
      "epoch": 9.857759904676795,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.017970990940519824,
      "loss": 0.8032,
      "num_input_tokens_seen": 38420808,
      "step": 66185
    },
    {
      "epoch": 9.858504617217754,
      "grad_norm": 0.02734375,
      "learning_rate": 0.017969079887766654,
      "loss": 0.7929,
      "num_input_tokens_seen": 38423848,
      "step": 66190
    },
    {
      "epoch": 9.859249329758713,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.017967168784853892,
      "loss": 0.7871,
      "num_input_tokens_seen": 38426568,
      "step": 66195
    },
    {
      "epoch": 9.859994042299672,
      "grad_norm": 0.033203125,
      "learning_rate": 0.017965257631813832,
      "loss": 0.8124,
      "num_input_tokens_seen": 38429736,
      "step": 66200
    },
    {
      "epoch": 9.860738754840632,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.017963346428678757,
      "loss": 0.7992,
      "num_input_tokens_seen": 38432680,
      "step": 66205
    },
    {
      "epoch": 9.861483467381591,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.017961435175480957,
      "loss": 0.7913,
      "num_input_tokens_seen": 38435400,
      "step": 66210
    },
    {
      "epoch": 9.86222817992255,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.01795952387225272,
      "loss": 0.8166,
      "num_input_tokens_seen": 38438344,
      "step": 66215
    },
    {
      "epoch": 9.862972892463509,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.017957612519026334,
      "loss": 0.7916,
      "num_input_tokens_seen": 38441288,
      "step": 66220
    },
    {
      "epoch": 9.863717605004469,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.017955701115834093,
      "loss": 0.786,
      "num_input_tokens_seen": 38444264,
      "step": 66225
    },
    {
      "epoch": 9.864462317545428,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.017953789662708282,
      "loss": 0.7997,
      "num_input_tokens_seen": 38447080,
      "step": 66230
    },
    {
      "epoch": 9.865207030086387,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.017951878159681196,
      "loss": 0.771,
      "num_input_tokens_seen": 38449896,
      "step": 66235
    },
    {
      "epoch": 9.865951742627345,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.01794996660678513,
      "loss": 0.8399,
      "num_input_tokens_seen": 38452712,
      "step": 66240
    },
    {
      "epoch": 9.866696455168306,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.017948055004052377,
      "loss": 0.8253,
      "num_input_tokens_seen": 38455432,
      "step": 66245
    },
    {
      "epoch": 9.867441167709265,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.01794614335151523,
      "loss": 0.7821,
      "num_input_tokens_seen": 38458504,
      "step": 66250
    },
    {
      "epoch": 9.868185880250223,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.017944231649205993,
      "loss": 0.8081,
      "num_input_tokens_seen": 38461416,
      "step": 66255
    },
    {
      "epoch": 9.868930592791182,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.017942319897156944,
      "loss": 0.8104,
      "num_input_tokens_seen": 38464296,
      "step": 66260
    },
    {
      "epoch": 9.86967530533214,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.017940408095400397,
      "loss": 0.8021,
      "num_input_tokens_seen": 38466952,
      "step": 66265
    },
    {
      "epoch": 9.870420017873101,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.01793849624396864,
      "loss": 0.8264,
      "num_input_tokens_seen": 38469960,
      "step": 66270
    },
    {
      "epoch": 9.87116473041406,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.01793658434289398,
      "loss": 0.7966,
      "num_input_tokens_seen": 38472872,
      "step": 66275
    },
    {
      "epoch": 9.871909442955019,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.01793467239220871,
      "loss": 0.7765,
      "num_input_tokens_seen": 38475752,
      "step": 66280
    },
    {
      "epoch": 9.87265415549598,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.017932760391945128,
      "loss": 0.7962,
      "num_input_tokens_seen": 38478472,
      "step": 66285
    },
    {
      "epoch": 9.873398868036938,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.017930848342135545,
      "loss": 0.8164,
      "num_input_tokens_seen": 38481128,
      "step": 66290
    },
    {
      "epoch": 9.874143580577897,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.01792893624281226,
      "loss": 0.7758,
      "num_input_tokens_seen": 38484104,
      "step": 66295
    },
    {
      "epoch": 9.874888293118856,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.01792702409400757,
      "loss": 0.8176,
      "num_input_tokens_seen": 38487016,
      "step": 66300
    },
    {
      "epoch": 9.875633005659815,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.017925111895753784,
      "loss": 0.8141,
      "num_input_tokens_seen": 38489832,
      "step": 66305
    },
    {
      "epoch": 9.876377718200775,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.017923199648083202,
      "loss": 0.816,
      "num_input_tokens_seen": 38492968,
      "step": 66310
    },
    {
      "epoch": 9.877122430741734,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.017921287351028135,
      "loss": 0.8079,
      "num_input_tokens_seen": 38495784,
      "step": 66315
    },
    {
      "epoch": 9.877867143282693,
      "grad_norm": 0.0234375,
      "learning_rate": 0.017919375004620892,
      "loss": 0.7843,
      "num_input_tokens_seen": 38498760,
      "step": 66320
    },
    {
      "epoch": 9.878611855823651,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.01791746260889377,
      "loss": 0.8128,
      "num_input_tokens_seen": 38501832,
      "step": 66325
    },
    {
      "epoch": 9.879356568364612,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.017915550163879086,
      "loss": 0.8196,
      "num_input_tokens_seen": 38504648,
      "step": 66330
    },
    {
      "epoch": 9.88010128090557,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.017913637669609146,
      "loss": 0.7936,
      "num_input_tokens_seen": 38507592,
      "step": 66335
    },
    {
      "epoch": 9.88084599344653,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.017911725126116253,
      "loss": 0.7789,
      "num_input_tokens_seen": 38510664,
      "step": 66340
    },
    {
      "epoch": 9.881590705987488,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.017909812533432726,
      "loss": 0.788,
      "num_input_tokens_seen": 38513384,
      "step": 66345
    },
    {
      "epoch": 9.882335418528449,
      "grad_norm": 0.021484375,
      "learning_rate": 0.017907899891590872,
      "loss": 0.8196,
      "num_input_tokens_seen": 38516520,
      "step": 66350
    },
    {
      "epoch": 9.883080131069407,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.01790598720062301,
      "loss": 0.8032,
      "num_input_tokens_seen": 38519592,
      "step": 66355
    },
    {
      "epoch": 9.883824843610366,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.017904074460561446,
      "loss": 0.8036,
      "num_input_tokens_seen": 38522408,
      "step": 66360
    },
    {
      "epoch": 9.884569556151325,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01790216167143849,
      "loss": 0.7757,
      "num_input_tokens_seen": 38525512,
      "step": 66365
    },
    {
      "epoch": 9.885314268692285,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.017900248833286468,
      "loss": 0.8095,
      "num_input_tokens_seen": 38528520,
      "step": 66370
    },
    {
      "epoch": 9.886058981233244,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01789833594613769,
      "loss": 0.7896,
      "num_input_tokens_seen": 38531496,
      "step": 66375
    },
    {
      "epoch": 9.886803693774203,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.01789642301002447,
      "loss": 0.7891,
      "num_input_tokens_seen": 38534184,
      "step": 66380
    },
    {
      "epoch": 9.887548406315162,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.01789451002497913,
      "loss": 0.7816,
      "num_input_tokens_seen": 38537256,
      "step": 66385
    },
    {
      "epoch": 9.888293118856122,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.01789259699103398,
      "loss": 0.8374,
      "num_input_tokens_seen": 38540136,
      "step": 66390
    },
    {
      "epoch": 9.889037831397081,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.017890683908221346,
      "loss": 0.7903,
      "num_input_tokens_seen": 38542920,
      "step": 66395
    },
    {
      "epoch": 9.88978254393804,
      "grad_norm": 0.029296875,
      "learning_rate": 0.017888770776573543,
      "loss": 0.7971,
      "num_input_tokens_seen": 38545768,
      "step": 66400
    },
    {
      "epoch": 9.890527256478999,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.017886857596122897,
      "loss": 0.8174,
      "num_input_tokens_seen": 38548552,
      "step": 66405
    },
    {
      "epoch": 9.891271969019959,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.017884944366901724,
      "loss": 0.8115,
      "num_input_tokens_seen": 38551208,
      "step": 66410
    },
    {
      "epoch": 9.892016681560918,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.01788303108894235,
      "loss": 0.7948,
      "num_input_tokens_seen": 38553896,
      "step": 66415
    },
    {
      "epoch": 9.892761394101877,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.017881117762277098,
      "loss": 0.7943,
      "num_input_tokens_seen": 38557000,
      "step": 66420
    },
    {
      "epoch": 9.893506106642835,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.017879204386938286,
      "loss": 0.7986,
      "num_input_tokens_seen": 38560008,
      "step": 66425
    },
    {
      "epoch": 9.894250819183796,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01787729096295824,
      "loss": 0.7955,
      "num_input_tokens_seen": 38562920,
      "step": 66430
    },
    {
      "epoch": 9.894995531724755,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.01787537749036929,
      "loss": 0.7797,
      "num_input_tokens_seen": 38565864,
      "step": 66435
    },
    {
      "epoch": 9.895740244265713,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.017873463969203758,
      "loss": 0.8157,
      "num_input_tokens_seen": 38568904,
      "step": 66440
    },
    {
      "epoch": 9.896484956806672,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.017871550399493976,
      "loss": 0.7927,
      "num_input_tokens_seen": 38571784,
      "step": 66445
    },
    {
      "epoch": 9.897229669347631,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.017869636781272267,
      "loss": 0.7836,
      "num_input_tokens_seen": 38574632,
      "step": 66450
    },
    {
      "epoch": 9.897974381888591,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.017867723114570966,
      "loss": 0.7842,
      "num_input_tokens_seen": 38577448,
      "step": 66455
    },
    {
      "epoch": 9.89871909442955,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01786580939942239,
      "loss": 0.8152,
      "num_input_tokens_seen": 38580520,
      "step": 66460
    },
    {
      "epoch": 9.899463806970509,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.01786389563585888,
      "loss": 0.7861,
      "num_input_tokens_seen": 38583528,
      "step": 66465
    },
    {
      "epoch": 9.900208519511468,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.017861981823912767,
      "loss": 0.8041,
      "num_input_tokens_seen": 38586248,
      "step": 66470
    },
    {
      "epoch": 9.900953232052428,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.01786006796361637,
      "loss": 0.7935,
      "num_input_tokens_seen": 38589000,
      "step": 66475
    },
    {
      "epoch": 9.901697944593387,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.01785815405500204,
      "loss": 0.8217,
      "num_input_tokens_seen": 38592232,
      "step": 66480
    },
    {
      "epoch": 9.902442657134346,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.017856240098102105,
      "loss": 0.7837,
      "num_input_tokens_seen": 38595144,
      "step": 66485
    },
    {
      "epoch": 9.903187369675305,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.01785432609294889,
      "loss": 0.8234,
      "num_input_tokens_seen": 38597768,
      "step": 66490
    },
    {
      "epoch": 9.903932082216265,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.01785241203957474,
      "loss": 0.8031,
      "num_input_tokens_seen": 38600680,
      "step": 66495
    },
    {
      "epoch": 9.904676794757224,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.017850497938011984,
      "loss": 0.7847,
      "num_input_tokens_seen": 38603432,
      "step": 66500
    },
    {
      "epoch": 9.905421507298183,
      "grad_norm": 0.037109375,
      "learning_rate": 0.017848583788292967,
      "loss": 0.7989,
      "num_input_tokens_seen": 38606312,
      "step": 66505
    },
    {
      "epoch": 9.906166219839141,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.017846669590450018,
      "loss": 0.8306,
      "num_input_tokens_seen": 38609256,
      "step": 66510
    },
    {
      "epoch": 9.906910932380102,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.017844755344515482,
      "loss": 0.7818,
      "num_input_tokens_seen": 38612360,
      "step": 66515
    },
    {
      "epoch": 9.90765564492106,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.017842841050521696,
      "loss": 0.8242,
      "num_input_tokens_seen": 38615048,
      "step": 66520
    },
    {
      "epoch": 9.90840035746202,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.017840926708500997,
      "loss": 0.7797,
      "num_input_tokens_seen": 38617960,
      "step": 66525
    },
    {
      "epoch": 9.909145070002978,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.017839012318485732,
      "loss": 0.8008,
      "num_input_tokens_seen": 38620776,
      "step": 66530
    },
    {
      "epoch": 9.909889782543939,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01783709788050824,
      "loss": 0.7737,
      "num_input_tokens_seen": 38623720,
      "step": 66535
    },
    {
      "epoch": 9.910634495084897,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.01783518339460086,
      "loss": 0.7993,
      "num_input_tokens_seen": 38626664,
      "step": 66540
    },
    {
      "epoch": 9.911379207625856,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.017833268860795934,
      "loss": 0.815,
      "num_input_tokens_seen": 38629736,
      "step": 66545
    },
    {
      "epoch": 9.912123920166815,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.017831354279125817,
      "loss": 0.8186,
      "num_input_tokens_seen": 38632616,
      "step": 66550
    },
    {
      "epoch": 9.912868632707776,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.017829439649622844,
      "loss": 0.7935,
      "num_input_tokens_seen": 38635176,
      "step": 66555
    },
    {
      "epoch": 9.913613345248734,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.017827524972319368,
      "loss": 0.7847,
      "num_input_tokens_seen": 38637928,
      "step": 66560
    },
    {
      "epoch": 9.914358057789693,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.017825610247247724,
      "loss": 0.8228,
      "num_input_tokens_seen": 38640776,
      "step": 66565
    },
    {
      "epoch": 9.915102770330652,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.017823695474440272,
      "loss": 0.8146,
      "num_input_tokens_seen": 38643720,
      "step": 66570
    },
    {
      "epoch": 9.915847482871612,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.017821780653929354,
      "loss": 0.7879,
      "num_input_tokens_seen": 38646600,
      "step": 66575
    },
    {
      "epoch": 9.916592195412571,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.01781986578574732,
      "loss": 0.789,
      "num_input_tokens_seen": 38649512,
      "step": 66580
    },
    {
      "epoch": 9.91733690795353,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.017817950869926522,
      "loss": 0.7737,
      "num_input_tokens_seen": 38652200,
      "step": 66585
    },
    {
      "epoch": 9.918081620494489,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.017816035906499304,
      "loss": 0.8146,
      "num_input_tokens_seen": 38654984,
      "step": 66590
    },
    {
      "epoch": 9.91882633303545,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.017814120895498022,
      "loss": 0.7809,
      "num_input_tokens_seen": 38658120,
      "step": 66595
    },
    {
      "epoch": 9.919571045576408,
      "grad_norm": 0.017333984375,
      "learning_rate": 0.017812205836955027,
      "loss": 0.7767,
      "num_input_tokens_seen": 38661096,
      "step": 66600
    },
    {
      "epoch": 9.920315758117367,
      "grad_norm": 0.016357421875,
      "learning_rate": 0.01781029073090267,
      "loss": 0.797,
      "num_input_tokens_seen": 38663880,
      "step": 66605
    },
    {
      "epoch": 9.921060470658325,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.01780837557737331,
      "loss": 0.8117,
      "num_input_tokens_seen": 38667208,
      "step": 66610
    },
    {
      "epoch": 9.921805183199286,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.017806460376399307,
      "loss": 0.7853,
      "num_input_tokens_seen": 38669928,
      "step": 66615
    },
    {
      "epoch": 9.922549895740245,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.017804545128013,
      "loss": 0.7433,
      "num_input_tokens_seen": 38672840,
      "step": 66620
    },
    {
      "epoch": 9.923294608281203,
      "grad_norm": 0.02734375,
      "learning_rate": 0.017802629832246755,
      "loss": 0.8031,
      "num_input_tokens_seen": 38675688,
      "step": 66625
    },
    {
      "epoch": 9.924039320822162,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.017800714489132927,
      "loss": 0.8046,
      "num_input_tokens_seen": 38678568,
      "step": 66630
    },
    {
      "epoch": 9.924784033363121,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.017798799098703872,
      "loss": 0.8212,
      "num_input_tokens_seen": 38681288,
      "step": 66635
    },
    {
      "epoch": 9.925528745904082,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.017796883660991955,
      "loss": 0.7763,
      "num_input_tokens_seen": 38683880,
      "step": 66640
    },
    {
      "epoch": 9.92627345844504,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.017794968176029526,
      "loss": 0.802,
      "num_input_tokens_seen": 38686792,
      "step": 66645
    },
    {
      "epoch": 9.927018170985999,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.017793052643848957,
      "loss": 0.775,
      "num_input_tokens_seen": 38689512,
      "step": 66650
    },
    {
      "epoch": 9.927762883526958,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.017791137064482602,
      "loss": 0.7952,
      "num_input_tokens_seen": 38692744,
      "step": 66655
    },
    {
      "epoch": 9.928507596067918,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.01778922143796282,
      "loss": 0.794,
      "num_input_tokens_seen": 38695816,
      "step": 66660
    },
    {
      "epoch": 9.929252308608877,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.017787305764321974,
      "loss": 0.8063,
      "num_input_tokens_seen": 38698792,
      "step": 66665
    },
    {
      "epoch": 9.929997021149836,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.017785390043592433,
      "loss": 0.8169,
      "num_input_tokens_seen": 38701608,
      "step": 66670
    },
    {
      "epoch": 9.930741733690795,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.01778347427580656,
      "loss": 0.7761,
      "num_input_tokens_seen": 38704424,
      "step": 66675
    },
    {
      "epoch": 9.931486446231755,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.017781558460996712,
      "loss": 0.8044,
      "num_input_tokens_seen": 38707400,
      "step": 66680
    },
    {
      "epoch": 9.932231158772714,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.017779642599195268,
      "loss": 0.796,
      "num_input_tokens_seen": 38710248,
      "step": 66685
    },
    {
      "epoch": 9.932975871313673,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.017777726690434585,
      "loss": 0.8027,
      "num_input_tokens_seen": 38713096,
      "step": 66690
    },
    {
      "epoch": 9.933720583854631,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01777581073474703,
      "loss": 0.7894,
      "num_input_tokens_seen": 38716328,
      "step": 66695
    },
    {
      "epoch": 9.934465296395592,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.017773894732164976,
      "loss": 0.8359,
      "num_input_tokens_seen": 38719400,
      "step": 66700
    },
    {
      "epoch": 9.93521000893655,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.01777197868272079,
      "loss": 0.8278,
      "num_input_tokens_seen": 38722344,
      "step": 66705
    },
    {
      "epoch": 9.93595472147751,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.017770062586446842,
      "loss": 0.805,
      "num_input_tokens_seen": 38725192,
      "step": 66710
    },
    {
      "epoch": 9.936699434018468,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0177681464433755,
      "loss": 0.7745,
      "num_input_tokens_seen": 38728072,
      "step": 66715
    },
    {
      "epoch": 9.937444146559429,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01776623025353914,
      "loss": 0.7708,
      "num_input_tokens_seen": 38730856,
      "step": 66720
    },
    {
      "epoch": 9.938188859100388,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.017764314016970126,
      "loss": 0.8187,
      "num_input_tokens_seen": 38733640,
      "step": 66725
    },
    {
      "epoch": 9.938933571641346,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.017762397733700842,
      "loss": 0.7953,
      "num_input_tokens_seen": 38736488,
      "step": 66730
    },
    {
      "epoch": 9.939678284182305,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.017760481403763653,
      "loss": 0.7808,
      "num_input_tokens_seen": 38739816,
      "step": 66735
    },
    {
      "epoch": 9.940422996723266,
      "grad_norm": 0.016845703125,
      "learning_rate": 0.01775856502719094,
      "loss": 0.8147,
      "num_input_tokens_seen": 38742760,
      "step": 66740
    },
    {
      "epoch": 9.941167709264224,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01775664860401507,
      "loss": 0.8016,
      "num_input_tokens_seen": 38745288,
      "step": 66745
    },
    {
      "epoch": 9.941912421805183,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.017754732134268425,
      "loss": 0.7964,
      "num_input_tokens_seen": 38747944,
      "step": 66750
    },
    {
      "epoch": 9.942657134346142,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.01775281561798338,
      "loss": 0.815,
      "num_input_tokens_seen": 38750792,
      "step": 66755
    },
    {
      "epoch": 9.943401846887102,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.017750899055192317,
      "loss": 0.7879,
      "num_input_tokens_seen": 38753704,
      "step": 66760
    },
    {
      "epoch": 9.944146559428061,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.017748982445927604,
      "loss": 0.7999,
      "num_input_tokens_seen": 38756744,
      "step": 66765
    },
    {
      "epoch": 9.94489127196902,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.017747065790221627,
      "loss": 0.7975,
      "num_input_tokens_seen": 38759656,
      "step": 66770
    },
    {
      "epoch": 9.945635984509979,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.01774514908810677,
      "loss": 0.7871,
      "num_input_tokens_seen": 38763720,
      "step": 66775
    },
    {
      "epoch": 9.946380697050937,
      "grad_norm": 0.041015625,
      "learning_rate": 0.017743232339615406,
      "loss": 0.7897,
      "num_input_tokens_seen": 38766376,
      "step": 66780
    },
    {
      "epoch": 9.947125409591898,
      "grad_norm": 0.033203125,
      "learning_rate": 0.017741315544779927,
      "loss": 0.7861,
      "num_input_tokens_seen": 38769512,
      "step": 66785
    },
    {
      "epoch": 9.947870122132857,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.0177393987036327,
      "loss": 0.8112,
      "num_input_tokens_seen": 38772648,
      "step": 66790
    },
    {
      "epoch": 9.948614834673815,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.017737481816206116,
      "loss": 0.7815,
      "num_input_tokens_seen": 38775528,
      "step": 66795
    },
    {
      "epoch": 9.949359547214776,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01773556488253256,
      "loss": 0.8025,
      "num_input_tokens_seen": 38778568,
      "step": 66800
    },
    {
      "epoch": 9.950104259755735,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.017733647902644416,
      "loss": 0.8255,
      "num_input_tokens_seen": 38781384,
      "step": 66805
    },
    {
      "epoch": 9.950848972296694,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.017731730876574068,
      "loss": 0.8141,
      "num_input_tokens_seen": 38784040,
      "step": 66810
    },
    {
      "epoch": 9.951593684837652,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.017729813804353907,
      "loss": 0.7863,
      "num_input_tokens_seen": 38787176,
      "step": 66815
    },
    {
      "epoch": 9.952338397378611,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.017727896686016317,
      "loss": 0.8188,
      "num_input_tokens_seen": 38789992,
      "step": 66820
    },
    {
      "epoch": 9.953083109919572,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01772597952159368,
      "loss": 0.7965,
      "num_input_tokens_seen": 38793320,
      "step": 66825
    },
    {
      "epoch": 9.95382782246053,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.01772406231111839,
      "loss": 0.7951,
      "num_input_tokens_seen": 38796200,
      "step": 66830
    },
    {
      "epoch": 9.954572535001489,
      "grad_norm": 0.03125,
      "learning_rate": 0.01772214505462284,
      "loss": 0.7871,
      "num_input_tokens_seen": 38799272,
      "step": 66835
    },
    {
      "epoch": 9.955317247542448,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.01772022775213941,
      "loss": 0.7959,
      "num_input_tokens_seen": 38802024,
      "step": 66840
    },
    {
      "epoch": 9.956061960083408,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0177183104037005,
      "loss": 0.8193,
      "num_input_tokens_seen": 38804776,
      "step": 66845
    },
    {
      "epoch": 9.956806672624367,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.017716393009338503,
      "loss": 0.7931,
      "num_input_tokens_seen": 38807688,
      "step": 66850
    },
    {
      "epoch": 9.957551385165326,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.017714475569085803,
      "loss": 0.8158,
      "num_input_tokens_seen": 38810280,
      "step": 66855
    },
    {
      "epoch": 9.958296097706285,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.017712558082974797,
      "loss": 0.7972,
      "num_input_tokens_seen": 38813128,
      "step": 66860
    },
    {
      "epoch": 9.959040810247245,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.017710640551037878,
      "loss": 0.7876,
      "num_input_tokens_seen": 38816072,
      "step": 66865
    },
    {
      "epoch": 9.959785522788204,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01770872297330744,
      "loss": 0.7937,
      "num_input_tokens_seen": 38819240,
      "step": 66870
    },
    {
      "epoch": 9.960530235329163,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.017706805349815884,
      "loss": 0.799,
      "num_input_tokens_seen": 38821928,
      "step": 66875
    },
    {
      "epoch": 9.961274947870121,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0177048876805956,
      "loss": 0.8081,
      "num_input_tokens_seen": 38825032,
      "step": 66880
    },
    {
      "epoch": 9.962019660411082,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.01770296996567899,
      "loss": 0.8102,
      "num_input_tokens_seen": 38827656,
      "step": 66885
    },
    {
      "epoch": 9.96276437295204,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.017701052205098454,
      "loss": 0.7997,
      "num_input_tokens_seen": 38830568,
      "step": 66890
    },
    {
      "epoch": 9.963509085493,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.017699134398886377,
      "loss": 0.7763,
      "num_input_tokens_seen": 38833640,
      "step": 66895
    },
    {
      "epoch": 9.964253798033958,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.017697216547075175,
      "loss": 0.789,
      "num_input_tokens_seen": 38836456,
      "step": 66900
    },
    {
      "epoch": 9.964998510574919,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.01769529864969724,
      "loss": 0.8209,
      "num_input_tokens_seen": 38839208,
      "step": 66905
    },
    {
      "epoch": 9.965743223115878,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.017693380706784965,
      "loss": 0.7958,
      "num_input_tokens_seen": 38841992,
      "step": 66910
    },
    {
      "epoch": 9.966487935656836,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.01769146271837077,
      "loss": 0.8077,
      "num_input_tokens_seen": 38844808,
      "step": 66915
    },
    {
      "epoch": 9.967232648197795,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.01768954468448705,
      "loss": 0.7901,
      "num_input_tokens_seen": 38847656,
      "step": 66920
    },
    {
      "epoch": 9.967977360738756,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.017687626605166196,
      "loss": 0.8273,
      "num_input_tokens_seen": 38850696,
      "step": 66925
    },
    {
      "epoch": 9.968722073279714,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.017685708480440623,
      "loss": 0.8062,
      "num_input_tokens_seen": 38854056,
      "step": 66930
    },
    {
      "epoch": 9.969466785820673,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.01768379031034274,
      "loss": 0.8223,
      "num_input_tokens_seen": 38857032,
      "step": 66935
    },
    {
      "epoch": 9.970211498361632,
      "grad_norm": 0.046875,
      "learning_rate": 0.01768187209490495,
      "loss": 0.7963,
      "num_input_tokens_seen": 38860136,
      "step": 66940
    },
    {
      "epoch": 9.970956210902592,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.01767995383415965,
      "loss": 0.8037,
      "num_input_tokens_seen": 38863048,
      "step": 66945
    },
    {
      "epoch": 9.971700923443551,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.01767803552813926,
      "loss": 0.8052,
      "num_input_tokens_seen": 38865992,
      "step": 66950
    },
    {
      "epoch": 9.97244563598451,
      "grad_norm": 0.046875,
      "learning_rate": 0.017676117176876175,
      "loss": 0.8069,
      "num_input_tokens_seen": 38869032,
      "step": 66955
    },
    {
      "epoch": 9.973190348525469,
      "grad_norm": 0.044921875,
      "learning_rate": 0.017674198780402817,
      "loss": 0.8188,
      "num_input_tokens_seen": 38871944,
      "step": 66960
    },
    {
      "epoch": 9.973935061066427,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01767228033875158,
      "loss": 0.8017,
      "num_input_tokens_seen": 38874408,
      "step": 66965
    },
    {
      "epoch": 9.974679773607388,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.017670361851954888,
      "loss": 0.7963,
      "num_input_tokens_seen": 38877256,
      "step": 66970
    },
    {
      "epoch": 9.975424486148347,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.017668443320045146,
      "loss": 0.8008,
      "num_input_tokens_seen": 38879912,
      "step": 66975
    },
    {
      "epoch": 9.976169198689306,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.017666524743054766,
      "loss": 0.8066,
      "num_input_tokens_seen": 38882600,
      "step": 66980
    },
    {
      "epoch": 9.976913911230266,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.017664606121016165,
      "loss": 0.7824,
      "num_input_tokens_seen": 38885672,
      "step": 66985
    },
    {
      "epoch": 9.977658623771225,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01766268745396175,
      "loss": 0.7892,
      "num_input_tokens_seen": 38888808,
      "step": 66990
    },
    {
      "epoch": 9.978403336312184,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.01766076874192393,
      "loss": 0.7833,
      "num_input_tokens_seen": 38891816,
      "step": 66995
    },
    {
      "epoch": 9.979148048853142,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.017658849984935134,
      "loss": 0.7801,
      "num_input_tokens_seen": 38894728,
      "step": 67000
    },
    {
      "epoch": 9.979892761394101,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.017656931183027764,
      "loss": 0.8046,
      "num_input_tokens_seen": 38897864,
      "step": 67005
    },
    {
      "epoch": 9.980637473935062,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.017655012336234243,
      "loss": 0.792,
      "num_input_tokens_seen": 38901096,
      "step": 67010
    },
    {
      "epoch": 9.98138218647602,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.01765309344458699,
      "loss": 0.8106,
      "num_input_tokens_seen": 38904040,
      "step": 67015
    },
    {
      "epoch": 9.98212689901698,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.017651174508118424,
      "loss": 0.7845,
      "num_input_tokens_seen": 38906888,
      "step": 67020
    },
    {
      "epoch": 9.982871611557938,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.01764925552686095,
      "loss": 0.7883,
      "num_input_tokens_seen": 38909864,
      "step": 67025
    },
    {
      "epoch": 9.983616324098898,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.017647336500847002,
      "loss": 0.824,
      "num_input_tokens_seen": 38912616,
      "step": 67030
    },
    {
      "epoch": 9.984361036639857,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.01764541743010899,
      "loss": 0.8025,
      "num_input_tokens_seen": 38915368,
      "step": 67035
    },
    {
      "epoch": 9.985105749180816,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.017643498314679342,
      "loss": 0.8157,
      "num_input_tokens_seen": 38918312,
      "step": 67040
    },
    {
      "epoch": 9.985850461721775,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.017641579154590473,
      "loss": 0.799,
      "num_input_tokens_seen": 38921320,
      "step": 67045
    },
    {
      "epoch": 9.986595174262735,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.017639659949874816,
      "loss": 0.7884,
      "num_input_tokens_seen": 38924360,
      "step": 67050
    },
    {
      "epoch": 9.987339886803694,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.017637740700564782,
      "loss": 0.7906,
      "num_input_tokens_seen": 38927336,
      "step": 67055
    },
    {
      "epoch": 9.988084599344653,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.017635821406692804,
      "loss": 0.7883,
      "num_input_tokens_seen": 38930056,
      "step": 67060
    },
    {
      "epoch": 9.988829311885612,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0176339020682913,
      "loss": 0.7948,
      "num_input_tokens_seen": 38932904,
      "step": 67065
    },
    {
      "epoch": 9.989574024426572,
      "grad_norm": 0.033203125,
      "learning_rate": 0.017631982685392695,
      "loss": 0.8057,
      "num_input_tokens_seen": 38935912,
      "step": 67070
    },
    {
      "epoch": 9.99031873696753,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01763006325802942,
      "loss": 0.8048,
      "num_input_tokens_seen": 38938920,
      "step": 67075
    },
    {
      "epoch": 9.99106344950849,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0176281437862339,
      "loss": 0.7768,
      "num_input_tokens_seen": 38941672,
      "step": 67080
    },
    {
      "epoch": 9.991808162049448,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.01762622427003856,
      "loss": 0.8081,
      "num_input_tokens_seen": 38945064,
      "step": 67085
    },
    {
      "epoch": 9.992552874590409,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.01762430470947583,
      "loss": 0.8227,
      "num_input_tokens_seen": 38947816,
      "step": 67090
    },
    {
      "epoch": 9.993297587131368,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.01762238510457814,
      "loss": 0.8048,
      "num_input_tokens_seen": 38951048,
      "step": 67095
    },
    {
      "epoch": 9.994042299672326,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.017620465455377918,
      "loss": 0.7956,
      "num_input_tokens_seen": 38953864,
      "step": 67100
    },
    {
      "epoch": 9.994787012213285,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0176185457619076,
      "loss": 0.7837,
      "num_input_tokens_seen": 38956616,
      "step": 67105
    },
    {
      "epoch": 9.995531724754246,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.017616626024199604,
      "loss": 0.8213,
      "num_input_tokens_seen": 38959272,
      "step": 67110
    },
    {
      "epoch": 9.996276437295204,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.01761470624228638,
      "loss": 0.8003,
      "num_input_tokens_seen": 38961960,
      "step": 67115
    },
    {
      "epoch": 9.997021149836163,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.017612786416200343,
      "loss": 0.8147,
      "num_input_tokens_seen": 38965192,
      "step": 67120
    },
    {
      "epoch": 9.997765862377122,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.017610866545973942,
      "loss": 0.8092,
      "num_input_tokens_seen": 38968424,
      "step": 67125
    },
    {
      "epoch": 9.998510574918082,
      "grad_norm": 0.04296875,
      "learning_rate": 0.017608946631639602,
      "loss": 0.7926,
      "num_input_tokens_seen": 38971208,
      "step": 67130
    },
    {
      "epoch": 9.999255287459041,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.017607026673229758,
      "loss": 0.7866,
      "num_input_tokens_seen": 38973960,
      "step": 67135
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.01760510667077685,
      "loss": 0.8102,
      "num_input_tokens_seen": 38976336,
      "step": 67140
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.8017409443855286,
      "eval_runtime": 70.6577,
      "eval_samples_per_second": 42.232,
      "eval_steps_per_second": 10.558,
      "num_input_tokens_seen": 38976336,
      "step": 67140
    },
    {
      "epoch": 10.000744712540959,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.017603186624313313,
      "loss": 0.8091,
      "num_input_tokens_seen": 38979152,
      "step": 67145
    },
    {
      "epoch": 10.001489425081918,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.017601266533871587,
      "loss": 0.7918,
      "num_input_tokens_seen": 38982416,
      "step": 67150
    },
    {
      "epoch": 10.002234137622878,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.0175993463994841,
      "loss": 0.7996,
      "num_input_tokens_seen": 38985136,
      "step": 67155
    },
    {
      "epoch": 10.002978850163837,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.0175974262211833,
      "loss": 0.8032,
      "num_input_tokens_seen": 38988048,
      "step": 67160
    },
    {
      "epoch": 10.003723562704796,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.017595505999001622,
      "loss": 0.8192,
      "num_input_tokens_seen": 38991152,
      "step": 67165
    },
    {
      "epoch": 10.004468275245754,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.017593585732971512,
      "loss": 0.8028,
      "num_input_tokens_seen": 38994256,
      "step": 67170
    },
    {
      "epoch": 10.005212987786715,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.017591665423125402,
      "loss": 0.7831,
      "num_input_tokens_seen": 38996976,
      "step": 67175
    },
    {
      "epoch": 10.005957700327674,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.017589745069495743,
      "loss": 0.7978,
      "num_input_tokens_seen": 38999792,
      "step": 67180
    },
    {
      "epoch": 10.006702412868632,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.017587824672114977,
      "loss": 0.7972,
      "num_input_tokens_seen": 39003152,
      "step": 67185
    },
    {
      "epoch": 10.007447125409591,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.01758590423101554,
      "loss": 0.7821,
      "num_input_tokens_seen": 39005840,
      "step": 67190
    },
    {
      "epoch": 10.008191837950552,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.017583983746229874,
      "loss": 0.7998,
      "num_input_tokens_seen": 39008496,
      "step": 67195
    },
    {
      "epoch": 10.00893655049151,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.01758206321779043,
      "loss": 0.7931,
      "num_input_tokens_seen": 39011504,
      "step": 67200
    },
    {
      "epoch": 10.00968126303247,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.017580142645729657,
      "loss": 0.8125,
      "num_input_tokens_seen": 39014608,
      "step": 67205
    },
    {
      "epoch": 10.010425975573428,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.017578222030079995,
      "loss": 0.7984,
      "num_input_tokens_seen": 39017648,
      "step": 67210
    },
    {
      "epoch": 10.011170688114388,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.017576301370873893,
      "loss": 0.8132,
      "num_input_tokens_seen": 39020528,
      "step": 67215
    },
    {
      "epoch": 10.011915400655347,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.017574380668143794,
      "loss": 0.8047,
      "num_input_tokens_seen": 39023568,
      "step": 67220
    },
    {
      "epoch": 10.012660113196306,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.017572459921922155,
      "loss": 0.7994,
      "num_input_tokens_seen": 39026640,
      "step": 67225
    },
    {
      "epoch": 10.013404825737265,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.017570539132241415,
      "loss": 0.7761,
      "num_input_tokens_seen": 39029360,
      "step": 67230
    },
    {
      "epoch": 10.014149538278225,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.01756861829913403,
      "loss": 0.7927,
      "num_input_tokens_seen": 39032528,
      "step": 67235
    },
    {
      "epoch": 10.014894250819184,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.017566697422632453,
      "loss": 0.7885,
      "num_input_tokens_seen": 39035600,
      "step": 67240
    },
    {
      "epoch": 10.015638963360143,
      "grad_norm": 0.068359375,
      "learning_rate": 0.017564776502769133,
      "loss": 0.8346,
      "num_input_tokens_seen": 39038416,
      "step": 67245
    },
    {
      "epoch": 10.016383675901102,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.017562855539576514,
      "loss": 0.772,
      "num_input_tokens_seen": 39041136,
      "step": 67250
    },
    {
      "epoch": 10.017128388442062,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.017560934533087057,
      "loss": 0.8128,
      "num_input_tokens_seen": 39044240,
      "step": 67255
    },
    {
      "epoch": 10.01787310098302,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.017559013483333216,
      "loss": 0.8176,
      "num_input_tokens_seen": 39047216,
      "step": 67260
    },
    {
      "epoch": 10.01861781352398,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.01755709239034744,
      "loss": 0.7866,
      "num_input_tokens_seen": 39050032,
      "step": 67265
    },
    {
      "epoch": 10.019362526064938,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.017555171254162193,
      "loss": 0.7886,
      "num_input_tokens_seen": 39053104,
      "step": 67270
    },
    {
      "epoch": 10.020107238605899,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.017553250074809918,
      "loss": 0.7956,
      "num_input_tokens_seen": 39056080,
      "step": 67275
    },
    {
      "epoch": 10.020851951146858,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.017551328852323085,
      "loss": 0.7957,
      "num_input_tokens_seen": 39058800,
      "step": 67280
    },
    {
      "epoch": 10.021596663687816,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.01754940758673414,
      "loss": 0.7857,
      "num_input_tokens_seen": 39061520,
      "step": 67285
    },
    {
      "epoch": 10.022341376228775,
      "grad_norm": 0.041015625,
      "learning_rate": 0.017547486278075546,
      "loss": 0.7751,
      "num_input_tokens_seen": 39064400,
      "step": 67290
    },
    {
      "epoch": 10.023086088769736,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.017545564926379757,
      "loss": 0.8055,
      "num_input_tokens_seen": 39067120,
      "step": 67295
    },
    {
      "epoch": 10.023830801310694,
      "grad_norm": 0.037109375,
      "learning_rate": 0.017543643531679236,
      "loss": 0.8062,
      "num_input_tokens_seen": 39070160,
      "step": 67300
    },
    {
      "epoch": 10.024575513851653,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.017541722094006448,
      "loss": 0.835,
      "num_input_tokens_seen": 39073200,
      "step": 67305
    },
    {
      "epoch": 10.025320226392612,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.017539800613393847,
      "loss": 0.7742,
      "num_input_tokens_seen": 39075824,
      "step": 67310
    },
    {
      "epoch": 10.02606493893357,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.017537879089873896,
      "loss": 0.7763,
      "num_input_tokens_seen": 39078544,
      "step": 67315
    },
    {
      "epoch": 10.026809651474531,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.017535957523479057,
      "loss": 0.8021,
      "num_input_tokens_seen": 39081328,
      "step": 67320
    },
    {
      "epoch": 10.02755436401549,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.017534035914241786,
      "loss": 0.8028,
      "num_input_tokens_seen": 39084272,
      "step": 67325
    },
    {
      "epoch": 10.028299076556449,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.017532114262194563,
      "loss": 0.7909,
      "num_input_tokens_seen": 39087344,
      "step": 67330
    },
    {
      "epoch": 10.029043789097408,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.01753019256736984,
      "loss": 0.8139,
      "num_input_tokens_seen": 39090128,
      "step": 67335
    },
    {
      "epoch": 10.029788501638368,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.01752827082980008,
      "loss": 0.7954,
      "num_input_tokens_seen": 39093200,
      "step": 67340
    },
    {
      "epoch": 10.030533214179327,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.017526349049517765,
      "loss": 0.796,
      "num_input_tokens_seen": 39096464,
      "step": 67345
    },
    {
      "epoch": 10.031277926720286,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.017524427226555343,
      "loss": 0.7922,
      "num_input_tokens_seen": 39099664,
      "step": 67350
    },
    {
      "epoch": 10.032022639261244,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.017522505360945292,
      "loss": 0.8077,
      "num_input_tokens_seen": 39102736,
      "step": 67355
    },
    {
      "epoch": 10.032767351802205,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.017520583452720073,
      "loss": 0.8031,
      "num_input_tokens_seen": 39105872,
      "step": 67360
    },
    {
      "epoch": 10.033512064343164,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.01751866150191216,
      "loss": 0.7976,
      "num_input_tokens_seen": 39108720,
      "step": 67365
    },
    {
      "epoch": 10.034256776884122,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.017516739508554018,
      "loss": 0.795,
      "num_input_tokens_seen": 39111504,
      "step": 67370
    },
    {
      "epoch": 10.035001489425081,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01751481747267812,
      "loss": 0.806,
      "num_input_tokens_seen": 39114416,
      "step": 67375
    },
    {
      "epoch": 10.035746201966042,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.017512895394316944,
      "loss": 0.7974,
      "num_input_tokens_seen": 39117360,
      "step": 67380
    },
    {
      "epoch": 10.036490914507,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.017510973273502952,
      "loss": 0.8072,
      "num_input_tokens_seen": 39120464,
      "step": 67385
    },
    {
      "epoch": 10.03723562704796,
      "grad_norm": 0.04296875,
      "learning_rate": 0.01750905111026862,
      "loss": 0.8144,
      "num_input_tokens_seen": 39123376,
      "step": 67390
    },
    {
      "epoch": 10.037980339588918,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01750712890464641,
      "loss": 0.8105,
      "num_input_tokens_seen": 39126224,
      "step": 67395
    },
    {
      "epoch": 10.038725052129879,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.017505206656668812,
      "loss": 0.7942,
      "num_input_tokens_seen": 39129072,
      "step": 67400
    },
    {
      "epoch": 10.039469764670837,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.017503284366368294,
      "loss": 0.8111,
      "num_input_tokens_seen": 39131952,
      "step": 67405
    },
    {
      "epoch": 10.040214477211796,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.01750136203377733,
      "loss": 0.8005,
      "num_input_tokens_seen": 39134864,
      "step": 67410
    },
    {
      "epoch": 10.040959189752755,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.017499439658928393,
      "loss": 0.7868,
      "num_input_tokens_seen": 39137552,
      "step": 67415
    },
    {
      "epoch": 10.041703902293715,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.017497517241853968,
      "loss": 0.8199,
      "num_input_tokens_seen": 39140720,
      "step": 67420
    },
    {
      "epoch": 10.042448614834674,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.01749559478258653,
      "loss": 0.7979,
      "num_input_tokens_seen": 39143856,
      "step": 67425
    },
    {
      "epoch": 10.043193327375633,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.01749367228115855,
      "loss": 0.8465,
      "num_input_tokens_seen": 39146544,
      "step": 67430
    },
    {
      "epoch": 10.043938039916592,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.01749174973760251,
      "loss": 0.7872,
      "num_input_tokens_seen": 39149296,
      "step": 67435
    },
    {
      "epoch": 10.044682752457552,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.017489827151950892,
      "loss": 0.7868,
      "num_input_tokens_seen": 39152336,
      "step": 67440
    },
    {
      "epoch": 10.045427464998511,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.01748790452423618,
      "loss": 0.7874,
      "num_input_tokens_seen": 39155440,
      "step": 67445
    },
    {
      "epoch": 10.04617217753947,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.017485981854490844,
      "loss": 0.8106,
      "num_input_tokens_seen": 39158480,
      "step": 67450
    },
    {
      "epoch": 10.046916890080428,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.017484059142747373,
      "loss": 0.8143,
      "num_input_tokens_seen": 39161712,
      "step": 67455
    },
    {
      "epoch": 10.047661602621389,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.01748213638903824,
      "loss": 0.8106,
      "num_input_tokens_seen": 39164624,
      "step": 67460
    },
    {
      "epoch": 10.048406315162348,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.01748021359339595,
      "loss": 0.8057,
      "num_input_tokens_seen": 39167696,
      "step": 67465
    },
    {
      "epoch": 10.049151027703306,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.017478290755852964,
      "loss": 0.8043,
      "num_input_tokens_seen": 39170512,
      "step": 67470
    },
    {
      "epoch": 10.049895740244265,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.017476367876441773,
      "loss": 0.8185,
      "num_input_tokens_seen": 39173648,
      "step": 67475
    },
    {
      "epoch": 10.050640452785226,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.01747444495519487,
      "loss": 0.7919,
      "num_input_tokens_seen": 39176496,
      "step": 67480
    },
    {
      "epoch": 10.051385165326185,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01747252199214473,
      "loss": 0.8102,
      "num_input_tokens_seen": 39179408,
      "step": 67485
    },
    {
      "epoch": 10.052129877867143,
      "grad_norm": 0.05078125,
      "learning_rate": 0.01747059898732384,
      "loss": 0.7994,
      "num_input_tokens_seen": 39182480,
      "step": 67490
    },
    {
      "epoch": 10.052874590408102,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.017468675940764695,
      "loss": 0.8034,
      "num_input_tokens_seen": 39185296,
      "step": 67495
    },
    {
      "epoch": 10.05361930294906,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.017466752852499778,
      "loss": 0.8041,
      "num_input_tokens_seen": 39188080,
      "step": 67500
    },
    {
      "epoch": 10.054364015490021,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.01746482972256158,
      "loss": 0.8064,
      "num_input_tokens_seen": 39191024,
      "step": 67505
    },
    {
      "epoch": 10.05510872803098,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.01746290655098259,
      "loss": 0.8109,
      "num_input_tokens_seen": 39194000,
      "step": 67510
    },
    {
      "epoch": 10.055853440571939,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.017460983337795298,
      "loss": 0.7948,
      "num_input_tokens_seen": 39196784,
      "step": 67515
    },
    {
      "epoch": 10.056598153112898,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01745906008303219,
      "loss": 0.8027,
      "num_input_tokens_seen": 39199632,
      "step": 67520
    },
    {
      "epoch": 10.057342865653858,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.017457136786725762,
      "loss": 0.7965,
      "num_input_tokens_seen": 39202288,
      "step": 67525
    },
    {
      "epoch": 10.058087578194817,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.017455213448908508,
      "loss": 0.8124,
      "num_input_tokens_seen": 39205200,
      "step": 67530
    },
    {
      "epoch": 10.058832290735776,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.017453290069612912,
      "loss": 0.7908,
      "num_input_tokens_seen": 39208624,
      "step": 67535
    },
    {
      "epoch": 10.059577003276734,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.01745136664887148,
      "loss": 0.787,
      "num_input_tokens_seen": 39211312,
      "step": 67540
    },
    {
      "epoch": 10.060321715817695,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.0174494431867167,
      "loss": 0.8061,
      "num_input_tokens_seen": 39214480,
      "step": 67545
    },
    {
      "epoch": 10.061066428358654,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.017447519683181065,
      "loss": 0.788,
      "num_input_tokens_seen": 39217424,
      "step": 67550
    },
    {
      "epoch": 10.061811140899612,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.01744559613829707,
      "loss": 0.8017,
      "num_input_tokens_seen": 39220016,
      "step": 67555
    },
    {
      "epoch": 10.062555853440571,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.01744367255209722,
      "loss": 0.7965,
      "num_input_tokens_seen": 39222832,
      "step": 67560
    },
    {
      "epoch": 10.063300565981532,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.017441748924614,
      "loss": 0.7918,
      "num_input_tokens_seen": 39225808,
      "step": 67565
    },
    {
      "epoch": 10.06404527852249,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01743982525587991,
      "loss": 0.8039,
      "num_input_tokens_seen": 39228944,
      "step": 67570
    },
    {
      "epoch": 10.06478999106345,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.01743790154592746,
      "loss": 0.8089,
      "num_input_tokens_seen": 39231920,
      "step": 67575
    },
    {
      "epoch": 10.065534703604408,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.01743597779478914,
      "loss": 0.8126,
      "num_input_tokens_seen": 39235088,
      "step": 67580
    },
    {
      "epoch": 10.066279416145369,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.01743405400249745,
      "loss": 0.8378,
      "num_input_tokens_seen": 39237840,
      "step": 67585
    },
    {
      "epoch": 10.067024128686327,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.017432130169084892,
      "loss": 0.8201,
      "num_input_tokens_seen": 39240400,
      "step": 67590
    },
    {
      "epoch": 10.067768841227286,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.017430206294583965,
      "loss": 0.8,
      "num_input_tokens_seen": 39243184,
      "step": 67595
    },
    {
      "epoch": 10.068513553768245,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.017428282379027174,
      "loss": 0.7877,
      "num_input_tokens_seen": 39246160,
      "step": 67600
    },
    {
      "epoch": 10.069258266309205,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.017426358422447018,
      "loss": 0.7819,
      "num_input_tokens_seen": 39249008,
      "step": 67605
    },
    {
      "epoch": 10.070002978850164,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.017424434424876005,
      "loss": 0.7858,
      "num_input_tokens_seen": 39251888,
      "step": 67610
    },
    {
      "epoch": 10.070747691391123,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.017422510386346634,
      "loss": 0.7716,
      "num_input_tokens_seen": 39254896,
      "step": 67615
    },
    {
      "epoch": 10.071492403932082,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.017420586306891413,
      "loss": 0.7867,
      "num_input_tokens_seen": 39257520,
      "step": 67620
    },
    {
      "epoch": 10.072237116473042,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.017418662186542852,
      "loss": 0.8233,
      "num_input_tokens_seen": 39260432,
      "step": 67625
    },
    {
      "epoch": 10.072981829014001,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.01741673802533344,
      "loss": 0.7868,
      "num_input_tokens_seen": 39263312,
      "step": 67630
    },
    {
      "epoch": 10.07372654155496,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.017414813823295707,
      "loss": 0.7765,
      "num_input_tokens_seen": 39266000,
      "step": 67635
    },
    {
      "epoch": 10.074471254095918,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01741288958046214,
      "loss": 0.7908,
      "num_input_tokens_seen": 39268976,
      "step": 67640
    },
    {
      "epoch": 10.075215966636879,
      "grad_norm": 0.044921875,
      "learning_rate": 0.017410965296865265,
      "loss": 0.8055,
      "num_input_tokens_seen": 39271792,
      "step": 67645
    },
    {
      "epoch": 10.075960679177838,
      "grad_norm": 0.0703125,
      "learning_rate": 0.017409040972537572,
      "loss": 0.7997,
      "num_input_tokens_seen": 39274704,
      "step": 67650
    },
    {
      "epoch": 10.076705391718797,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.017407116607511586,
      "loss": 0.8012,
      "num_input_tokens_seen": 39277680,
      "step": 67655
    },
    {
      "epoch": 10.077450104259755,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.017405192201819807,
      "loss": 0.7711,
      "num_input_tokens_seen": 39280624,
      "step": 67660
    },
    {
      "epoch": 10.078194816800714,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.017403267755494757,
      "loss": 0.7882,
      "num_input_tokens_seen": 39283600,
      "step": 67665
    },
    {
      "epoch": 10.078939529341675,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.017401343268568938,
      "loss": 0.8428,
      "num_input_tokens_seen": 39286800,
      "step": 67670
    },
    {
      "epoch": 10.079684241882633,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.017399418741074864,
      "loss": 0.8042,
      "num_input_tokens_seen": 39289488,
      "step": 67675
    },
    {
      "epoch": 10.080428954423592,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.01739749417304505,
      "loss": 0.7808,
      "num_input_tokens_seen": 39292720,
      "step": 67680
    },
    {
      "epoch": 10.08117366696455,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.017395569564512006,
      "loss": 0.784,
      "num_input_tokens_seen": 39295472,
      "step": 67685
    },
    {
      "epoch": 10.081918379505511,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.017393644915508252,
      "loss": 0.7767,
      "num_input_tokens_seen": 39298288,
      "step": 67690
    },
    {
      "epoch": 10.08266309204647,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0173917202260663,
      "loss": 0.7957,
      "num_input_tokens_seen": 39300944,
      "step": 67695
    },
    {
      "epoch": 10.083407804587429,
      "grad_norm": 0.05078125,
      "learning_rate": 0.01738979549621866,
      "loss": 0.782,
      "num_input_tokens_seen": 39303952,
      "step": 67700
    },
    {
      "epoch": 10.084152517128388,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.017387870725997862,
      "loss": 0.8403,
      "num_input_tokens_seen": 39306832,
      "step": 67705
    },
    {
      "epoch": 10.084897229669348,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.017385945915436407,
      "loss": 0.789,
      "num_input_tokens_seen": 39310224,
      "step": 67710
    },
    {
      "epoch": 10.085641942210307,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.01738402106456683,
      "loss": 0.8228,
      "num_input_tokens_seen": 39313232,
      "step": 67715
    },
    {
      "epoch": 10.086386654751266,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.017382096173421638,
      "loss": 0.8427,
      "num_input_tokens_seen": 39315760,
      "step": 67720
    },
    {
      "epoch": 10.087131367292224,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.017380171242033348,
      "loss": 0.7956,
      "num_input_tokens_seen": 39318544,
      "step": 67725
    },
    {
      "epoch": 10.087876079833185,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.017378246270434484,
      "loss": 0.8193,
      "num_input_tokens_seen": 39321328,
      "step": 67730
    },
    {
      "epoch": 10.088620792374144,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.017376321258657568,
      "loss": 0.817,
      "num_input_tokens_seen": 39324240,
      "step": 67735
    },
    {
      "epoch": 10.089365504915103,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.01737439620673512,
      "loss": 0.7781,
      "num_input_tokens_seen": 39327184,
      "step": 67740
    },
    {
      "epoch": 10.090110217456061,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.01737247111469966,
      "loss": 0.7765,
      "num_input_tokens_seen": 39329968,
      "step": 67745
    },
    {
      "epoch": 10.090854929997022,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.017370545982583717,
      "loss": 0.8033,
      "num_input_tokens_seen": 39332944,
      "step": 67750
    },
    {
      "epoch": 10.09159964253798,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.0173686208104198,
      "loss": 0.7845,
      "num_input_tokens_seen": 39335696,
      "step": 67755
    },
    {
      "epoch": 10.09234435507894,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.017366695598240445,
      "loss": 0.7765,
      "num_input_tokens_seen": 39338736,
      "step": 67760
    },
    {
      "epoch": 10.093089067619898,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.017364770346078175,
      "loss": 0.816,
      "num_input_tokens_seen": 39341744,
      "step": 67765
    },
    {
      "epoch": 10.093833780160859,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01736284505396551,
      "loss": 0.7681,
      "num_input_tokens_seen": 39344720,
      "step": 67770
    },
    {
      "epoch": 10.094578492701817,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.017360919721934986,
      "loss": 0.7934,
      "num_input_tokens_seen": 39347792,
      "step": 67775
    },
    {
      "epoch": 10.095323205242776,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.01735899435001912,
      "loss": 0.7965,
      "num_input_tokens_seen": 39351120,
      "step": 67780
    },
    {
      "epoch": 10.096067917783735,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.017357068938250442,
      "loss": 0.7851,
      "num_input_tokens_seen": 39354128,
      "step": 67785
    },
    {
      "epoch": 10.096812630324695,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.01735514348666148,
      "loss": 0.7766,
      "num_input_tokens_seen": 39358192,
      "step": 67790
    },
    {
      "epoch": 10.097557342865654,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.01735321799528476,
      "loss": 0.8076,
      "num_input_tokens_seen": 39361008,
      "step": 67795
    },
    {
      "epoch": 10.098302055406613,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.017351292464152814,
      "loss": 0.784,
      "num_input_tokens_seen": 39363952,
      "step": 67800
    },
    {
      "epoch": 10.099046767947572,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01734936689329817,
      "loss": 0.8319,
      "num_input_tokens_seen": 39366896,
      "step": 67805
    },
    {
      "epoch": 10.099791480488532,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.017347441282753364,
      "loss": 0.7895,
      "num_input_tokens_seen": 39369808,
      "step": 67810
    },
    {
      "epoch": 10.100536193029491,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.017345515632550924,
      "loss": 0.7758,
      "num_input_tokens_seen": 39372688,
      "step": 67815
    },
    {
      "epoch": 10.10128090557045,
      "grad_norm": 0.01708984375,
      "learning_rate": 0.01734358994272338,
      "loss": 0.7894,
      "num_input_tokens_seen": 39375728,
      "step": 67820
    },
    {
      "epoch": 10.102025618111409,
      "grad_norm": 0.029296875,
      "learning_rate": 0.017341664213303267,
      "loss": 0.7978,
      "num_input_tokens_seen": 39378768,
      "step": 67825
    },
    {
      "epoch": 10.102770330652369,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.017339738444323113,
      "loss": 0.8304,
      "num_input_tokens_seen": 39381808,
      "step": 67830
    },
    {
      "epoch": 10.103515043193328,
      "grad_norm": 0.025390625,
      "learning_rate": 0.017337812635815457,
      "loss": 0.8092,
      "num_input_tokens_seen": 39384720,
      "step": 67835
    },
    {
      "epoch": 10.104259755734287,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.017335886787812834,
      "loss": 0.8006,
      "num_input_tokens_seen": 39387824,
      "step": 67840
    },
    {
      "epoch": 10.105004468275245,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.017333960900347778,
      "loss": 0.7943,
      "num_input_tokens_seen": 39390704,
      "step": 67845
    },
    {
      "epoch": 10.105749180816204,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.017332034973452825,
      "loss": 0.8058,
      "num_input_tokens_seen": 39393648,
      "step": 67850
    },
    {
      "epoch": 10.106493893357165,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01733010900716051,
      "loss": 0.8699,
      "num_input_tokens_seen": 39396432,
      "step": 67855
    },
    {
      "epoch": 10.107238605898123,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.017328183001503374,
      "loss": 0.8049,
      "num_input_tokens_seen": 39399184,
      "step": 67860
    },
    {
      "epoch": 10.107983318439082,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.01732625695651395,
      "loss": 0.7977,
      "num_input_tokens_seen": 39402160,
      "step": 67865
    },
    {
      "epoch": 10.108728030980041,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01732433087222478,
      "loss": 0.7693,
      "num_input_tokens_seen": 39405104,
      "step": 67870
    },
    {
      "epoch": 10.109472743521001,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0173224047486684,
      "loss": 0.8397,
      "num_input_tokens_seen": 39407984,
      "step": 67875
    },
    {
      "epoch": 10.11021745606196,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.017320478585877357,
      "loss": 0.8008,
      "num_input_tokens_seen": 39410960,
      "step": 67880
    },
    {
      "epoch": 10.110962168602919,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.017318552383884186,
      "loss": 0.8033,
      "num_input_tokens_seen": 39413776,
      "step": 67885
    },
    {
      "epoch": 10.111706881143878,
      "grad_norm": 0.015869140625,
      "learning_rate": 0.017316626142721423,
      "loss": 0.7951,
      "num_input_tokens_seen": 39416912,
      "step": 67890
    },
    {
      "epoch": 10.112451593684838,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01731469986242162,
      "loss": 0.8133,
      "num_input_tokens_seen": 39419888,
      "step": 67895
    },
    {
      "epoch": 10.113196306225797,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.017312773543017315,
      "loss": 0.8204,
      "num_input_tokens_seen": 39422576,
      "step": 67900
    },
    {
      "epoch": 10.113941018766756,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.01731084718454105,
      "loss": 0.7963,
      "num_input_tokens_seen": 39426000,
      "step": 67905
    },
    {
      "epoch": 10.114685731307715,
      "grad_norm": 0.033203125,
      "learning_rate": 0.017308920787025374,
      "loss": 0.7777,
      "num_input_tokens_seen": 39428784,
      "step": 67910
    },
    {
      "epoch": 10.115430443848675,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.017306994350502827,
      "loss": 0.7988,
      "num_input_tokens_seen": 39431920,
      "step": 67915
    },
    {
      "epoch": 10.116175156389634,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.017305067875005953,
      "loss": 0.8097,
      "num_input_tokens_seen": 39434608,
      "step": 67920
    },
    {
      "epoch": 10.116919868930593,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.017303141360567303,
      "loss": 0.8181,
      "num_input_tokens_seen": 39437616,
      "step": 67925
    },
    {
      "epoch": 10.117664581471551,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01730121480721942,
      "loss": 0.8051,
      "num_input_tokens_seen": 39440528,
      "step": 67930
    },
    {
      "epoch": 10.118409294012512,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.017299288214994848,
      "loss": 0.8283,
      "num_input_tokens_seen": 39443376,
      "step": 67935
    },
    {
      "epoch": 10.11915400655347,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.017297361583926144,
      "loss": 0.7936,
      "num_input_tokens_seen": 39446032,
      "step": 67940
    },
    {
      "epoch": 10.11989871909443,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.017295434914045844,
      "loss": 0.8056,
      "num_input_tokens_seen": 39448848,
      "step": 67945
    },
    {
      "epoch": 10.120643431635388,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.017293508205386512,
      "loss": 0.8167,
      "num_input_tokens_seen": 39451792,
      "step": 67950
    },
    {
      "epoch": 10.121388144176349,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.017291581457980685,
      "loss": 0.8078,
      "num_input_tokens_seen": 39454704,
      "step": 67955
    },
    {
      "epoch": 10.122132856717307,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.01728965467186092,
      "loss": 0.7899,
      "num_input_tokens_seen": 39457872,
      "step": 67960
    },
    {
      "epoch": 10.122877569258266,
      "grad_norm": 0.02734375,
      "learning_rate": 0.017287727847059762,
      "loss": 0.805,
      "num_input_tokens_seen": 39461968,
      "step": 67965
    },
    {
      "epoch": 10.123622281799225,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.01728580098360977,
      "loss": 0.8011,
      "num_input_tokens_seen": 39464784,
      "step": 67970
    },
    {
      "epoch": 10.124366994340185,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.0172838740815435,
      "loss": 0.8223,
      "num_input_tokens_seen": 39467504,
      "step": 67975
    },
    {
      "epoch": 10.125111706881144,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.01728194714089349,
      "loss": 0.8199,
      "num_input_tokens_seen": 39470608,
      "step": 67980
    },
    {
      "epoch": 10.125856419422103,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.017280020161692302,
      "loss": 0.7904,
      "num_input_tokens_seen": 39473520,
      "step": 67985
    },
    {
      "epoch": 10.126601131963062,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0172780931439725,
      "loss": 0.7872,
      "num_input_tokens_seen": 39476304,
      "step": 67990
    },
    {
      "epoch": 10.127345844504022,
      "grad_norm": 0.0234375,
      "learning_rate": 0.017276166087766623,
      "loss": 0.8004,
      "num_input_tokens_seen": 39479344,
      "step": 67995
    },
    {
      "epoch": 10.128090557044981,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.017274238993107233,
      "loss": 0.797,
      "num_input_tokens_seen": 39481968,
      "step": 68000
    },
    {
      "epoch": 10.12883526958594,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.017272311860026885,
      "loss": 0.8018,
      "num_input_tokens_seen": 39484688,
      "step": 68005
    },
    {
      "epoch": 10.129579982126899,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.017270384688558144,
      "loss": 0.7913,
      "num_input_tokens_seen": 39488208,
      "step": 68010
    },
    {
      "epoch": 10.130324694667857,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.017268457478733554,
      "loss": 0.8132,
      "num_input_tokens_seen": 39491024,
      "step": 68015
    },
    {
      "epoch": 10.131069407208818,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.017266530230585685,
      "loss": 0.8033,
      "num_input_tokens_seen": 39493968,
      "step": 68020
    },
    {
      "epoch": 10.131814119749777,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.01726460294414709,
      "loss": 0.8065,
      "num_input_tokens_seen": 39496720,
      "step": 68025
    },
    {
      "epoch": 10.132558832290735,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.017262675619450333,
      "loss": 0.8112,
      "num_input_tokens_seen": 39499728,
      "step": 68030
    },
    {
      "epoch": 10.133303544831694,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.01726074825652797,
      "loss": 0.7976,
      "num_input_tokens_seen": 39502672,
      "step": 68035
    },
    {
      "epoch": 10.134048257372655,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.017258820855412554,
      "loss": 0.7963,
      "num_input_tokens_seen": 39505776,
      "step": 68040
    },
    {
      "epoch": 10.134792969913613,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.01725689341613667,
      "loss": 0.7734,
      "num_input_tokens_seen": 39508560,
      "step": 68045
    },
    {
      "epoch": 10.135537682454572,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.017254965938732852,
      "loss": 0.8155,
      "num_input_tokens_seen": 39511600,
      "step": 68050
    },
    {
      "epoch": 10.136282394995531,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.017253038423233686,
      "loss": 0.8018,
      "num_input_tokens_seen": 39514480,
      "step": 68055
    },
    {
      "epoch": 10.137027107536491,
      "grad_norm": 0.02734375,
      "learning_rate": 0.01725111086967172,
      "loss": 0.8039,
      "num_input_tokens_seen": 39517168,
      "step": 68060
    },
    {
      "epoch": 10.13777182007745,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.017249183278079526,
      "loss": 0.8163,
      "num_input_tokens_seen": 39519984,
      "step": 68065
    },
    {
      "epoch": 10.138516532618409,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.017247255648489668,
      "loss": 0.7925,
      "num_input_tokens_seen": 39522768,
      "step": 68070
    },
    {
      "epoch": 10.139261245159368,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.017245327980934706,
      "loss": 0.7968,
      "num_input_tokens_seen": 39526000,
      "step": 68075
    },
    {
      "epoch": 10.140005957700328,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.017243400275447216,
      "loss": 0.8025,
      "num_input_tokens_seen": 39529040,
      "step": 68080
    },
    {
      "epoch": 10.140750670241287,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.017241472532059755,
      "loss": 0.8066,
      "num_input_tokens_seen": 39532016,
      "step": 68085
    },
    {
      "epoch": 10.141495382782246,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.01723954475080489,
      "loss": 0.7875,
      "num_input_tokens_seen": 39534736,
      "step": 68090
    },
    {
      "epoch": 10.142240095323205,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.017237616931715195,
      "loss": 0.7879,
      "num_input_tokens_seen": 39537840,
      "step": 68095
    },
    {
      "epoch": 10.142984807864165,
      "grad_norm": 0.0390625,
      "learning_rate": 0.017235689074823236,
      "loss": 0.8068,
      "num_input_tokens_seen": 39540528,
      "step": 68100
    },
    {
      "epoch": 10.143729520405124,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.017233761180161587,
      "loss": 0.797,
      "num_input_tokens_seen": 39543568,
      "step": 68105
    },
    {
      "epoch": 10.144474232946083,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.01723183324776281,
      "loss": 0.8188,
      "num_input_tokens_seen": 39546384,
      "step": 68110
    },
    {
      "epoch": 10.145218945487041,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.01722990527765948,
      "loss": 0.7834,
      "num_input_tokens_seen": 39549296,
      "step": 68115
    },
    {
      "epoch": 10.145963658028002,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.017227977269884166,
      "loss": 0.7985,
      "num_input_tokens_seen": 39552048,
      "step": 68120
    },
    {
      "epoch": 10.14670837056896,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.017226049224469433,
      "loss": 0.8184,
      "num_input_tokens_seen": 39555088,
      "step": 68125
    },
    {
      "epoch": 10.14745308310992,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.017224121141447866,
      "loss": 0.7893,
      "num_input_tokens_seen": 39558064,
      "step": 68130
    },
    {
      "epoch": 10.148197795650878,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.017222193020852036,
      "loss": 0.7908,
      "num_input_tokens_seen": 39560752,
      "step": 68135
    },
    {
      "epoch": 10.148942508191839,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.017220264862714517,
      "loss": 0.8037,
      "num_input_tokens_seen": 39564016,
      "step": 68140
    },
    {
      "epoch": 10.149687220732797,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.01721833666706787,
      "loss": 0.8136,
      "num_input_tokens_seen": 39567184,
      "step": 68145
    },
    {
      "epoch": 10.150431933273756,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.017216408433944688,
      "loss": 0.7941,
      "num_input_tokens_seen": 39570448,
      "step": 68150
    },
    {
      "epoch": 10.151176645814715,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.01721448016337753,
      "loss": 0.792,
      "num_input_tokens_seen": 39573168,
      "step": 68155
    },
    {
      "epoch": 10.151921358355676,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.017212551855398984,
      "loss": 0.8131,
      "num_input_tokens_seen": 39576080,
      "step": 68160
    },
    {
      "epoch": 10.152666070896634,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.017210623510041625,
      "loss": 0.8067,
      "num_input_tokens_seen": 39578928,
      "step": 68165
    },
    {
      "epoch": 10.153410783437593,
      "grad_norm": 0.02734375,
      "learning_rate": 0.01720869512733802,
      "loss": 0.7903,
      "num_input_tokens_seen": 39581840,
      "step": 68170
    },
    {
      "epoch": 10.154155495978552,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.017206766707320766,
      "loss": 0.807,
      "num_input_tokens_seen": 39584880,
      "step": 68175
    },
    {
      "epoch": 10.15490020851951,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01720483825002243,
      "loss": 0.7948,
      "num_input_tokens_seen": 39588144,
      "step": 68180
    },
    {
      "epoch": 10.155644921060471,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.01720290975547559,
      "loss": 0.8249,
      "num_input_tokens_seen": 39591056,
      "step": 68185
    },
    {
      "epoch": 10.15638963360143,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.01720098122371283,
      "loss": 0.8091,
      "num_input_tokens_seen": 39594160,
      "step": 68190
    },
    {
      "epoch": 10.157134346142389,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.01719905265476673,
      "loss": 0.8062,
      "num_input_tokens_seen": 39597136,
      "step": 68195
    },
    {
      "epoch": 10.157879058683347,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.017197124048669868,
      "loss": 0.7877,
      "num_input_tokens_seen": 39600144,
      "step": 68200
    },
    {
      "epoch": 10.158623771224308,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.01719519540545483,
      "loss": 0.81,
      "num_input_tokens_seen": 39603248,
      "step": 68205
    },
    {
      "epoch": 10.159368483765267,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.017193266725154196,
      "loss": 0.792,
      "num_input_tokens_seen": 39606096,
      "step": 68210
    },
    {
      "epoch": 10.160113196306225,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.017191338007800545,
      "loss": 0.8131,
      "num_input_tokens_seen": 39609072,
      "step": 68215
    },
    {
      "epoch": 10.160857908847184,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.017189409253426473,
      "loss": 0.835,
      "num_input_tokens_seen": 39611952,
      "step": 68220
    },
    {
      "epoch": 10.161602621388145,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.01718748046206455,
      "loss": 0.8037,
      "num_input_tokens_seen": 39614736,
      "step": 68225
    },
    {
      "epoch": 10.162347333929103,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.017185551633747378,
      "loss": 0.7934,
      "num_input_tokens_seen": 39617424,
      "step": 68230
    },
    {
      "epoch": 10.163092046470062,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.01718362276850752,
      "loss": 0.8073,
      "num_input_tokens_seen": 39620592,
      "step": 68235
    },
    {
      "epoch": 10.163836759011021,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.017181693866377582,
      "loss": 0.802,
      "num_input_tokens_seen": 39623632,
      "step": 68240
    },
    {
      "epoch": 10.164581471551982,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.017179764927390147,
      "loss": 0.8133,
      "num_input_tokens_seen": 39626704,
      "step": 68245
    },
    {
      "epoch": 10.16532618409294,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.01717783595157779,
      "loss": 0.8114,
      "num_input_tokens_seen": 39629488,
      "step": 68250
    },
    {
      "epoch": 10.166070896633899,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.017175906938973116,
      "loss": 0.7951,
      "num_input_tokens_seen": 39632400,
      "step": 68255
    },
    {
      "epoch": 10.166815609174858,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.0171739778896087,
      "loss": 0.7904,
      "num_input_tokens_seen": 39635536,
      "step": 68260
    },
    {
      "epoch": 10.167560321715818,
      "grad_norm": 0.04296875,
      "learning_rate": 0.017172048803517137,
      "loss": 0.8139,
      "num_input_tokens_seen": 39638480,
      "step": 68265
    },
    {
      "epoch": 10.168305034256777,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01717011968073102,
      "loss": 0.8093,
      "num_input_tokens_seen": 39641616,
      "step": 68270
    },
    {
      "epoch": 10.169049746797736,
      "grad_norm": 0.0234375,
      "learning_rate": 0.017168190521282937,
      "loss": 0.7986,
      "num_input_tokens_seen": 39644528,
      "step": 68275
    },
    {
      "epoch": 10.169794459338695,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.017166261325205483,
      "loss": 0.8125,
      "num_input_tokens_seen": 39647504,
      "step": 68280
    },
    {
      "epoch": 10.170539171879655,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.017164332092531237,
      "loss": 0.7981,
      "num_input_tokens_seen": 39650544,
      "step": 68285
    },
    {
      "epoch": 10.171283884420614,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.017162402823292802,
      "loss": 0.7956,
      "num_input_tokens_seen": 39653296,
      "step": 68290
    },
    {
      "epoch": 10.172028596961573,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.01716047351752277,
      "loss": 0.7924,
      "num_input_tokens_seen": 39656176,
      "step": 68295
    },
    {
      "epoch": 10.172773309502531,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.017158544175253733,
      "loss": 0.8152,
      "num_input_tokens_seen": 39659568,
      "step": 68300
    },
    {
      "epoch": 10.173518022043492,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.01715661479651829,
      "loss": 0.8082,
      "num_input_tokens_seen": 39662352,
      "step": 68305
    },
    {
      "epoch": 10.17426273458445,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.017154685381349035,
      "loss": 0.8126,
      "num_input_tokens_seen": 39665360,
      "step": 68310
    },
    {
      "epoch": 10.17500744712541,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.01715275592977855,
      "loss": 0.8051,
      "num_input_tokens_seen": 39668144,
      "step": 68315
    },
    {
      "epoch": 10.175752159666368,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.017150826441839444,
      "loss": 0.8076,
      "num_input_tokens_seen": 39670960,
      "step": 68320
    },
    {
      "epoch": 10.176496872207329,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.017148896917564316,
      "loss": 0.8054,
      "num_input_tokens_seen": 39674096,
      "step": 68325
    },
    {
      "epoch": 10.177241584748288,
      "grad_norm": 0.03515625,
      "learning_rate": 0.017146967356985756,
      "loss": 0.8134,
      "num_input_tokens_seen": 39676624,
      "step": 68330
    },
    {
      "epoch": 10.177986297289246,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.017145037760136362,
      "loss": 0.8032,
      "num_input_tokens_seen": 39679376,
      "step": 68335
    },
    {
      "epoch": 10.178731009830205,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.017143108127048737,
      "loss": 0.7912,
      "num_input_tokens_seen": 39682256,
      "step": 68340
    },
    {
      "epoch": 10.179475722371166,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01714117845775548,
      "loss": 0.788,
      "num_input_tokens_seen": 39685360,
      "step": 68345
    },
    {
      "epoch": 10.180220434912124,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.017139248752289192,
      "loss": 0.8096,
      "num_input_tokens_seen": 39688272,
      "step": 68350
    },
    {
      "epoch": 10.180965147453083,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.017137319010682463,
      "loss": 0.7801,
      "num_input_tokens_seen": 39691536,
      "step": 68355
    },
    {
      "epoch": 10.181709859994042,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.017135389232967906,
      "loss": 0.808,
      "num_input_tokens_seen": 39694384,
      "step": 68360
    },
    {
      "epoch": 10.182454572535,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01713345941917812,
      "loss": 0.7854,
      "num_input_tokens_seen": 39697136,
      "step": 68365
    },
    {
      "epoch": 10.183199285075961,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0171315295693457,
      "loss": 0.7913,
      "num_input_tokens_seen": 39700176,
      "step": 68370
    },
    {
      "epoch": 10.18394399761692,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01712959968350326,
      "loss": 0.7957,
      "num_input_tokens_seen": 39703152,
      "step": 68375
    },
    {
      "epoch": 10.184688710157879,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.01712766976168339,
      "loss": 0.8165,
      "num_input_tokens_seen": 39706160,
      "step": 68380
    },
    {
      "epoch": 10.185433422698837,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.017125739803918707,
      "loss": 0.8006,
      "num_input_tokens_seen": 39709264,
      "step": 68385
    },
    {
      "epoch": 10.186178135239798,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.017123809810241815,
      "loss": 0.8175,
      "num_input_tokens_seen": 39712112,
      "step": 68390
    },
    {
      "epoch": 10.186922847780757,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.017121879780685308,
      "loss": 0.7815,
      "num_input_tokens_seen": 39714960,
      "step": 68395
    },
    {
      "epoch": 10.187667560321715,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.017119949715281798,
      "loss": 0.7735,
      "num_input_tokens_seen": 39718032,
      "step": 68400
    },
    {
      "epoch": 10.188412272862674,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.017118019614063895,
      "loss": 0.8433,
      "num_input_tokens_seen": 39721072,
      "step": 68405
    },
    {
      "epoch": 10.189156985403635,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.01711608947706421,
      "loss": 0.7886,
      "num_input_tokens_seen": 39724176,
      "step": 68410
    },
    {
      "epoch": 10.189901697944594,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.017114159304315337,
      "loss": 0.8141,
      "num_input_tokens_seen": 39726992,
      "step": 68415
    },
    {
      "epoch": 10.190646410485552,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.017112229095849887,
      "loss": 0.7986,
      "num_input_tokens_seen": 39730064,
      "step": 68420
    },
    {
      "epoch": 10.191391123026511,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.01711029885170048,
      "loss": 0.79,
      "num_input_tokens_seen": 39732976,
      "step": 68425
    },
    {
      "epoch": 10.192135835567472,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.017108368571899717,
      "loss": 0.7797,
      "num_input_tokens_seen": 39736144,
      "step": 68430
    },
    {
      "epoch": 10.19288054810843,
      "grad_norm": 0.048828125,
      "learning_rate": 0.017106438256480207,
      "loss": 0.7994,
      "num_input_tokens_seen": 39738992,
      "step": 68435
    },
    {
      "epoch": 10.19362526064939,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.017104507905474565,
      "loss": 0.7811,
      "num_input_tokens_seen": 39741808,
      "step": 68440
    },
    {
      "epoch": 10.194369973190348,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.017102577518915404,
      "loss": 0.8239,
      "num_input_tokens_seen": 39744720,
      "step": 68445
    },
    {
      "epoch": 10.195114685731308,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.01710064709683533,
      "loss": 0.7959,
      "num_input_tokens_seen": 39747408,
      "step": 68450
    },
    {
      "epoch": 10.195859398272267,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.01709871663926696,
      "loss": 0.7915,
      "num_input_tokens_seen": 39750128,
      "step": 68455
    },
    {
      "epoch": 10.196604110813226,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.017096786146242902,
      "loss": 0.8023,
      "num_input_tokens_seen": 39753296,
      "step": 68460
    },
    {
      "epoch": 10.197348823354185,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.017094855617795775,
      "loss": 0.8327,
      "num_input_tokens_seen": 39756016,
      "step": 68465
    },
    {
      "epoch": 10.198093535895145,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.017092925053958197,
      "loss": 0.7825,
      "num_input_tokens_seen": 39759312,
      "step": 68470
    },
    {
      "epoch": 10.198838248436104,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.017090994454762774,
      "loss": 0.811,
      "num_input_tokens_seen": 39762192,
      "step": 68475
    },
    {
      "epoch": 10.199582960977063,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01708906382024212,
      "loss": 0.826,
      "num_input_tokens_seen": 39765232,
      "step": 68480
    },
    {
      "epoch": 10.200327673518021,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.017087133150428858,
      "loss": 0.8115,
      "num_input_tokens_seen": 39768208,
      "step": 68485
    },
    {
      "epoch": 10.201072386058982,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.017085202445355604,
      "loss": 0.7896,
      "num_input_tokens_seen": 39771312,
      "step": 68490
    },
    {
      "epoch": 10.20181709859994,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01708327170505497,
      "loss": 0.8181,
      "num_input_tokens_seen": 39774192,
      "step": 68495
    },
    {
      "epoch": 10.2025618111409,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.017081340929559582,
      "loss": 0.8041,
      "num_input_tokens_seen": 39777424,
      "step": 68500
    },
    {
      "epoch": 10.203306523681858,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.017079410118902058,
      "loss": 0.7893,
      "num_input_tokens_seen": 39780208,
      "step": 68505
    },
    {
      "epoch": 10.204051236222819,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01707747927311501,
      "loss": 0.8063,
      "num_input_tokens_seen": 39782928,
      "step": 68510
    },
    {
      "epoch": 10.204795948763778,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.017075548392231063,
      "loss": 0.8049,
      "num_input_tokens_seen": 39785488,
      "step": 68515
    },
    {
      "epoch": 10.205540661304736,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.017073617476282832,
      "loss": 0.799,
      "num_input_tokens_seen": 39788176,
      "step": 68520
    },
    {
      "epoch": 10.206285373845695,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01707168652530294,
      "loss": 0.7984,
      "num_input_tokens_seen": 39790832,
      "step": 68525
    },
    {
      "epoch": 10.207030086386654,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01706975553932401,
      "loss": 0.8041,
      "num_input_tokens_seen": 39793968,
      "step": 68530
    },
    {
      "epoch": 10.207774798927614,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.01706782451837867,
      "loss": 0.8196,
      "num_input_tokens_seen": 39796816,
      "step": 68535
    },
    {
      "epoch": 10.208519511468573,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.01706589346249953,
      "loss": 0.7919,
      "num_input_tokens_seen": 39799824,
      "step": 68540
    },
    {
      "epoch": 10.209264224009532,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01706396237171922,
      "loss": 0.7981,
      "num_input_tokens_seen": 39802896,
      "step": 68545
    },
    {
      "epoch": 10.21000893655049,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.017062031246070367,
      "loss": 0.781,
      "num_input_tokens_seen": 39806128,
      "step": 68550
    },
    {
      "epoch": 10.210753649091451,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.017060100085585592,
      "loss": 0.7974,
      "num_input_tokens_seen": 39809040,
      "step": 68555
    },
    {
      "epoch": 10.21149836163241,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.017058168890297515,
      "loss": 0.7837,
      "num_input_tokens_seen": 39811920,
      "step": 68560
    },
    {
      "epoch": 10.212243074173369,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.017056237660238772,
      "loss": 0.8015,
      "num_input_tokens_seen": 39814672,
      "step": 68565
    },
    {
      "epoch": 10.212987786714327,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.017054306395441976,
      "loss": 0.7854,
      "num_input_tokens_seen": 39817456,
      "step": 68570
    },
    {
      "epoch": 10.213732499255288,
      "grad_norm": 0.04296875,
      "learning_rate": 0.01705237509593977,
      "loss": 0.8019,
      "num_input_tokens_seen": 39820496,
      "step": 68575
    },
    {
      "epoch": 10.214477211796247,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01705044376176477,
      "loss": 0.7951,
      "num_input_tokens_seen": 39823120,
      "step": 68580
    },
    {
      "epoch": 10.215221924337206,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0170485123929496,
      "loss": 0.7989,
      "num_input_tokens_seen": 39826000,
      "step": 68585
    },
    {
      "epoch": 10.215966636878164,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.017046580989526907,
      "loss": 0.7964,
      "num_input_tokens_seen": 39828720,
      "step": 68590
    },
    {
      "epoch": 10.216711349419125,
      "grad_norm": 0.044921875,
      "learning_rate": 0.017044649551529303,
      "loss": 0.7892,
      "num_input_tokens_seen": 39831984,
      "step": 68595
    },
    {
      "epoch": 10.217456061960084,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.01704271807898942,
      "loss": 0.8013,
      "num_input_tokens_seen": 39835056,
      "step": 68600
    },
    {
      "epoch": 10.218200774501042,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.017040786571939893,
      "loss": 0.7971,
      "num_input_tokens_seen": 39838032,
      "step": 68605
    },
    {
      "epoch": 10.218945487042001,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.017038855030413354,
      "loss": 0.8479,
      "num_input_tokens_seen": 39840912,
      "step": 68610
    },
    {
      "epoch": 10.219690199582962,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01703692345444243,
      "loss": 0.8159,
      "num_input_tokens_seen": 39843600,
      "step": 68615
    },
    {
      "epoch": 10.22043491212392,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.01703499184405975,
      "loss": 0.8045,
      "num_input_tokens_seen": 39846416,
      "step": 68620
    },
    {
      "epoch": 10.22117962466488,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.017033060199297952,
      "loss": 0.8075,
      "num_input_tokens_seen": 39849264,
      "step": 68625
    },
    {
      "epoch": 10.221924337205838,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01703112852018967,
      "loss": 0.8083,
      "num_input_tokens_seen": 39852016,
      "step": 68630
    },
    {
      "epoch": 10.222669049746798,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.017029196806767538,
      "loss": 0.8244,
      "num_input_tokens_seen": 39854896,
      "step": 68635
    },
    {
      "epoch": 10.223413762287757,
      "grad_norm": 0.0859375,
      "learning_rate": 0.01702726505906419,
      "loss": 0.8162,
      "num_input_tokens_seen": 39857776,
      "step": 68640
    },
    {
      "epoch": 10.224158474828716,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.017025333277112257,
      "loss": 0.7991,
      "num_input_tokens_seen": 39860688,
      "step": 68645
    },
    {
      "epoch": 10.224903187369675,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.017023401460944373,
      "loss": 0.7917,
      "num_input_tokens_seen": 39863504,
      "step": 68650
    },
    {
      "epoch": 10.225647899910635,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.01702146961059318,
      "loss": 0.8002,
      "num_input_tokens_seen": 39866544,
      "step": 68655
    },
    {
      "epoch": 10.226392612451594,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.01701953772609131,
      "loss": 0.7923,
      "num_input_tokens_seen": 39869392,
      "step": 68660
    },
    {
      "epoch": 10.227137324992553,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.017017605807471407,
      "loss": 0.7973,
      "num_input_tokens_seen": 39872816,
      "step": 68665
    },
    {
      "epoch": 10.227882037533512,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.017015673854766107,
      "loss": 0.7937,
      "num_input_tokens_seen": 39875696,
      "step": 68670
    },
    {
      "epoch": 10.228626750074472,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.017013741868008043,
      "loss": 0.8052,
      "num_input_tokens_seen": 39878224,
      "step": 68675
    },
    {
      "epoch": 10.22937146261543,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.017011809847229855,
      "loss": 0.8077,
      "num_input_tokens_seen": 39880880,
      "step": 68680
    },
    {
      "epoch": 10.23011617515639,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.017009877792464184,
      "loss": 0.7982,
      "num_input_tokens_seen": 39883728,
      "step": 68685
    },
    {
      "epoch": 10.230860887697348,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01700794570374367,
      "loss": 0.7886,
      "num_input_tokens_seen": 39886768,
      "step": 68690
    },
    {
      "epoch": 10.231605600238307,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.017006013581100956,
      "loss": 0.7964,
      "num_input_tokens_seen": 39889808,
      "step": 68695
    },
    {
      "epoch": 10.232350312779268,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.01700408142456868,
      "loss": 0.7819,
      "num_input_tokens_seen": 39892784,
      "step": 68700
    },
    {
      "epoch": 10.233095025320226,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01700214923417949,
      "loss": 0.7777,
      "num_input_tokens_seen": 39895440,
      "step": 68705
    },
    {
      "epoch": 10.233839737861185,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.01700021700996602,
      "loss": 0.7841,
      "num_input_tokens_seen": 39898096,
      "step": 68710
    },
    {
      "epoch": 10.234584450402144,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.016998284751960916,
      "loss": 0.7881,
      "num_input_tokens_seen": 39901424,
      "step": 68715
    },
    {
      "epoch": 10.235329162943104,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.016996352460196824,
      "loss": 0.8599,
      "num_input_tokens_seen": 39904400,
      "step": 68720
    },
    {
      "epoch": 10.236073875484063,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.016994420134706385,
      "loss": 0.7825,
      "num_input_tokens_seen": 39906928,
      "step": 68725
    },
    {
      "epoch": 10.236818588025022,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.016992487775522244,
      "loss": 0.7965,
      "num_input_tokens_seen": 39909744,
      "step": 68730
    },
    {
      "epoch": 10.23756330056598,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.016990555382677044,
      "loss": 0.7941,
      "num_input_tokens_seen": 39912560,
      "step": 68735
    },
    {
      "epoch": 10.238308013106941,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.01698862295620344,
      "loss": 0.7771,
      "num_input_tokens_seen": 39915152,
      "step": 68740
    },
    {
      "epoch": 10.2390527256479,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.01698669049613407,
      "loss": 0.7878,
      "num_input_tokens_seen": 39918000,
      "step": 68745
    },
    {
      "epoch": 10.239797438188859,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.016984758002501585,
      "loss": 0.7801,
      "num_input_tokens_seen": 39920624,
      "step": 68750
    },
    {
      "epoch": 10.240542150729818,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01698282547533863,
      "loss": 0.7826,
      "num_input_tokens_seen": 39923792,
      "step": 68755
    },
    {
      "epoch": 10.241286863270778,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01698089291467785,
      "loss": 0.7735,
      "num_input_tokens_seen": 39926608,
      "step": 68760
    },
    {
      "epoch": 10.242031575811737,
      "grad_norm": 0.07421875,
      "learning_rate": 0.016978960320551902,
      "loss": 0.7747,
      "num_input_tokens_seen": 39929296,
      "step": 68765
    },
    {
      "epoch": 10.242776288352696,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.016977027692993425,
      "loss": 0.8115,
      "num_input_tokens_seen": 39932144,
      "step": 68770
    },
    {
      "epoch": 10.243521000893654,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.016975095032035082,
      "loss": 0.7851,
      "num_input_tokens_seen": 39934992,
      "step": 68775
    },
    {
      "epoch": 10.244265713434615,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01697316233770951,
      "loss": 0.7775,
      "num_input_tokens_seen": 39938000,
      "step": 68780
    },
    {
      "epoch": 10.245010425975574,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.01697122961004937,
      "loss": 0.7671,
      "num_input_tokens_seen": 39940880,
      "step": 68785
    },
    {
      "epoch": 10.245755138516532,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.016969296849087306,
      "loss": 0.8363,
      "num_input_tokens_seen": 39943536,
      "step": 68790
    },
    {
      "epoch": 10.246499851057491,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.016967364054855976,
      "loss": 0.7993,
      "num_input_tokens_seen": 39946352,
      "step": 68795
    },
    {
      "epoch": 10.247244563598452,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.016965431227388028,
      "loss": 0.8043,
      "num_input_tokens_seen": 39949200,
      "step": 68800
    },
    {
      "epoch": 10.24798927613941,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.01696349836671612,
      "loss": 0.7954,
      "num_input_tokens_seen": 39952144,
      "step": 68805
    },
    {
      "epoch": 10.24873398868037,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0169615654728729,
      "loss": 0.7709,
      "num_input_tokens_seen": 39955376,
      "step": 68810
    },
    {
      "epoch": 10.249478701221328,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.016959632545891022,
      "loss": 0.7987,
      "num_input_tokens_seen": 39958096,
      "step": 68815
    },
    {
      "epoch": 10.250223413762289,
      "grad_norm": 0.044921875,
      "learning_rate": 0.016957699585803145,
      "loss": 0.7583,
      "num_input_tokens_seen": 39960816,
      "step": 68820
    },
    {
      "epoch": 10.250968126303247,
      "grad_norm": 0.037109375,
      "learning_rate": 0.016955766592641926,
      "loss": 0.8028,
      "num_input_tokens_seen": 39963984,
      "step": 68825
    },
    {
      "epoch": 10.251712838844206,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.016953833566440018,
      "loss": 0.7738,
      "num_input_tokens_seen": 39966928,
      "step": 68830
    },
    {
      "epoch": 10.252457551385165,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01695190050723008,
      "loss": 0.7712,
      "num_input_tokens_seen": 39969840,
      "step": 68835
    },
    {
      "epoch": 10.253202263926125,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.016949967415044762,
      "loss": 0.8295,
      "num_input_tokens_seen": 39972784,
      "step": 68840
    },
    {
      "epoch": 10.253946976467084,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.016948034289916727,
      "loss": 0.7968,
      "num_input_tokens_seen": 39975632,
      "step": 68845
    },
    {
      "epoch": 10.254691689008043,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.016946101131878635,
      "loss": 0.8,
      "num_input_tokens_seen": 39978480,
      "step": 68850
    },
    {
      "epoch": 10.255436401549002,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.01694416794096314,
      "loss": 0.8193,
      "num_input_tokens_seen": 39981328,
      "step": 68855
    },
    {
      "epoch": 10.256181114089962,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0169422347172029,
      "loss": 0.7924,
      "num_input_tokens_seen": 39984048,
      "step": 68860
    },
    {
      "epoch": 10.256925826630921,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.016940301460630584,
      "loss": 0.8132,
      "num_input_tokens_seen": 39986928,
      "step": 68865
    },
    {
      "epoch": 10.25767053917188,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.01693836817127885,
      "loss": 0.7956,
      "num_input_tokens_seen": 39990096,
      "step": 68870
    },
    {
      "epoch": 10.258415251712838,
      "grad_norm": 0.056640625,
      "learning_rate": 0.016936434849180355,
      "loss": 0.8036,
      "num_input_tokens_seen": 39992912,
      "step": 68875
    },
    {
      "epoch": 10.259159964253797,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.016934501494367756,
      "loss": 0.7808,
      "num_input_tokens_seen": 39995920,
      "step": 68880
    },
    {
      "epoch": 10.259904676794758,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.01693256810687372,
      "loss": 0.8096,
      "num_input_tokens_seen": 39999088,
      "step": 68885
    },
    {
      "epoch": 10.260649389335716,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.01693063468673091,
      "loss": 0.7999,
      "num_input_tokens_seen": 40002192,
      "step": 68890
    },
    {
      "epoch": 10.261394101876675,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.016928701233971992,
      "loss": 0.7696,
      "num_input_tokens_seen": 40005040,
      "step": 68895
    },
    {
      "epoch": 10.262138814417634,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.01692676774862963,
      "loss": 0.7732,
      "num_input_tokens_seen": 40007760,
      "step": 68900
    },
    {
      "epoch": 10.262883526958595,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01692483423073648,
      "loss": 0.7759,
      "num_input_tokens_seen": 40010640,
      "step": 68905
    },
    {
      "epoch": 10.263628239499553,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.016922900680325214,
      "loss": 0.7858,
      "num_input_tokens_seen": 40013488,
      "step": 68910
    },
    {
      "epoch": 10.264372952040512,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.016920967097428498,
      "loss": 0.8278,
      "num_input_tokens_seen": 40016368,
      "step": 68915
    },
    {
      "epoch": 10.26511766458147,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.01691903348207899,
      "loss": 0.7872,
      "num_input_tokens_seen": 40019376,
      "step": 68920
    },
    {
      "epoch": 10.265862377122431,
      "grad_norm": 0.037109375,
      "learning_rate": 0.016917099834309363,
      "loss": 0.7823,
      "num_input_tokens_seen": 40022768,
      "step": 68925
    },
    {
      "epoch": 10.26660708966339,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.01691516615415228,
      "loss": 0.7876,
      "num_input_tokens_seen": 40025616,
      "step": 68930
    },
    {
      "epoch": 10.267351802204349,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.016913232441640415,
      "loss": 0.8443,
      "num_input_tokens_seen": 40028560,
      "step": 68935
    },
    {
      "epoch": 10.268096514745308,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.016911298696806437,
      "loss": 0.7819,
      "num_input_tokens_seen": 40031600,
      "step": 68940
    },
    {
      "epoch": 10.268841227286268,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.016909364919683003,
      "loss": 0.8111,
      "num_input_tokens_seen": 40034576,
      "step": 68945
    },
    {
      "epoch": 10.269585939827227,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.016907431110302792,
      "loss": 0.806,
      "num_input_tokens_seen": 40037584,
      "step": 68950
    },
    {
      "epoch": 10.270330652368186,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.016905497268698475,
      "loss": 0.8112,
      "num_input_tokens_seen": 40040496,
      "step": 68955
    },
    {
      "epoch": 10.271075364909144,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.016903563394902715,
      "loss": 0.8009,
      "num_input_tokens_seen": 40043696,
      "step": 68960
    },
    {
      "epoch": 10.271820077450105,
      "grad_norm": 0.037109375,
      "learning_rate": 0.016901629488948182,
      "loss": 0.7814,
      "num_input_tokens_seen": 40046480,
      "step": 68965
    },
    {
      "epoch": 10.272564789991064,
      "grad_norm": 0.041015625,
      "learning_rate": 0.016899695550867556,
      "loss": 0.8004,
      "num_input_tokens_seen": 40049488,
      "step": 68970
    },
    {
      "epoch": 10.273309502532022,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0168977615806935,
      "loss": 0.8304,
      "num_input_tokens_seen": 40052336,
      "step": 68975
    },
    {
      "epoch": 10.274054215072981,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.016895827578458696,
      "loss": 0.7952,
      "num_input_tokens_seen": 40055120,
      "step": 68980
    },
    {
      "epoch": 10.274798927613942,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01689389354419581,
      "loss": 0.7568,
      "num_input_tokens_seen": 40057808,
      "step": 68985
    },
    {
      "epoch": 10.2755436401549,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.016891959477937514,
      "loss": 0.8376,
      "num_input_tokens_seen": 40060528,
      "step": 68990
    },
    {
      "epoch": 10.27628835269586,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.016890025379716494,
      "loss": 0.7837,
      "num_input_tokens_seen": 40063280,
      "step": 68995
    },
    {
      "epoch": 10.277033065236818,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.016888091249565407,
      "loss": 0.8033,
      "num_input_tokens_seen": 40066256,
      "step": 69000
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.016886157087516947,
      "loss": 0.8035,
      "num_input_tokens_seen": 40069232,
      "step": 69005
    },
    {
      "epoch": 10.278522490318737,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01688422289360377,
      "loss": 0.8301,
      "num_input_tokens_seen": 40071952,
      "step": 69010
    },
    {
      "epoch": 10.279267202859696,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01688228866785857,
      "loss": 0.8221,
      "num_input_tokens_seen": 40074768,
      "step": 69015
    },
    {
      "epoch": 10.280011915400655,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.016880354410314012,
      "loss": 0.8088,
      "num_input_tokens_seen": 40077616,
      "step": 69020
    },
    {
      "epoch": 10.280756627941615,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.016878420121002775,
      "loss": 0.8041,
      "num_input_tokens_seen": 40080560,
      "step": 69025
    },
    {
      "epoch": 10.281501340482574,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01687648579995754,
      "loss": 0.7874,
      "num_input_tokens_seen": 40083408,
      "step": 69030
    },
    {
      "epoch": 10.282246053023533,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.016874551447210992,
      "loss": 0.8077,
      "num_input_tokens_seen": 40086352,
      "step": 69035
    },
    {
      "epoch": 10.282990765564492,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0168726170627958,
      "loss": 0.8116,
      "num_input_tokens_seen": 40089264,
      "step": 69040
    },
    {
      "epoch": 10.283735478105452,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.016870682646744642,
      "loss": 0.7984,
      "num_input_tokens_seen": 40092272,
      "step": 69045
    },
    {
      "epoch": 10.284480190646411,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.016868748199090202,
      "loss": 0.7954,
      "num_input_tokens_seen": 40095056,
      "step": 69050
    },
    {
      "epoch": 10.28522490318737,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01686681371986516,
      "loss": 0.7896,
      "num_input_tokens_seen": 40098288,
      "step": 69055
    },
    {
      "epoch": 10.285969615728328,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0168648792091022,
      "loss": 0.7924,
      "num_input_tokens_seen": 40101232,
      "step": 69060
    },
    {
      "epoch": 10.286714328269287,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.016862944666834,
      "loss": 0.8187,
      "num_input_tokens_seen": 40104048,
      "step": 69065
    },
    {
      "epoch": 10.287459040810248,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.016861010093093244,
      "loss": 0.7829,
      "num_input_tokens_seen": 40107024,
      "step": 69070
    },
    {
      "epoch": 10.288203753351207,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.016859075487912614,
      "loss": 0.8043,
      "num_input_tokens_seen": 40109904,
      "step": 69075
    },
    {
      "epoch": 10.288948465892165,
      "grad_norm": 0.0390625,
      "learning_rate": 0.016857140851324795,
      "loss": 0.7974,
      "num_input_tokens_seen": 40112816,
      "step": 69080
    },
    {
      "epoch": 10.289693178433124,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.016855206183362465,
      "loss": 0.8022,
      "num_input_tokens_seen": 40115664,
      "step": 69085
    },
    {
      "epoch": 10.290437890974085,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.016853271484058314,
      "loss": 0.802,
      "num_input_tokens_seen": 40118128,
      "step": 69090
    },
    {
      "epoch": 10.291182603515043,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.016851336753445023,
      "loss": 0.7991,
      "num_input_tokens_seen": 40121232,
      "step": 69095
    },
    {
      "epoch": 10.291927316056002,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01684940199155528,
      "loss": 0.8127,
      "num_input_tokens_seen": 40124304,
      "step": 69100
    },
    {
      "epoch": 10.29267202859696,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.016847467198421773,
      "loss": 0.7807,
      "num_input_tokens_seen": 40127056,
      "step": 69105
    },
    {
      "epoch": 10.293416741137921,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.016845532374077182,
      "loss": 0.851,
      "num_input_tokens_seen": 40129872,
      "step": 69110
    },
    {
      "epoch": 10.29416145367888,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.016843597518554198,
      "loss": 0.7902,
      "num_input_tokens_seen": 40132912,
      "step": 69115
    },
    {
      "epoch": 10.294906166219839,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.01684166263188551,
      "loss": 0.8041,
      "num_input_tokens_seen": 40135568,
      "step": 69120
    },
    {
      "epoch": 10.295650878760798,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0168397277141038,
      "loss": 0.8035,
      "num_input_tokens_seen": 40138192,
      "step": 69125
    },
    {
      "epoch": 10.296395591301758,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.01683779276524176,
      "loss": 0.8076,
      "num_input_tokens_seen": 40141264,
      "step": 69130
    },
    {
      "epoch": 10.297140303842717,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.016835857785332087,
      "loss": 0.776,
      "num_input_tokens_seen": 40144112,
      "step": 69135
    },
    {
      "epoch": 10.297885016383676,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.016833922774407453,
      "loss": 0.7984,
      "num_input_tokens_seen": 40147120,
      "step": 69140
    },
    {
      "epoch": 10.298629728924634,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.01683198773250056,
      "loss": 0.7814,
      "num_input_tokens_seen": 40150064,
      "step": 69145
    },
    {
      "epoch": 10.299374441465595,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.016830052659644098,
      "loss": 0.8339,
      "num_input_tokens_seen": 40153040,
      "step": 69150
    },
    {
      "epoch": 10.300119154006554,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.016828117555870752,
      "loss": 0.8105,
      "num_input_tokens_seen": 40156144,
      "step": 69155
    },
    {
      "epoch": 10.300863866547513,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.016826182421213223,
      "loss": 0.7789,
      "num_input_tokens_seen": 40159056,
      "step": 69160
    },
    {
      "epoch": 10.301608579088471,
      "grad_norm": 0.03515625,
      "learning_rate": 0.016824247255704194,
      "loss": 0.8041,
      "num_input_tokens_seen": 40161968,
      "step": 69165
    },
    {
      "epoch": 10.302353291629432,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.01682231205937637,
      "loss": 0.7938,
      "num_input_tokens_seen": 40165008,
      "step": 69170
    },
    {
      "epoch": 10.30309800417039,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.01682037683226243,
      "loss": 0.8059,
      "num_input_tokens_seen": 40168048,
      "step": 69175
    },
    {
      "epoch": 10.30384271671135,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.016818441574395073,
      "loss": 0.766,
      "num_input_tokens_seen": 40170832,
      "step": 69180
    },
    {
      "epoch": 10.304587429252308,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.01681650628580699,
      "loss": 0.7922,
      "num_input_tokens_seen": 40173840,
      "step": 69185
    },
    {
      "epoch": 10.305332141793269,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.016814570966530884,
      "loss": 0.7846,
      "num_input_tokens_seen": 40176688,
      "step": 69190
    },
    {
      "epoch": 10.306076854334227,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.016812635616599448,
      "loss": 0.8087,
      "num_input_tokens_seen": 40179184,
      "step": 69195
    },
    {
      "epoch": 10.306821566875186,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.016810700236045376,
      "loss": 0.7832,
      "num_input_tokens_seen": 40182128,
      "step": 69200
    },
    {
      "epoch": 10.307566279416145,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.016808764824901363,
      "loss": 0.78,
      "num_input_tokens_seen": 40185008,
      "step": 69205
    },
    {
      "epoch": 10.308310991957104,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.016806829383200102,
      "loss": 0.7728,
      "num_input_tokens_seen": 40187792,
      "step": 69210
    },
    {
      "epoch": 10.309055704498064,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.016804893910974297,
      "loss": 0.8032,
      "num_input_tokens_seen": 40190896,
      "step": 69215
    },
    {
      "epoch": 10.309800417039023,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.016802958408256644,
      "loss": 0.7853,
      "num_input_tokens_seen": 40193936,
      "step": 69220
    },
    {
      "epoch": 10.310545129579982,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.016801022875079837,
      "loss": 0.8333,
      "num_input_tokens_seen": 40196880,
      "step": 69225
    },
    {
      "epoch": 10.31128984212094,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.016799087311476585,
      "loss": 0.7748,
      "num_input_tokens_seen": 40199696,
      "step": 69230
    },
    {
      "epoch": 10.312034554661901,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.016797151717479582,
      "loss": 0.7997,
      "num_input_tokens_seen": 40202928,
      "step": 69235
    },
    {
      "epoch": 10.31277926720286,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.01679521609312153,
      "loss": 0.8185,
      "num_input_tokens_seen": 40205872,
      "step": 69240
    },
    {
      "epoch": 10.313523979743819,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.01679328043843512,
      "loss": 0.8179,
      "num_input_tokens_seen": 40208656,
      "step": 69245
    },
    {
      "epoch": 10.314268692284777,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.01679134475345306,
      "loss": 0.8053,
      "num_input_tokens_seen": 40212016,
      "step": 69250
    },
    {
      "epoch": 10.315013404825738,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01678940903820805,
      "loss": 0.797,
      "num_input_tokens_seen": 40214928,
      "step": 69255
    },
    {
      "epoch": 10.315758117366697,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.016787473292732794,
      "loss": 0.8192,
      "num_input_tokens_seen": 40218000,
      "step": 69260
    },
    {
      "epoch": 10.316502829907655,
      "grad_norm": 0.0390625,
      "learning_rate": 0.016785537517059995,
      "loss": 0.779,
      "num_input_tokens_seen": 40221072,
      "step": 69265
    },
    {
      "epoch": 10.317247542448614,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.016783601711222357,
      "loss": 0.8136,
      "num_input_tokens_seen": 40223888,
      "step": 69270
    },
    {
      "epoch": 10.317992254989575,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.016781665875252576,
      "loss": 0.7996,
      "num_input_tokens_seen": 40226480,
      "step": 69275
    },
    {
      "epoch": 10.318736967530533,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.016779730009183366,
      "loss": 0.8024,
      "num_input_tokens_seen": 40229584,
      "step": 69280
    },
    {
      "epoch": 10.319481680071492,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.016777794113047424,
      "loss": 0.7857,
      "num_input_tokens_seen": 40232528,
      "step": 69285
    },
    {
      "epoch": 10.320226392612451,
      "grad_norm": 0.041015625,
      "learning_rate": 0.016775858186877456,
      "loss": 0.8125,
      "num_input_tokens_seen": 40235248,
      "step": 69290
    },
    {
      "epoch": 10.320971105153411,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.01677392223070617,
      "loss": 0.7935,
      "num_input_tokens_seen": 40238096,
      "step": 69295
    },
    {
      "epoch": 10.32171581769437,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.016771986244566273,
      "loss": 0.818,
      "num_input_tokens_seen": 40240720,
      "step": 69300
    },
    {
      "epoch": 10.322460530235329,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.016770050228490464,
      "loss": 0.7964,
      "num_input_tokens_seen": 40243728,
      "step": 69305
    },
    {
      "epoch": 10.323205242776288,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.016768114182511458,
      "loss": 0.8012,
      "num_input_tokens_seen": 40246480,
      "step": 69310
    },
    {
      "epoch": 10.323949955317248,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.016766178106661957,
      "loss": 0.806,
      "num_input_tokens_seen": 40249552,
      "step": 69315
    },
    {
      "epoch": 10.324694667858207,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.016764242000974677,
      "loss": 0.8029,
      "num_input_tokens_seen": 40252624,
      "step": 69320
    },
    {
      "epoch": 10.325439380399166,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.01676230586548232,
      "loss": 0.8199,
      "num_input_tokens_seen": 40255664,
      "step": 69325
    },
    {
      "epoch": 10.326184092940125,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.016760369700217596,
      "loss": 0.8064,
      "num_input_tokens_seen": 40258608,
      "step": 69330
    },
    {
      "epoch": 10.326928805481085,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.016758433505213215,
      "loss": 0.7883,
      "num_input_tokens_seen": 40261264,
      "step": 69335
    },
    {
      "epoch": 10.327673518022044,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.016756497280501888,
      "loss": 0.7991,
      "num_input_tokens_seen": 40264336,
      "step": 69340
    },
    {
      "epoch": 10.328418230563003,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.016754561026116322,
      "loss": 0.7875,
      "num_input_tokens_seen": 40267184,
      "step": 69345
    },
    {
      "epoch": 10.329162943103961,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.01675262474208923,
      "loss": 0.7963,
      "num_input_tokens_seen": 40270000,
      "step": 69350
    },
    {
      "epoch": 10.329907655644922,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.016750688428453328,
      "loss": 0.7985,
      "num_input_tokens_seen": 40272976,
      "step": 69355
    },
    {
      "epoch": 10.33065236818588,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.016748752085241323,
      "loss": 0.7922,
      "num_input_tokens_seen": 40275696,
      "step": 69360
    },
    {
      "epoch": 10.33139708072684,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.01674681571248593,
      "loss": 0.7947,
      "num_input_tokens_seen": 40278416,
      "step": 69365
    },
    {
      "epoch": 10.332141793267798,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.016744879310219864,
      "loss": 0.7784,
      "num_input_tokens_seen": 40281392,
      "step": 69370
    },
    {
      "epoch": 10.332886505808759,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.01674294287847583,
      "loss": 0.8386,
      "num_input_tokens_seen": 40284240,
      "step": 69375
    },
    {
      "epoch": 10.333631218349717,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.01674100641728655,
      "loss": 0.794,
      "num_input_tokens_seen": 40287376,
      "step": 69380
    },
    {
      "epoch": 10.334375930890676,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01673906992668473,
      "loss": 0.7951,
      "num_input_tokens_seen": 40291056,
      "step": 69385
    },
    {
      "epoch": 10.335120643431635,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.016737133406703096,
      "loss": 0.7898,
      "num_input_tokens_seen": 40293776,
      "step": 69390
    },
    {
      "epoch": 10.335865355972594,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01673519685737436,
      "loss": 0.8078,
      "num_input_tokens_seen": 40296720,
      "step": 69395
    },
    {
      "epoch": 10.336610068513554,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.016733260278731236,
      "loss": 0.7895,
      "num_input_tokens_seen": 40299568,
      "step": 69400
    },
    {
      "epoch": 10.337354781054513,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.016731323670806444,
      "loss": 0.785,
      "num_input_tokens_seen": 40302320,
      "step": 69405
    },
    {
      "epoch": 10.338099493595472,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.016729387033632696,
      "loss": 0.7939,
      "num_input_tokens_seen": 40304848,
      "step": 69410
    },
    {
      "epoch": 10.33884420613643,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01672745036724271,
      "loss": 0.779,
      "num_input_tokens_seen": 40307632,
      "step": 69415
    },
    {
      "epoch": 10.339588918677391,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.016725513671669204,
      "loss": 0.7735,
      "num_input_tokens_seen": 40310960,
      "step": 69420
    },
    {
      "epoch": 10.34033363121835,
      "grad_norm": 0.048828125,
      "learning_rate": 0.016723576946944898,
      "loss": 0.7717,
      "num_input_tokens_seen": 40313744,
      "step": 69425
    },
    {
      "epoch": 10.341078343759309,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.016721640193102513,
      "loss": 0.7977,
      "num_input_tokens_seen": 40316976,
      "step": 69430
    },
    {
      "epoch": 10.341823056300267,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.01671970341017477,
      "loss": 0.8091,
      "num_input_tokens_seen": 40319664,
      "step": 69435
    },
    {
      "epoch": 10.342567768841228,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01671776659819439,
      "loss": 0.8043,
      "num_input_tokens_seen": 40322768,
      "step": 69440
    },
    {
      "epoch": 10.343312481382187,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.016715829757194076,
      "loss": 0.8061,
      "num_input_tokens_seen": 40325680,
      "step": 69445
    },
    {
      "epoch": 10.344057193923145,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.01671389288720657,
      "loss": 0.806,
      "num_input_tokens_seen": 40328720,
      "step": 69450
    },
    {
      "epoch": 10.344801906464104,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.016711955988264582,
      "loss": 0.7886,
      "num_input_tokens_seen": 40331408,
      "step": 69455
    },
    {
      "epoch": 10.345546619005065,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.01671001906040084,
      "loss": 0.7881,
      "num_input_tokens_seen": 40334384,
      "step": 69460
    },
    {
      "epoch": 10.346291331546023,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.016708082103648065,
      "loss": 0.7799,
      "num_input_tokens_seen": 40337104,
      "step": 69465
    },
    {
      "epoch": 10.347036044086982,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.016706145118038977,
      "loss": 0.8057,
      "num_input_tokens_seen": 40339888,
      "step": 69470
    },
    {
      "epoch": 10.347780756627941,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.016704208103606302,
      "loss": 0.7794,
      "num_input_tokens_seen": 40342992,
      "step": 69475
    },
    {
      "epoch": 10.348525469168901,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.016702271060382766,
      "loss": 0.7744,
      "num_input_tokens_seen": 40345840,
      "step": 69480
    },
    {
      "epoch": 10.34927018170986,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.016700333988401084,
      "loss": 0.8245,
      "num_input_tokens_seen": 40348880,
      "step": 69485
    },
    {
      "epoch": 10.350014894250819,
      "grad_norm": 0.03515625,
      "learning_rate": 0.016698396887693992,
      "loss": 0.8011,
      "num_input_tokens_seen": 40351664,
      "step": 69490
    },
    {
      "epoch": 10.350759606791778,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.01669645975829421,
      "loss": 0.8089,
      "num_input_tokens_seen": 40354576,
      "step": 69495
    },
    {
      "epoch": 10.351504319332738,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.016694522600234466,
      "loss": 0.8223,
      "num_input_tokens_seen": 40357296,
      "step": 69500
    },
    {
      "epoch": 10.352249031873697,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.01669258541354748,
      "loss": 0.7925,
      "num_input_tokens_seen": 40360624,
      "step": 69505
    },
    {
      "epoch": 10.352993744414656,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.01669064819826599,
      "loss": 0.8166,
      "num_input_tokens_seen": 40363664,
      "step": 69510
    },
    {
      "epoch": 10.353738456955615,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.01668871095442271,
      "loss": 0.8176,
      "num_input_tokens_seen": 40366416,
      "step": 69515
    },
    {
      "epoch": 10.354483169496575,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.016686773682050383,
      "loss": 0.7807,
      "num_input_tokens_seen": 40369168,
      "step": 69520
    },
    {
      "epoch": 10.355227882037534,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.016684836381181723,
      "loss": 0.8332,
      "num_input_tokens_seen": 40371984,
      "step": 69525
    },
    {
      "epoch": 10.355972594578493,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.016682899051849467,
      "loss": 0.812,
      "num_input_tokens_seen": 40374768,
      "step": 69530
    },
    {
      "epoch": 10.356717307119451,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.016680961694086338,
      "loss": 0.8262,
      "num_input_tokens_seen": 40377648,
      "step": 69535
    },
    {
      "epoch": 10.357462019660412,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.016679024307925075,
      "loss": 0.7959,
      "num_input_tokens_seen": 40380144,
      "step": 69540
    },
    {
      "epoch": 10.35820673220137,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0166770868933984,
      "loss": 0.8045,
      "num_input_tokens_seen": 40383152,
      "step": 69545
    },
    {
      "epoch": 10.35895144474233,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.016675149450539047,
      "loss": 0.8053,
      "num_input_tokens_seen": 40386064,
      "step": 69550
    },
    {
      "epoch": 10.359696157283288,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.016673211979379744,
      "loss": 0.8315,
      "num_input_tokens_seen": 40389136,
      "step": 69555
    },
    {
      "epoch": 10.360440869824249,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.01667127447995323,
      "loss": 0.8082,
      "num_input_tokens_seen": 40391856,
      "step": 69560
    },
    {
      "epoch": 10.361185582365207,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01666933695229223,
      "loss": 0.7775,
      "num_input_tokens_seen": 40394832,
      "step": 69565
    },
    {
      "epoch": 10.361930294906166,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.016667399396429482,
      "loss": 0.7993,
      "num_input_tokens_seen": 40397776,
      "step": 69570
    },
    {
      "epoch": 10.362675007447125,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.01666546181239771,
      "loss": 0.8132,
      "num_input_tokens_seen": 40400720,
      "step": 69575
    },
    {
      "epoch": 10.363419719988084,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.016663524200229658,
      "loss": 0.799,
      "num_input_tokens_seen": 40403472,
      "step": 69580
    },
    {
      "epoch": 10.364164432529044,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.016661586559958055,
      "loss": 0.7963,
      "num_input_tokens_seen": 40406160,
      "step": 69585
    },
    {
      "epoch": 10.364909145070003,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01665964889161563,
      "loss": 0.7795,
      "num_input_tokens_seen": 40409008,
      "step": 69590
    },
    {
      "epoch": 10.365653857610962,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.016657711195235125,
      "loss": 0.795,
      "num_input_tokens_seen": 40411856,
      "step": 69595
    },
    {
      "epoch": 10.36639857015192,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.016655773470849282,
      "loss": 0.7845,
      "num_input_tokens_seen": 40414928,
      "step": 69600
    },
    {
      "epoch": 10.367143282692881,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.016653835718490828,
      "loss": 0.7835,
      "num_input_tokens_seen": 40418160,
      "step": 69605
    },
    {
      "epoch": 10.36788799523384,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.016651897938192495,
      "loss": 0.7881,
      "num_input_tokens_seen": 40421392,
      "step": 69610
    },
    {
      "epoch": 10.368632707774799,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.016649960129987026,
      "loss": 0.8003,
      "num_input_tokens_seen": 40424496,
      "step": 69615
    },
    {
      "epoch": 10.369377420315757,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.016648022293907157,
      "loss": 0.7985,
      "num_input_tokens_seen": 40427056,
      "step": 69620
    },
    {
      "epoch": 10.370122132856718,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.016646084429985632,
      "loss": 0.8122,
      "num_input_tokens_seen": 40430000,
      "step": 69625
    },
    {
      "epoch": 10.370866845397677,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01664414653825518,
      "loss": 0.7835,
      "num_input_tokens_seen": 40432976,
      "step": 69630
    },
    {
      "epoch": 10.371611557938635,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.01664220861874854,
      "loss": 0.7836,
      "num_input_tokens_seen": 40436016,
      "step": 69635
    },
    {
      "epoch": 10.372356270479594,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.01664027067149846,
      "loss": 0.7934,
      "num_input_tokens_seen": 40438992,
      "step": 69640
    },
    {
      "epoch": 10.373100983020555,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.01663833269653767,
      "loss": 0.8012,
      "num_input_tokens_seen": 40441872,
      "step": 69645
    },
    {
      "epoch": 10.373845695561513,
      "grad_norm": 0.068359375,
      "learning_rate": 0.016636394693898916,
      "loss": 0.7596,
      "num_input_tokens_seen": 40444752,
      "step": 69650
    },
    {
      "epoch": 10.374590408102472,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.016634456663614936,
      "loss": 0.7943,
      "num_input_tokens_seen": 40447792,
      "step": 69655
    },
    {
      "epoch": 10.375335120643431,
      "grad_norm": 0.044921875,
      "learning_rate": 0.01663251860571847,
      "loss": 0.824,
      "num_input_tokens_seen": 40450480,
      "step": 69660
    },
    {
      "epoch": 10.376079833184392,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.01663058052024227,
      "loss": 0.8127,
      "num_input_tokens_seen": 40453456,
      "step": 69665
    },
    {
      "epoch": 10.37682454572535,
      "grad_norm": 0.021484375,
      "learning_rate": 0.01662864240721906,
      "loss": 0.7645,
      "num_input_tokens_seen": 40456432,
      "step": 69670
    },
    {
      "epoch": 10.377569258266309,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.016626704266681593,
      "loss": 0.8354,
      "num_input_tokens_seen": 40459728,
      "step": 69675
    },
    {
      "epoch": 10.378313970807268,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.016624766098662612,
      "loss": 0.7927,
      "num_input_tokens_seen": 40462640,
      "step": 69680
    },
    {
      "epoch": 10.379058683348228,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.016622827903194862,
      "loss": 0.8011,
      "num_input_tokens_seen": 40465200,
      "step": 69685
    },
    {
      "epoch": 10.379803395889187,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.016620889680311085,
      "loss": 0.7942,
      "num_input_tokens_seen": 40468336,
      "step": 69690
    },
    {
      "epoch": 10.380548108430146,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.016618951430044018,
      "loss": 0.7922,
      "num_input_tokens_seen": 40470992,
      "step": 69695
    },
    {
      "epoch": 10.381292820971105,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.016617013152426415,
      "loss": 0.7773,
      "num_input_tokens_seen": 40474032,
      "step": 69700
    },
    {
      "epoch": 10.382037533512065,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.01661507484749102,
      "loss": 0.7879,
      "num_input_tokens_seen": 40476688,
      "step": 69705
    },
    {
      "epoch": 10.382782246053024,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.016613136515270573,
      "loss": 0.7995,
      "num_input_tokens_seen": 40479216,
      "step": 69710
    },
    {
      "epoch": 10.383526958593983,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.016611198155797827,
      "loss": 0.8082,
      "num_input_tokens_seen": 40481872,
      "step": 69715
    },
    {
      "epoch": 10.384271671134941,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.01660925976910552,
      "loss": 0.8057,
      "num_input_tokens_seen": 40484528,
      "step": 69720
    },
    {
      "epoch": 10.3850163836759,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.016607321355226416,
      "loss": 0.8152,
      "num_input_tokens_seen": 40487312,
      "step": 69725
    },
    {
      "epoch": 10.38576109621686,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.016605382914193244,
      "loss": 0.8156,
      "num_input_tokens_seen": 40490000,
      "step": 69730
    },
    {
      "epoch": 10.38650580875782,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.01660344444603876,
      "loss": 0.7888,
      "num_input_tokens_seen": 40493136,
      "step": 69735
    },
    {
      "epoch": 10.387250521298778,
      "grad_norm": 0.0703125,
      "learning_rate": 0.016601505950795712,
      "loss": 0.7911,
      "num_input_tokens_seen": 40496048,
      "step": 69740
    },
    {
      "epoch": 10.387995233839739,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.016599567428496845,
      "loss": 0.8054,
      "num_input_tokens_seen": 40499120,
      "step": 69745
    },
    {
      "epoch": 10.388739946380698,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.01659762887917492,
      "loss": 0.7998,
      "num_input_tokens_seen": 40502160,
      "step": 69750
    },
    {
      "epoch": 10.389484658921656,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.016595690302862672,
      "loss": 0.817,
      "num_input_tokens_seen": 40505168,
      "step": 69755
    },
    {
      "epoch": 10.390229371462615,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.01659375169959286,
      "loss": 0.8185,
      "num_input_tokens_seen": 40507728,
      "step": 69760
    },
    {
      "epoch": 10.390974084003574,
      "grad_norm": 0.04296875,
      "learning_rate": 0.016591813069398234,
      "loss": 0.7766,
      "num_input_tokens_seen": 40510896,
      "step": 69765
    },
    {
      "epoch": 10.391718796544534,
      "grad_norm": 0.0390625,
      "learning_rate": 0.016589874412311547,
      "loss": 0.8076,
      "num_input_tokens_seen": 40513456,
      "step": 69770
    },
    {
      "epoch": 10.392463509085493,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.016587935728365543,
      "loss": 0.8019,
      "num_input_tokens_seen": 40516240,
      "step": 69775
    },
    {
      "epoch": 10.393208221626452,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01658599701759298,
      "loss": 0.781,
      "num_input_tokens_seen": 40519120,
      "step": 69780
    },
    {
      "epoch": 10.39395293416741,
      "grad_norm": 0.05078125,
      "learning_rate": 0.016584058280026605,
      "loss": 0.7621,
      "num_input_tokens_seen": 40522128,
      "step": 69785
    },
    {
      "epoch": 10.394697646708371,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.016582119515699184,
      "loss": 0.8091,
      "num_input_tokens_seen": 40524880,
      "step": 69790
    },
    {
      "epoch": 10.39544235924933,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.016580180724643456,
      "loss": 0.7804,
      "num_input_tokens_seen": 40527600,
      "step": 69795
    },
    {
      "epoch": 10.396187071790289,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.016578241906892183,
      "loss": 0.818,
      "num_input_tokens_seen": 40530704,
      "step": 69800
    },
    {
      "epoch": 10.396931784331247,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.01657630306247812,
      "loss": 0.8034,
      "num_input_tokens_seen": 40533648,
      "step": 69805
    },
    {
      "epoch": 10.397676496872208,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.016574364191434012,
      "loss": 0.8024,
      "num_input_tokens_seen": 40536528,
      "step": 69810
    },
    {
      "epoch": 10.398421209413167,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.016572425293792624,
      "loss": 0.8139,
      "num_input_tokens_seen": 40539504,
      "step": 69815
    },
    {
      "epoch": 10.399165921954125,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.016570486369586707,
      "loss": 0.7856,
      "num_input_tokens_seen": 40542416,
      "step": 69820
    },
    {
      "epoch": 10.399910634495084,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.01656854741884902,
      "loss": 0.8052,
      "num_input_tokens_seen": 40545392,
      "step": 69825
    },
    {
      "epoch": 10.400655347036045,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.016566608441612323,
      "loss": 0.8261,
      "num_input_tokens_seen": 40548432,
      "step": 69830
    },
    {
      "epoch": 10.401400059577004,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.016564669437909362,
      "loss": 0.7999,
      "num_input_tokens_seen": 40551696,
      "step": 69835
    },
    {
      "epoch": 10.402144772117962,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.016562730407772903,
      "loss": 0.8129,
      "num_input_tokens_seen": 40554576,
      "step": 69840
    },
    {
      "epoch": 10.402889484658921,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.01656079135123571,
      "loss": 1.0187,
      "num_input_tokens_seen": 40557552,
      "step": 69845
    },
    {
      "epoch": 10.403634197199882,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.016558852268330523,
      "loss": 0.8113,
      "num_input_tokens_seen": 40560464,
      "step": 69850
    },
    {
      "epoch": 10.40437890974084,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.016556913159090116,
      "loss": 0.8073,
      "num_input_tokens_seen": 40563280,
      "step": 69855
    },
    {
      "epoch": 10.405123622281799,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.01655497402354724,
      "loss": 0.8178,
      "num_input_tokens_seen": 40566128,
      "step": 69860
    },
    {
      "epoch": 10.405868334822758,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.016553034861734664,
      "loss": 0.8049,
      "num_input_tokens_seen": 40568816,
      "step": 69865
    },
    {
      "epoch": 10.406613047363718,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.016551095673685137,
      "loss": 0.8073,
      "num_input_tokens_seen": 40571568,
      "step": 69870
    },
    {
      "epoch": 10.407357759904677,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.016549156459431426,
      "loss": 0.7954,
      "num_input_tokens_seen": 40574128,
      "step": 69875
    },
    {
      "epoch": 10.408102472445636,
      "grad_norm": 0.041015625,
      "learning_rate": 0.016547217219006292,
      "loss": 0.8024,
      "num_input_tokens_seen": 40577072,
      "step": 69880
    },
    {
      "epoch": 10.408847184986595,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.016545277952442497,
      "loss": 0.7997,
      "num_input_tokens_seen": 40580528,
      "step": 69885
    },
    {
      "epoch": 10.409591897527555,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0165433386597728,
      "loss": 0.8176,
      "num_input_tokens_seen": 40583408,
      "step": 69890
    },
    {
      "epoch": 10.410336610068514,
      "grad_norm": 0.04296875,
      "learning_rate": 0.016541399341029966,
      "loss": 0.7999,
      "num_input_tokens_seen": 40586384,
      "step": 69895
    },
    {
      "epoch": 10.411081322609473,
      "grad_norm": 0.033203125,
      "learning_rate": 0.016539459996246758,
      "loss": 0.787,
      "num_input_tokens_seen": 40589264,
      "step": 69900
    },
    {
      "epoch": 10.411826035150431,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.016537520625455937,
      "loss": 0.8051,
      "num_input_tokens_seen": 40592016,
      "step": 69905
    },
    {
      "epoch": 10.41257074769139,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.016535581228690265,
      "loss": 0.7865,
      "num_input_tokens_seen": 40594512,
      "step": 69910
    },
    {
      "epoch": 10.41331546023235,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.01653364180598251,
      "loss": 0.8091,
      "num_input_tokens_seen": 40597360,
      "step": 69915
    },
    {
      "epoch": 10.41406017277331,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.016531702357365435,
      "loss": 0.8005,
      "num_input_tokens_seen": 40600048,
      "step": 69920
    },
    {
      "epoch": 10.414804885314268,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.016529762882871807,
      "loss": 0.7837,
      "num_input_tokens_seen": 40603216,
      "step": 69925
    },
    {
      "epoch": 10.415549597855227,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.01652782338253439,
      "loss": 0.801,
      "num_input_tokens_seen": 40606064,
      "step": 69930
    },
    {
      "epoch": 10.416294310396188,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.016525883856385953,
      "loss": 0.7967,
      "num_input_tokens_seen": 40608880,
      "step": 69935
    },
    {
      "epoch": 10.417039022937146,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.01652394430445925,
      "loss": 0.7751,
      "num_input_tokens_seen": 40611472,
      "step": 69940
    },
    {
      "epoch": 10.417783735478105,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.016522004726787064,
      "loss": 0.8267,
      "num_input_tokens_seen": 40614128,
      "step": 69945
    },
    {
      "epoch": 10.418528448019064,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.016520065123402154,
      "loss": 0.794,
      "num_input_tokens_seen": 40616848,
      "step": 69950
    },
    {
      "epoch": 10.419273160560024,
      "grad_norm": 0.072265625,
      "learning_rate": 0.016518125494337284,
      "loss": 0.8157,
      "num_input_tokens_seen": 40619728,
      "step": 69955
    },
    {
      "epoch": 10.420017873100983,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.016516185839625233,
      "loss": 0.8319,
      "num_input_tokens_seen": 40622608,
      "step": 69960
    },
    {
      "epoch": 10.420762585641942,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.01651424615929876,
      "loss": 0.7923,
      "num_input_tokens_seen": 40625648,
      "step": 69965
    },
    {
      "epoch": 10.4215072981829,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01651230645339064,
      "loss": 0.8062,
      "num_input_tokens_seen": 40628688,
      "step": 69970
    },
    {
      "epoch": 10.422252010723861,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.016510366721933636,
      "loss": 0.7873,
      "num_input_tokens_seen": 40631632,
      "step": 69975
    },
    {
      "epoch": 10.42299672326482,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.016508426964960524,
      "loss": 0.7973,
      "num_input_tokens_seen": 40634544,
      "step": 69980
    },
    {
      "epoch": 10.423741435805779,
      "grad_norm": 0.03125,
      "learning_rate": 0.016506487182504068,
      "loss": 0.8049,
      "num_input_tokens_seen": 40637520,
      "step": 69985
    },
    {
      "epoch": 10.424486148346737,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.016504547374597046,
      "loss": 0.8101,
      "num_input_tokens_seen": 40640304,
      "step": 69990
    },
    {
      "epoch": 10.425230860887698,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.016502607541272224,
      "loss": 0.7849,
      "num_input_tokens_seen": 40643024,
      "step": 69995
    },
    {
      "epoch": 10.425975573428657,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.01650066768256237,
      "loss": 0.8092,
      "num_input_tokens_seen": 40645840,
      "step": 70000
    },
    {
      "epoch": 10.426720285969616,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.01649872779850027,
      "loss": 0.7934,
      "num_input_tokens_seen": 40648688,
      "step": 70005
    },
    {
      "epoch": 10.427464998510574,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.01649678788911868,
      "loss": 0.8036,
      "num_input_tokens_seen": 40651632,
      "step": 70010
    },
    {
      "epoch": 10.428209711051535,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.016494847954450383,
      "loss": 0.8073,
      "num_input_tokens_seen": 40654480,
      "step": 70015
    },
    {
      "epoch": 10.428954423592494,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.01649290799452815,
      "loss": 0.8018,
      "num_input_tokens_seen": 40657392,
      "step": 70020
    },
    {
      "epoch": 10.429699136133452,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.016490968009384748,
      "loss": 0.8184,
      "num_input_tokens_seen": 40660144,
      "step": 70025
    },
    {
      "epoch": 10.430443848674411,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.016489027999052965,
      "loss": 0.8124,
      "num_input_tokens_seen": 40662992,
      "step": 70030
    },
    {
      "epoch": 10.431188561215372,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.01648708796356556,
      "loss": 0.8163,
      "num_input_tokens_seen": 40665648,
      "step": 70035
    },
    {
      "epoch": 10.43193327375633,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.01648514790295532,
      "loss": 0.8036,
      "num_input_tokens_seen": 40668464,
      "step": 70040
    },
    {
      "epoch": 10.43267798629729,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.016483207817255017,
      "loss": 0.8077,
      "num_input_tokens_seen": 40671600,
      "step": 70045
    },
    {
      "epoch": 10.433422698838248,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01648126770649742,
      "loss": 0.8095,
      "num_input_tokens_seen": 40674416,
      "step": 70050
    },
    {
      "epoch": 10.434167411379208,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.016479327570715314,
      "loss": 0.816,
      "num_input_tokens_seen": 40677584,
      "step": 70055
    },
    {
      "epoch": 10.434912123920167,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01647738740994147,
      "loss": 0.7958,
      "num_input_tokens_seen": 40680432,
      "step": 70060
    },
    {
      "epoch": 10.435656836461126,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.01647544722420867,
      "loss": 0.7934,
      "num_input_tokens_seen": 40683120,
      "step": 70065
    },
    {
      "epoch": 10.436401549002085,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.016473507013549684,
      "loss": 0.7975,
      "num_input_tokens_seen": 40686032,
      "step": 70070
    },
    {
      "epoch": 10.437146261543045,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.0164715667779973,
      "loss": 0.8137,
      "num_input_tokens_seen": 40688912,
      "step": 70075
    },
    {
      "epoch": 10.437890974084004,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.016469626517584286,
      "loss": 0.7909,
      "num_input_tokens_seen": 40691824,
      "step": 70080
    },
    {
      "epoch": 10.438635686624963,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.01646768623234343,
      "loss": 0.8105,
      "num_input_tokens_seen": 40694992,
      "step": 70085
    },
    {
      "epoch": 10.439380399165922,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.016465745922307505,
      "loss": 0.8032,
      "num_input_tokens_seen": 40697936,
      "step": 70090
    },
    {
      "epoch": 10.44012511170688,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.016463805587509293,
      "loss": 0.8083,
      "num_input_tokens_seen": 40700816,
      "step": 70095
    },
    {
      "epoch": 10.44086982424784,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.016461865227981577,
      "loss": 0.793,
      "num_input_tokens_seen": 40703536,
      "step": 70100
    },
    {
      "epoch": 10.4416145367888,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.016459924843757127,
      "loss": 0.7929,
      "num_input_tokens_seen": 40706416,
      "step": 70105
    },
    {
      "epoch": 10.442359249329758,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.016457984434868733,
      "loss": 0.813,
      "num_input_tokens_seen": 40709168,
      "step": 70110
    },
    {
      "epoch": 10.443103961870717,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.016456044001349174,
      "loss": 0.7982,
      "num_input_tokens_seen": 40712080,
      "step": 70115
    },
    {
      "epoch": 10.443848674411678,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.01645410354323123,
      "loss": 0.8084,
      "num_input_tokens_seen": 40714864,
      "step": 70120
    },
    {
      "epoch": 10.444593386952636,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.016452163060547687,
      "loss": 0.8295,
      "num_input_tokens_seen": 40717936,
      "step": 70125
    },
    {
      "epoch": 10.445338099493595,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.016450222553331322,
      "loss": 0.7878,
      "num_input_tokens_seen": 40720752,
      "step": 70130
    },
    {
      "epoch": 10.446082812034554,
      "grad_norm": 0.0157470703125,
      "learning_rate": 0.016448282021614927,
      "loss": 0.788,
      "num_input_tokens_seen": 40723440,
      "step": 70135
    },
    {
      "epoch": 10.446827524575514,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.016446341465431273,
      "loss": 0.8083,
      "num_input_tokens_seen": 40726608,
      "step": 70140
    },
    {
      "epoch": 10.447572237116473,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01644440088481315,
      "loss": 0.8038,
      "num_input_tokens_seen": 40729488,
      "step": 70145
    },
    {
      "epoch": 10.448316949657432,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.016442460279793344,
      "loss": 0.8077,
      "num_input_tokens_seen": 40732208,
      "step": 70150
    },
    {
      "epoch": 10.44906166219839,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.016440519650404637,
      "loss": 0.7924,
      "num_input_tokens_seen": 40735216,
      "step": 70155
    },
    {
      "epoch": 10.449806374739351,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.016438578996679817,
      "loss": 0.7943,
      "num_input_tokens_seen": 40738192,
      "step": 70160
    },
    {
      "epoch": 10.45055108728031,
      "grad_norm": 0.01953125,
      "learning_rate": 0.01643663831865167,
      "loss": 0.8077,
      "num_input_tokens_seen": 40740944,
      "step": 70165
    },
    {
      "epoch": 10.451295799821269,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.016434697616352974,
      "loss": 0.7948,
      "num_input_tokens_seen": 40743632,
      "step": 70170
    },
    {
      "epoch": 10.452040512362228,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.01643275688981652,
      "loss": 0.8015,
      "num_input_tokens_seen": 40746544,
      "step": 70175
    },
    {
      "epoch": 10.452785224903188,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01643081613907509,
      "loss": 0.8093,
      "num_input_tokens_seen": 40749360,
      "step": 70180
    },
    {
      "epoch": 10.453529937444147,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.016428875364161478,
      "loss": 0.7992,
      "num_input_tokens_seen": 40752272,
      "step": 70185
    },
    {
      "epoch": 10.454274649985106,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.016426934565108475,
      "loss": 0.8005,
      "num_input_tokens_seen": 40755024,
      "step": 70190
    },
    {
      "epoch": 10.455019362526064,
      "grad_norm": 0.0164794921875,
      "learning_rate": 0.01642499374194886,
      "loss": 0.7978,
      "num_input_tokens_seen": 40758064,
      "step": 70195
    },
    {
      "epoch": 10.455764075067025,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.01642305289471542,
      "loss": 0.8106,
      "num_input_tokens_seen": 40760784,
      "step": 70200
    },
    {
      "epoch": 10.456508787607984,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.016421112023440952,
      "loss": 0.7973,
      "num_input_tokens_seen": 40763504,
      "step": 70205
    },
    {
      "epoch": 10.457253500148942,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01641917112815824,
      "loss": 0.7798,
      "num_input_tokens_seen": 40766224,
      "step": 70210
    },
    {
      "epoch": 10.457998212689901,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.016417230208900072,
      "loss": 0.8003,
      "num_input_tokens_seen": 40769200,
      "step": 70215
    },
    {
      "epoch": 10.458742925230862,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.01641528926569924,
      "loss": 0.7941,
      "num_input_tokens_seen": 40772240,
      "step": 70220
    },
    {
      "epoch": 10.45948763777182,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.016413348298588534,
      "loss": 0.7934,
      "num_input_tokens_seen": 40775344,
      "step": 70225
    },
    {
      "epoch": 10.46023235031278,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01641140730760075,
      "loss": 0.7858,
      "num_input_tokens_seen": 40778128,
      "step": 70230
    },
    {
      "epoch": 10.460977062853738,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01640946629276867,
      "loss": 0.8001,
      "num_input_tokens_seen": 40780976,
      "step": 70235
    },
    {
      "epoch": 10.461721775394698,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.016407525254125088,
      "loss": 0.7796,
      "num_input_tokens_seen": 40783824,
      "step": 70240
    },
    {
      "epoch": 10.462466487935657,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.016405584191702804,
      "loss": 0.7893,
      "num_input_tokens_seen": 40786416,
      "step": 70245
    },
    {
      "epoch": 10.463211200476616,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.016403643105534597,
      "loss": 0.7772,
      "num_input_tokens_seen": 40789200,
      "step": 70250
    },
    {
      "epoch": 10.463955913017575,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.01640170199565327,
      "loss": 0.7868,
      "num_input_tokens_seen": 40792240,
      "step": 70255
    },
    {
      "epoch": 10.464700625558535,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.01639976086209161,
      "loss": 0.7738,
      "num_input_tokens_seen": 40795088,
      "step": 70260
    },
    {
      "epoch": 10.465445338099494,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.016397819704882417,
      "loss": 0.783,
      "num_input_tokens_seen": 40798064,
      "step": 70265
    },
    {
      "epoch": 10.466190050640453,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.016395878524058473,
      "loss": 0.7923,
      "num_input_tokens_seen": 40801104,
      "step": 70270
    },
    {
      "epoch": 10.466934763181412,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.016393937319652586,
      "loss": 0.7728,
      "num_input_tokens_seen": 40803600,
      "step": 70275
    },
    {
      "epoch": 10.46767947572237,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01639199609169754,
      "loss": 0.7849,
      "num_input_tokens_seen": 40806320,
      "step": 70280
    },
    {
      "epoch": 10.46842418826333,
      "grad_norm": 0.060546875,
      "learning_rate": 0.01639005484022614,
      "loss": 0.7951,
      "num_input_tokens_seen": 40809200,
      "step": 70285
    },
    {
      "epoch": 10.46916890080429,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.016388113565271176,
      "loss": 0.7932,
      "num_input_tokens_seen": 40812144,
      "step": 70290
    },
    {
      "epoch": 10.469913613345248,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.016386172266865444,
      "loss": 0.8129,
      "num_input_tokens_seen": 40815152,
      "step": 70295
    },
    {
      "epoch": 10.470658325886207,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.01638423094504174,
      "loss": 0.8525,
      "num_input_tokens_seen": 40817936,
      "step": 70300
    },
    {
      "epoch": 10.471403038427168,
      "grad_norm": 0.05078125,
      "learning_rate": 0.016382289599832858,
      "loss": 0.8147,
      "num_input_tokens_seen": 40820688,
      "step": 70305
    },
    {
      "epoch": 10.472147750968126,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.0163803482312716,
      "loss": 0.8052,
      "num_input_tokens_seen": 40823728,
      "step": 70310
    },
    {
      "epoch": 10.472892463509085,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.01637840683939076,
      "loss": 0.8115,
      "num_input_tokens_seen": 40826544,
      "step": 70315
    },
    {
      "epoch": 10.473637176050044,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01637646542422314,
      "loss": 0.7935,
      "num_input_tokens_seen": 40829488,
      "step": 70320
    },
    {
      "epoch": 10.474381888591004,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.016374523985801536,
      "loss": 0.8061,
      "num_input_tokens_seen": 40832368,
      "step": 70325
    },
    {
      "epoch": 10.475126601131963,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.016372582524158748,
      "loss": 0.7988,
      "num_input_tokens_seen": 40835312,
      "step": 70330
    },
    {
      "epoch": 10.475871313672922,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.016370641039327567,
      "loss": 0.8127,
      "num_input_tokens_seen": 40838064,
      "step": 70335
    },
    {
      "epoch": 10.47661602621388,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.0163686995313408,
      "loss": 0.7986,
      "num_input_tokens_seen": 40840816,
      "step": 70340
    },
    {
      "epoch": 10.477360738754841,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.016366758000231248,
      "loss": 0.7849,
      "num_input_tokens_seen": 40843472,
      "step": 70345
    },
    {
      "epoch": 10.4781054512958,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.01636481644603171,
      "loss": 0.8012,
      "num_input_tokens_seen": 40846320,
      "step": 70350
    },
    {
      "epoch": 10.478850163836759,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.016362874868774982,
      "loss": 0.8294,
      "num_input_tokens_seen": 40849296,
      "step": 70355
    },
    {
      "epoch": 10.479594876377718,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01636093326849387,
      "loss": 0.7902,
      "num_input_tokens_seen": 40852048,
      "step": 70360
    },
    {
      "epoch": 10.480339588918678,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.016358991645221175,
      "loss": 0.8299,
      "num_input_tokens_seen": 40855184,
      "step": 70365
    },
    {
      "epoch": 10.481084301459637,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.016357049998989695,
      "loss": 0.7784,
      "num_input_tokens_seen": 40858032,
      "step": 70370
    },
    {
      "epoch": 10.481829014000596,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.016355108329832235,
      "loss": 0.8243,
      "num_input_tokens_seen": 40861008,
      "step": 70375
    },
    {
      "epoch": 10.482573726541554,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.016353166637781597,
      "loss": 0.8195,
      "num_input_tokens_seen": 40863760,
      "step": 70380
    },
    {
      "epoch": 10.483318439082515,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.016351224922870583,
      "loss": 0.7924,
      "num_input_tokens_seen": 40866544,
      "step": 70385
    },
    {
      "epoch": 10.484063151623474,
      "grad_norm": 0.1875,
      "learning_rate": 0.016349283185132,
      "loss": 0.8266,
      "num_input_tokens_seen": 40869232,
      "step": 70390
    },
    {
      "epoch": 10.484807864164432,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.01634734142459865,
      "loss": 0.8159,
      "num_input_tokens_seen": 40872112,
      "step": 70395
    },
    {
      "epoch": 10.485552576705391,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.01634539964130333,
      "loss": 0.8044,
      "num_input_tokens_seen": 40874640,
      "step": 70400
    },
    {
      "epoch": 10.486297289246352,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.016343457835278852,
      "loss": 0.7901,
      "num_input_tokens_seen": 40877776,
      "step": 70405
    },
    {
      "epoch": 10.48704200178731,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.016341516006558025,
      "loss": 0.8103,
      "num_input_tokens_seen": 40880304,
      "step": 70410
    },
    {
      "epoch": 10.48778671432827,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.016339574155173644,
      "loss": 0.7938,
      "num_input_tokens_seen": 40883248,
      "step": 70415
    },
    {
      "epoch": 10.488531426869228,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.016337632281158518,
      "loss": 0.807,
      "num_input_tokens_seen": 40886576,
      "step": 70420
    },
    {
      "epoch": 10.489276139410187,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.016335690384545454,
      "loss": 0.8164,
      "num_input_tokens_seen": 40889680,
      "step": 70425
    },
    {
      "epoch": 10.490020851951147,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.016333748465367264,
      "loss": 0.8111,
      "num_input_tokens_seen": 40892272,
      "step": 70430
    },
    {
      "epoch": 10.490765564492106,
      "grad_norm": 0.01953125,
      "learning_rate": 0.016331806523656747,
      "loss": 0.7921,
      "num_input_tokens_seen": 40894928,
      "step": 70435
    },
    {
      "epoch": 10.491510277033065,
      "grad_norm": 0.01611328125,
      "learning_rate": 0.016329864559446707,
      "loss": 0.7913,
      "num_input_tokens_seen": 40897776,
      "step": 70440
    },
    {
      "epoch": 10.492254989574024,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.01632792257276996,
      "loss": 0.7943,
      "num_input_tokens_seen": 40900560,
      "step": 70445
    },
    {
      "epoch": 10.492999702114984,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.016325980563659317,
      "loss": 0.7914,
      "num_input_tokens_seen": 40903728,
      "step": 70450
    },
    {
      "epoch": 10.493744414655943,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.01632403853214757,
      "loss": 0.7603,
      "num_input_tokens_seen": 40906512,
      "step": 70455
    },
    {
      "epoch": 10.494489127196902,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.01632209647826754,
      "loss": 0.8183,
      "num_input_tokens_seen": 40909680,
      "step": 70460
    },
    {
      "epoch": 10.49523383973786,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.01632015440205204,
      "loss": 0.816,
      "num_input_tokens_seen": 40912528,
      "step": 70465
    },
    {
      "epoch": 10.495978552278821,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01631821230353387,
      "loss": 0.7842,
      "num_input_tokens_seen": 40915728,
      "step": 70470
    },
    {
      "epoch": 10.49672326481978,
      "grad_norm": 0.0390625,
      "learning_rate": 0.01631627018274584,
      "loss": 0.832,
      "num_input_tokens_seen": 40918416,
      "step": 70475
    },
    {
      "epoch": 10.497467977360738,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.016314328039720764,
      "loss": 0.7838,
      "num_input_tokens_seen": 40921680,
      "step": 70480
    },
    {
      "epoch": 10.498212689901697,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.016312385874491454,
      "loss": 0.7936,
      "num_input_tokens_seen": 40924592,
      "step": 70485
    },
    {
      "epoch": 10.498957402442658,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.01631044368709072,
      "loss": 0.7556,
      "num_input_tokens_seen": 40927376,
      "step": 70490
    },
    {
      "epoch": 10.499702114983616,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01630850147755137,
      "loss": 0.8158,
      "num_input_tokens_seen": 40930576,
      "step": 70495
    },
    {
      "epoch": 10.500446827524575,
      "grad_norm": 0.03125,
      "learning_rate": 0.016306559245906217,
      "loss": 0.803,
      "num_input_tokens_seen": 40933264,
      "step": 70500
    },
    {
      "epoch": 10.501191540065534,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.016304616992188072,
      "loss": 0.8372,
      "num_input_tokens_seen": 40936208,
      "step": 70505
    },
    {
      "epoch": 10.501936252606495,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01630267471642975,
      "loss": 0.8169,
      "num_input_tokens_seen": 40938960,
      "step": 70510
    },
    {
      "epoch": 10.502680965147453,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.01630073241866406,
      "loss": 0.8335,
      "num_input_tokens_seen": 40941744,
      "step": 70515
    },
    {
      "epoch": 10.503425677688412,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.01629879009892382,
      "loss": 0.8198,
      "num_input_tokens_seen": 40944560,
      "step": 70520
    },
    {
      "epoch": 10.50417039022937,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.016296847757241842,
      "loss": 0.8019,
      "num_input_tokens_seen": 40947280,
      "step": 70525
    },
    {
      "epoch": 10.504915102770331,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.016294905393650944,
      "loss": 0.8057,
      "num_input_tokens_seen": 40949872,
      "step": 70530
    },
    {
      "epoch": 10.50565981531129,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.016292963008183926,
      "loss": 0.805,
      "num_input_tokens_seen": 40952912,
      "step": 70535
    },
    {
      "epoch": 10.506404527852249,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.016291020600873618,
      "loss": 0.8054,
      "num_input_tokens_seen": 40955728,
      "step": 70540
    },
    {
      "epoch": 10.507149240393208,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.016289078171752826,
      "loss": 0.8337,
      "num_input_tokens_seen": 40958736,
      "step": 70545
    },
    {
      "epoch": 10.507893952934168,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.01628713572085437,
      "loss": 0.7993,
      "num_input_tokens_seen": 40961328,
      "step": 70550
    },
    {
      "epoch": 10.508638665475127,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.016285193248211063,
      "loss": 0.7956,
      "num_input_tokens_seen": 40964016,
      "step": 70555
    },
    {
      "epoch": 10.509383378016086,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.016283250753855723,
      "loss": 0.8068,
      "num_input_tokens_seen": 40966928,
      "step": 70560
    },
    {
      "epoch": 10.510128090557044,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.01628130823782117,
      "loss": 0.7972,
      "num_input_tokens_seen": 40969904,
      "step": 70565
    },
    {
      "epoch": 10.510872803098005,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.016279365700140214,
      "loss": 0.8076,
      "num_input_tokens_seen": 40972976,
      "step": 70570
    },
    {
      "epoch": 10.511617515638964,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.01627742314084567,
      "loss": 0.7913,
      "num_input_tokens_seen": 40975952,
      "step": 70575
    },
    {
      "epoch": 10.512362228179922,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.016275480559970364,
      "loss": 0.8057,
      "num_input_tokens_seen": 40978832,
      "step": 70580
    },
    {
      "epoch": 10.513106940720881,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01627353795754711,
      "loss": 0.7958,
      "num_input_tokens_seen": 40981744,
      "step": 70585
    },
    {
      "epoch": 10.513851653261842,
      "grad_norm": 0.018310546875,
      "learning_rate": 0.016271595333608727,
      "loss": 0.8071,
      "num_input_tokens_seen": 40984496,
      "step": 70590
    },
    {
      "epoch": 10.5145963658028,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.016269652688188033,
      "loss": 0.796,
      "num_input_tokens_seen": 40987248,
      "step": 70595
    },
    {
      "epoch": 10.51534107834376,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.016267710021317848,
      "loss": 0.8103,
      "num_input_tokens_seen": 40990032,
      "step": 70600
    },
    {
      "epoch": 10.516085790884718,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.01626576733303099,
      "loss": 0.7972,
      "num_input_tokens_seen": 40992976,
      "step": 70605
    },
    {
      "epoch": 10.516830503425677,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.01626382462336028,
      "loss": 0.8067,
      "num_input_tokens_seen": 40995952,
      "step": 70610
    },
    {
      "epoch": 10.517575215966637,
      "grad_norm": 0.044921875,
      "learning_rate": 0.016261881892338535,
      "loss": 0.8032,
      "num_input_tokens_seen": 40998608,
      "step": 70615
    },
    {
      "epoch": 10.518319928507596,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.01625993913999858,
      "loss": 0.7851,
      "num_input_tokens_seen": 41001488,
      "step": 70620
    },
    {
      "epoch": 10.519064641048555,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.01625799636637323,
      "loss": 0.7986,
      "num_input_tokens_seen": 41004144,
      "step": 70625
    },
    {
      "epoch": 10.519809353589514,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.016256053571495316,
      "loss": 0.7963,
      "num_input_tokens_seen": 41006960,
      "step": 70630
    },
    {
      "epoch": 10.520554066130474,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.01625411075539765,
      "loss": 0.8034,
      "num_input_tokens_seen": 41009840,
      "step": 70635
    },
    {
      "epoch": 10.521298778671433,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01625216791811306,
      "loss": 0.7994,
      "num_input_tokens_seen": 41012688,
      "step": 70640
    },
    {
      "epoch": 10.522043491212392,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.016250225059674363,
      "loss": 0.7839,
      "num_input_tokens_seen": 41015472,
      "step": 70645
    },
    {
      "epoch": 10.52278820375335,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.016248282180114388,
      "loss": 0.8028,
      "num_input_tokens_seen": 41018384,
      "step": 70650
    },
    {
      "epoch": 10.523532916294311,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.01624633927946595,
      "loss": 0.8088,
      "num_input_tokens_seen": 41021616,
      "step": 70655
    },
    {
      "epoch": 10.52427762883527,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.01624439635776188,
      "loss": 0.8088,
      "num_input_tokens_seen": 41024432,
      "step": 70660
    },
    {
      "epoch": 10.525022341376228,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.016242453415034993,
      "loss": 0.8096,
      "num_input_tokens_seen": 41026928,
      "step": 70665
    },
    {
      "epoch": 10.525767053917187,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.016240510451318125,
      "loss": 0.8121,
      "num_input_tokens_seen": 41029744,
      "step": 70670
    },
    {
      "epoch": 10.526511766458148,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.016238567466644087,
      "loss": 0.8221,
      "num_input_tokens_seen": 41032528,
      "step": 70675
    },
    {
      "epoch": 10.527256478999107,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.016236624461045714,
      "loss": 0.8077,
      "num_input_tokens_seen": 41035344,
      "step": 70680
    },
    {
      "epoch": 10.528001191540065,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.016234681434555826,
      "loss": 0.8163,
      "num_input_tokens_seen": 41038128,
      "step": 70685
    },
    {
      "epoch": 10.528745904081024,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.016232738387207256,
      "loss": 0.8006,
      "num_input_tokens_seen": 41041008,
      "step": 70690
    },
    {
      "epoch": 10.529490616621985,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01623079531903282,
      "loss": 0.8117,
      "num_input_tokens_seen": 41043984,
      "step": 70695
    },
    {
      "epoch": 10.530235329162943,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.016228852230065344,
      "loss": 0.795,
      "num_input_tokens_seen": 41046960,
      "step": 70700
    },
    {
      "epoch": 10.530980041703902,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.01622690912033766,
      "loss": 0.7835,
      "num_input_tokens_seen": 41049872,
      "step": 70705
    },
    {
      "epoch": 10.53172475424486,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.016224965989882597,
      "loss": 0.8049,
      "num_input_tokens_seen": 41052592,
      "step": 70710
    },
    {
      "epoch": 10.532469466785821,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.016223022838732976,
      "loss": 0.7955,
      "num_input_tokens_seen": 41055952,
      "step": 70715
    },
    {
      "epoch": 10.53321417932678,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.01622107966692163,
      "loss": 0.8145,
      "num_input_tokens_seen": 41058992,
      "step": 70720
    },
    {
      "epoch": 10.533958891867739,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.01621913647448138,
      "loss": 0.8082,
      "num_input_tokens_seen": 41061712,
      "step": 70725
    },
    {
      "epoch": 10.534703604408698,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01621719326144506,
      "loss": 0.7861,
      "num_input_tokens_seen": 41064624,
      "step": 70730
    },
    {
      "epoch": 10.535448316949658,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0162152500278455,
      "loss": 0.802,
      "num_input_tokens_seen": 41067664,
      "step": 70735
    },
    {
      "epoch": 10.536193029490617,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.016213306773715522,
      "loss": 0.8236,
      "num_input_tokens_seen": 41070544,
      "step": 70740
    },
    {
      "epoch": 10.536937742031576,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.01621136349908796,
      "loss": 0.8063,
      "num_input_tokens_seen": 41074064,
      "step": 70745
    },
    {
      "epoch": 10.537682454572534,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.016209420203995642,
      "loss": 0.8029,
      "num_input_tokens_seen": 41076656,
      "step": 70750
    },
    {
      "epoch": 10.538427167113493,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0162074768884714,
      "loss": 0.8033,
      "num_input_tokens_seen": 41079504,
      "step": 70755
    },
    {
      "epoch": 10.539171879654454,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.016205533552548067,
      "loss": 0.8143,
      "num_input_tokens_seen": 41082320,
      "step": 70760
    },
    {
      "epoch": 10.539916592195413,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.016203590196258465,
      "loss": 0.7886,
      "num_input_tokens_seen": 41085072,
      "step": 70765
    },
    {
      "epoch": 10.540661304736371,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.016201646819635433,
      "loss": 0.7965,
      "num_input_tokens_seen": 41087760,
      "step": 70770
    },
    {
      "epoch": 10.541406017277332,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.016199703422711795,
      "loss": 0.8106,
      "num_input_tokens_seen": 41090320,
      "step": 70775
    },
    {
      "epoch": 10.54215072981829,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.01619776000552039,
      "loss": 0.8103,
      "num_input_tokens_seen": 41093456,
      "step": 70780
    },
    {
      "epoch": 10.54289544235925,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.016195816568094046,
      "loss": 0.7931,
      "num_input_tokens_seen": 41096272,
      "step": 70785
    },
    {
      "epoch": 10.543640154900208,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.016193873110465596,
      "loss": 0.7919,
      "num_input_tokens_seen": 41099408,
      "step": 70790
    },
    {
      "epoch": 10.544384867441167,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.016191929632667877,
      "loss": 0.7753,
      "num_input_tokens_seen": 41102256,
      "step": 70795
    },
    {
      "epoch": 10.545129579982127,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.016189986134733714,
      "loss": 0.8304,
      "num_input_tokens_seen": 41105040,
      "step": 70800
    },
    {
      "epoch": 10.545874292523086,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.016188042616695944,
      "loss": 0.8026,
      "num_input_tokens_seen": 41107760,
      "step": 70805
    },
    {
      "epoch": 10.546619005064045,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.016186099078587406,
      "loss": 0.803,
      "num_input_tokens_seen": 41110608,
      "step": 70810
    },
    {
      "epoch": 10.547363717605004,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01618415552044093,
      "loss": 0.8088,
      "num_input_tokens_seen": 41113680,
      "step": 70815
    },
    {
      "epoch": 10.548108430145964,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.016182211942289346,
      "loss": 0.809,
      "num_input_tokens_seen": 41116720,
      "step": 70820
    },
    {
      "epoch": 10.548853142686923,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.016180268344165496,
      "loss": 0.7921,
      "num_input_tokens_seen": 41119600,
      "step": 70825
    },
    {
      "epoch": 10.549597855227882,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.016178324726102207,
      "loss": 0.792,
      "num_input_tokens_seen": 41122672,
      "step": 70830
    },
    {
      "epoch": 10.55034256776884,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01617638108813232,
      "loss": 0.7862,
      "num_input_tokens_seen": 41125840,
      "step": 70835
    },
    {
      "epoch": 10.551087280309801,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.016174437430288676,
      "loss": 0.8146,
      "num_input_tokens_seen": 41129136,
      "step": 70840
    },
    {
      "epoch": 10.55183199285076,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0161724937526041,
      "loss": 0.799,
      "num_input_tokens_seen": 41131664,
      "step": 70845
    },
    {
      "epoch": 10.552576705391719,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.016170550055111435,
      "loss": 0.7926,
      "num_input_tokens_seen": 41134256,
      "step": 70850
    },
    {
      "epoch": 10.553321417932677,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.01616860633784352,
      "loss": 0.8043,
      "num_input_tokens_seen": 41137200,
      "step": 70855
    },
    {
      "epoch": 10.554066130473638,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.016166662600833187,
      "loss": 0.8401,
      "num_input_tokens_seen": 41140016,
      "step": 70860
    },
    {
      "epoch": 10.554810843014597,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.01616471884411327,
      "loss": 0.7927,
      "num_input_tokens_seen": 41143344,
      "step": 70865
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.016162775067716614,
      "loss": 0.8111,
      "num_input_tokens_seen": 41146416,
      "step": 70870
    },
    {
      "epoch": 10.556300268096514,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.016160831271676057,
      "loss": 0.8289,
      "num_input_tokens_seen": 41149264,
      "step": 70875
    },
    {
      "epoch": 10.557044980637475,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.016158887456024434,
      "loss": 0.7987,
      "num_input_tokens_seen": 41152272,
      "step": 70880
    },
    {
      "epoch": 10.557789693178433,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.016156943620794586,
      "loss": 0.7899,
      "num_input_tokens_seen": 41154864,
      "step": 70885
    },
    {
      "epoch": 10.558534405719392,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.016154999766019355,
      "loss": 0.8144,
      "num_input_tokens_seen": 41157968,
      "step": 70890
    },
    {
      "epoch": 10.559279118260351,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.016153055891731577,
      "loss": 0.7916,
      "num_input_tokens_seen": 41160656,
      "step": 70895
    },
    {
      "epoch": 10.560023830801311,
      "grad_norm": 0.0177001953125,
      "learning_rate": 0.016151111997964086,
      "loss": 0.7972,
      "num_input_tokens_seen": 41163408,
      "step": 70900
    },
    {
      "epoch": 10.56076854334227,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.016149168084749727,
      "loss": 0.8072,
      "num_input_tokens_seen": 41166448,
      "step": 70905
    },
    {
      "epoch": 10.561513255883229,
      "grad_norm": 0.02734375,
      "learning_rate": 0.016147224152121345,
      "loss": 0.7848,
      "num_input_tokens_seen": 41169200,
      "step": 70910
    },
    {
      "epoch": 10.562257968424188,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.016145280200111772,
      "loss": 0.7918,
      "num_input_tokens_seen": 41172176,
      "step": 70915
    },
    {
      "epoch": 10.563002680965148,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.016143336228753857,
      "loss": 0.8226,
      "num_input_tokens_seen": 41175024,
      "step": 70920
    },
    {
      "epoch": 10.563747393506107,
      "grad_norm": 0.03125,
      "learning_rate": 0.01614139223808044,
      "loss": 0.7764,
      "num_input_tokens_seen": 41177936,
      "step": 70925
    },
    {
      "epoch": 10.564492106047066,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.016139448228124364,
      "loss": 0.7918,
      "num_input_tokens_seen": 41181040,
      "step": 70930
    },
    {
      "epoch": 10.565236818588025,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.016137504198918465,
      "loss": 0.8305,
      "num_input_tokens_seen": 41184016,
      "step": 70935
    },
    {
      "epoch": 10.565981531128983,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.01613556015049559,
      "loss": 0.792,
      "num_input_tokens_seen": 41186736,
      "step": 70940
    },
    {
      "epoch": 10.566726243669944,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.016133616082888574,
      "loss": 0.797,
      "num_input_tokens_seen": 41189712,
      "step": 70945
    },
    {
      "epoch": 10.567470956210903,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.016131671996130272,
      "loss": 0.8442,
      "num_input_tokens_seen": 41192368,
      "step": 70950
    },
    {
      "epoch": 10.568215668751861,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.016129727890253524,
      "loss": 0.7853,
      "num_input_tokens_seen": 41195088,
      "step": 70955
    },
    {
      "epoch": 10.568960381292822,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01612778376529117,
      "loss": 0.8024,
      "num_input_tokens_seen": 41198000,
      "step": 70960
    },
    {
      "epoch": 10.56970509383378,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.016125839621276054,
      "loss": 0.8147,
      "num_input_tokens_seen": 41200528,
      "step": 70965
    },
    {
      "epoch": 10.57044980637474,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01612389545824102,
      "loss": 0.8083,
      "num_input_tokens_seen": 41203472,
      "step": 70970
    },
    {
      "epoch": 10.571194518915698,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.016121951276218918,
      "loss": 0.8056,
      "num_input_tokens_seen": 41206224,
      "step": 70975
    },
    {
      "epoch": 10.571939231456657,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01612000707524259,
      "loss": 0.7929,
      "num_input_tokens_seen": 41209264,
      "step": 70980
    },
    {
      "epoch": 10.572683943997617,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.016118062855344874,
      "loss": 0.8066,
      "num_input_tokens_seen": 41211984,
      "step": 70985
    },
    {
      "epoch": 10.573428656538576,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.016116118616558633,
      "loss": 0.7893,
      "num_input_tokens_seen": 41214928,
      "step": 70990
    },
    {
      "epoch": 10.574173369079535,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.016114174358916693,
      "loss": 0.8183,
      "num_input_tokens_seen": 41217872,
      "step": 70995
    },
    {
      "epoch": 10.574918081620494,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.016112230082451915,
      "loss": 0.8047,
      "num_input_tokens_seen": 41220656,
      "step": 71000
    },
    {
      "epoch": 10.575662794161454,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.016110285787197136,
      "loss": 0.8123,
      "num_input_tokens_seen": 41223760,
      "step": 71005
    },
    {
      "epoch": 10.576407506702413,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.016108341473185208,
      "loss": 0.796,
      "num_input_tokens_seen": 41226544,
      "step": 71010
    },
    {
      "epoch": 10.577152219243372,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.01610639714044898,
      "loss": 0.8025,
      "num_input_tokens_seen": 41229360,
      "step": 71015
    },
    {
      "epoch": 10.57789693178433,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.016104452789021294,
      "loss": 0.8106,
      "num_input_tokens_seen": 41232080,
      "step": 71020
    },
    {
      "epoch": 10.578641644325291,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.016102508418935006,
      "loss": 0.7802,
      "num_input_tokens_seen": 41234640,
      "step": 71025
    },
    {
      "epoch": 10.57938635686625,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.016100564030222954,
      "loss": 0.8116,
      "num_input_tokens_seen": 41237776,
      "step": 71030
    },
    {
      "epoch": 10.580131069407209,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.016098619622917993,
      "loss": 0.8089,
      "num_input_tokens_seen": 41240560,
      "step": 71035
    },
    {
      "epoch": 10.580875781948167,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.016096675197052965,
      "loss": 0.7972,
      "num_input_tokens_seen": 41243472,
      "step": 71040
    },
    {
      "epoch": 10.581620494489128,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.01609473075266073,
      "loss": 0.8122,
      "num_input_tokens_seen": 41246384,
      "step": 71045
    },
    {
      "epoch": 10.582365207030087,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01609278628977413,
      "loss": 0.8074,
      "num_input_tokens_seen": 41249392,
      "step": 71050
    },
    {
      "epoch": 10.583109919571045,
      "grad_norm": 0.02734375,
      "learning_rate": 0.01609084180842602,
      "loss": 0.8006,
      "num_input_tokens_seen": 41252112,
      "step": 71055
    },
    {
      "epoch": 10.583854632112004,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.016088897308649243,
      "loss": 0.7888,
      "num_input_tokens_seen": 41254960,
      "step": 71060
    },
    {
      "epoch": 10.584599344652965,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.01608695279047665,
      "loss": 0.802,
      "num_input_tokens_seen": 41257872,
      "step": 71065
    },
    {
      "epoch": 10.585344057193923,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.0160850082539411,
      "loss": 0.7957,
      "num_input_tokens_seen": 41261104,
      "step": 71070
    },
    {
      "epoch": 10.586088769734882,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.01608306369907543,
      "loss": 0.8165,
      "num_input_tokens_seen": 41264208,
      "step": 71075
    },
    {
      "epoch": 10.586833482275841,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.016081119125912504,
      "loss": 0.8095,
      "num_input_tokens_seen": 41266960,
      "step": 71080
    },
    {
      "epoch": 10.587578194816802,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.016079174534485173,
      "loss": 0.8025,
      "num_input_tokens_seen": 41269904,
      "step": 71085
    },
    {
      "epoch": 10.58832290735776,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.016077229924826283,
      "loss": 0.7817,
      "num_input_tokens_seen": 41272752,
      "step": 71090
    },
    {
      "epoch": 10.589067619898719,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.016075285296968687,
      "loss": 0.784,
      "num_input_tokens_seen": 41275472,
      "step": 71095
    },
    {
      "epoch": 10.589812332439678,
      "grad_norm": 0.0390625,
      "learning_rate": 0.01607334065094524,
      "loss": 0.8004,
      "num_input_tokens_seen": 41278032,
      "step": 71100
    },
    {
      "epoch": 10.590557044980638,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.016071395986788792,
      "loss": 0.8192,
      "num_input_tokens_seen": 41281136,
      "step": 71105
    },
    {
      "epoch": 10.591301757521597,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.016069451304532197,
      "loss": 0.7795,
      "num_input_tokens_seen": 41284240,
      "step": 71110
    },
    {
      "epoch": 10.592046470062556,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.016067506604208307,
      "loss": 0.8001,
      "num_input_tokens_seen": 41287184,
      "step": 71115
    },
    {
      "epoch": 10.592791182603515,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.01606556188584998,
      "loss": 0.8116,
      "num_input_tokens_seen": 41290128,
      "step": 71120
    },
    {
      "epoch": 10.593535895144473,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.016063617149490075,
      "loss": 0.7875,
      "num_input_tokens_seen": 41293040,
      "step": 71125
    },
    {
      "epoch": 10.594280607685434,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01606167239516143,
      "loss": 0.7889,
      "num_input_tokens_seen": 41296112,
      "step": 71130
    },
    {
      "epoch": 10.595025320226393,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.016059727622896914,
      "loss": 0.808,
      "num_input_tokens_seen": 41298800,
      "step": 71135
    },
    {
      "epoch": 10.595770032767351,
      "grad_norm": 0.03125,
      "learning_rate": 0.016057782832729378,
      "loss": 0.7873,
      "num_input_tokens_seen": 41301616,
      "step": 71140
    },
    {
      "epoch": 10.59651474530831,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.01605583802469167,
      "loss": 0.8131,
      "num_input_tokens_seen": 41304496,
      "step": 71145
    },
    {
      "epoch": 10.59725945784927,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.016053893198816654,
      "loss": 0.8101,
      "num_input_tokens_seen": 41307408,
      "step": 71150
    },
    {
      "epoch": 10.59800417039023,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.016051948355137187,
      "loss": 0.7894,
      "num_input_tokens_seen": 41310288,
      "step": 71155
    },
    {
      "epoch": 10.598748882931188,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.016050003493686116,
      "loss": 0.7874,
      "num_input_tokens_seen": 41312816,
      "step": 71160
    },
    {
      "epoch": 10.599493595472147,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01604805861449631,
      "loss": 0.82,
      "num_input_tokens_seen": 41315952,
      "step": 71165
    },
    {
      "epoch": 10.600238308013108,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01604611371760061,
      "loss": 0.7985,
      "num_input_tokens_seen": 41318768,
      "step": 71170
    },
    {
      "epoch": 10.600983020554066,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.01604416880303189,
      "loss": 0.8211,
      "num_input_tokens_seen": 41321744,
      "step": 71175
    },
    {
      "epoch": 10.601727733095025,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.016042223870823,
      "loss": 0.7902,
      "num_input_tokens_seen": 41324528,
      "step": 71180
    },
    {
      "epoch": 10.602472445635984,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01604027892100679,
      "loss": 0.7861,
      "num_input_tokens_seen": 41327440,
      "step": 71185
    },
    {
      "epoch": 10.603217158176944,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.01603833395361613,
      "loss": 0.7849,
      "num_input_tokens_seen": 41330416,
      "step": 71190
    },
    {
      "epoch": 10.603961870717903,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01603638896868387,
      "loss": 0.8078,
      "num_input_tokens_seen": 41333296,
      "step": 71195
    },
    {
      "epoch": 10.604706583258862,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.016034443966242873,
      "loss": 0.7984,
      "num_input_tokens_seen": 41336240,
      "step": 71200
    },
    {
      "epoch": 10.60545129579982,
      "grad_norm": 0.03125,
      "learning_rate": 0.016032498946325996,
      "loss": 0.7814,
      "num_input_tokens_seen": 41338832,
      "step": 71205
    },
    {
      "epoch": 10.606196008340781,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.0160305539089661,
      "loss": 0.8272,
      "num_input_tokens_seen": 41341552,
      "step": 71210
    },
    {
      "epoch": 10.60694072088174,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.016028608854196044,
      "loss": 0.8065,
      "num_input_tokens_seen": 41344560,
      "step": 71215
    },
    {
      "epoch": 10.607685433422699,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.016026663782048686,
      "loss": 0.8124,
      "num_input_tokens_seen": 41347728,
      "step": 71220
    },
    {
      "epoch": 10.608430145963657,
      "grad_norm": 0.025390625,
      "learning_rate": 0.016024718692556887,
      "loss": 0.7918,
      "num_input_tokens_seen": 41350544,
      "step": 71225
    },
    {
      "epoch": 10.609174858504618,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.016022773585753505,
      "loss": 0.8035,
      "num_input_tokens_seen": 41353488,
      "step": 71230
    },
    {
      "epoch": 10.609919571045577,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.016020828461671407,
      "loss": 0.8025,
      "num_input_tokens_seen": 41356816,
      "step": 71235
    },
    {
      "epoch": 10.610664283586535,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.016018883320343444,
      "loss": 0.807,
      "num_input_tokens_seen": 41359888,
      "step": 71240
    },
    {
      "epoch": 10.611408996127494,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.016016938161802486,
      "loss": 0.8048,
      "num_input_tokens_seen": 41362704,
      "step": 71245
    },
    {
      "epoch": 10.612153708668455,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01601499298608139,
      "loss": 0.7981,
      "num_input_tokens_seen": 41365552,
      "step": 71250
    },
    {
      "epoch": 10.612898421209414,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.016013047793213022,
      "loss": 0.7938,
      "num_input_tokens_seen": 41368848,
      "step": 71255
    },
    {
      "epoch": 10.613643133750372,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.01601110258323024,
      "loss": 0.7992,
      "num_input_tokens_seen": 41371632,
      "step": 71260
    },
    {
      "epoch": 10.614387846291331,
      "grad_norm": 0.021484375,
      "learning_rate": 0.01600915735616591,
      "loss": 0.8052,
      "num_input_tokens_seen": 41374544,
      "step": 71265
    },
    {
      "epoch": 10.615132558832292,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.016007212112052888,
      "loss": 0.7882,
      "num_input_tokens_seen": 41377552,
      "step": 71270
    },
    {
      "epoch": 10.61587727137325,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.01600526685092404,
      "loss": 0.7856,
      "num_input_tokens_seen": 41380656,
      "step": 71275
    },
    {
      "epoch": 10.616621983914209,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.01600332157281223,
      "loss": 0.7992,
      "num_input_tokens_seen": 41383984,
      "step": 71280
    },
    {
      "epoch": 10.617366696455168,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.016001376277750324,
      "loss": 0.7944,
      "num_input_tokens_seen": 41386704,
      "step": 71285
    },
    {
      "epoch": 10.618111408996128,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.015999430965771187,
      "loss": 0.7847,
      "num_input_tokens_seen": 41389712,
      "step": 71290
    },
    {
      "epoch": 10.618856121537087,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.015997485636907678,
      "loss": 0.8044,
      "num_input_tokens_seen": 41392912,
      "step": 71295
    },
    {
      "epoch": 10.619600834078046,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.01599554029119266,
      "loss": 0.7884,
      "num_input_tokens_seen": 41395856,
      "step": 71300
    },
    {
      "epoch": 10.620345546619005,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.015993594928659,
      "loss": 0.7955,
      "num_input_tokens_seen": 41398896,
      "step": 71305
    },
    {
      "epoch": 10.621090259159963,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.015991649549339567,
      "loss": 0.7917,
      "num_input_tokens_seen": 41401840,
      "step": 71310
    },
    {
      "epoch": 10.621834971700924,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.01598970415326722,
      "loss": 0.7964,
      "num_input_tokens_seen": 41404720,
      "step": 71315
    },
    {
      "epoch": 10.622579684241883,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01598775874047483,
      "loss": 0.7897,
      "num_input_tokens_seen": 41407888,
      "step": 71320
    },
    {
      "epoch": 10.623324396782841,
      "grad_norm": 0.01953125,
      "learning_rate": 0.015985813310995255,
      "loss": 0.8048,
      "num_input_tokens_seen": 41410736,
      "step": 71325
    },
    {
      "epoch": 10.6240691093238,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01598386786486137,
      "loss": 0.8271,
      "num_input_tokens_seen": 41413520,
      "step": 71330
    },
    {
      "epoch": 10.62481382186476,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.01598192240210604,
      "loss": 0.7774,
      "num_input_tokens_seen": 41416336,
      "step": 71335
    },
    {
      "epoch": 10.62555853440572,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.015979976922762124,
      "loss": 0.7842,
      "num_input_tokens_seen": 41419280,
      "step": 71340
    },
    {
      "epoch": 10.626303246946678,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.015978031426862493,
      "loss": 0.7762,
      "num_input_tokens_seen": 41422256,
      "step": 71345
    },
    {
      "epoch": 10.627047959487637,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.015976085914440018,
      "loss": 0.8329,
      "num_input_tokens_seen": 41424944,
      "step": 71350
    },
    {
      "epoch": 10.627792672028598,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.015974140385527566,
      "loss": 0.7913,
      "num_input_tokens_seen": 41427888,
      "step": 71355
    },
    {
      "epoch": 10.628537384569556,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.015972194840157997,
      "loss": 0.8113,
      "num_input_tokens_seen": 41430864,
      "step": 71360
    },
    {
      "epoch": 10.629282097110515,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.015970249278364185,
      "loss": 0.8132,
      "num_input_tokens_seen": 41433744,
      "step": 71365
    },
    {
      "epoch": 10.630026809651474,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.015968303700179,
      "loss": 0.8001,
      "num_input_tokens_seen": 41436368,
      "step": 71370
    },
    {
      "epoch": 10.630771522192434,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.015966358105635313,
      "loss": 0.7689,
      "num_input_tokens_seen": 41439056,
      "step": 71375
    },
    {
      "epoch": 10.631516234733393,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.01596441249476598,
      "loss": 0.794,
      "num_input_tokens_seen": 41442064,
      "step": 71380
    },
    {
      "epoch": 10.632260947274352,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.015962466867603876,
      "loss": 0.779,
      "num_input_tokens_seen": 41444688,
      "step": 71385
    },
    {
      "epoch": 10.63300565981531,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.015960521224181883,
      "loss": 0.8038,
      "num_input_tokens_seen": 41447632,
      "step": 71390
    },
    {
      "epoch": 10.633750372356271,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01595857556453285,
      "loss": 0.7895,
      "num_input_tokens_seen": 41450640,
      "step": 71395
    },
    {
      "epoch": 10.63449508489723,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.01595662988868966,
      "loss": 0.8095,
      "num_input_tokens_seen": 41453648,
      "step": 71400
    },
    {
      "epoch": 10.635239797438189,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.015954684196685182,
      "loss": 0.826,
      "num_input_tokens_seen": 41456464,
      "step": 71405
    },
    {
      "epoch": 10.635984509979147,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.015952738488552283,
      "loss": 0.7988,
      "num_input_tokens_seen": 41459632,
      "step": 71410
    },
    {
      "epoch": 10.636729222520108,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.015950792764323842,
      "loss": 0.804,
      "num_input_tokens_seen": 41462448,
      "step": 71415
    },
    {
      "epoch": 10.637473935061067,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.015948847024032715,
      "loss": 0.8039,
      "num_input_tokens_seen": 41465232,
      "step": 71420
    },
    {
      "epoch": 10.638218647602026,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.01594690126771179,
      "loss": 0.8044,
      "num_input_tokens_seen": 41468112,
      "step": 71425
    },
    {
      "epoch": 10.638963360142984,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.01594495549539392,
      "loss": 0.8104,
      "num_input_tokens_seen": 41470992,
      "step": 71430
    },
    {
      "epoch": 10.639708072683945,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.015943009707111995,
      "loss": 0.8029,
      "num_input_tokens_seen": 41474032,
      "step": 71435
    },
    {
      "epoch": 10.640452785224904,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.01594106390289888,
      "loss": 0.7871,
      "num_input_tokens_seen": 41476848,
      "step": 71440
    },
    {
      "epoch": 10.641197497765862,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.015939118082787442,
      "loss": 0.8026,
      "num_input_tokens_seen": 41479664,
      "step": 71445
    },
    {
      "epoch": 10.641942210306821,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.015937172246810558,
      "loss": 0.7987,
      "num_input_tokens_seen": 41482352,
      "step": 71450
    },
    {
      "epoch": 10.64268692284778,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.015935226395001105,
      "loss": 0.8035,
      "num_input_tokens_seen": 41485360,
      "step": 71455
    },
    {
      "epoch": 10.64343163538874,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.015933280527391958,
      "loss": 0.7965,
      "num_input_tokens_seen": 41488048,
      "step": 71460
    },
    {
      "epoch": 10.6441763479297,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.015931334644015974,
      "loss": 0.7927,
      "num_input_tokens_seen": 41490992,
      "step": 71465
    },
    {
      "epoch": 10.644921060470658,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.015929388744906042,
      "loss": 0.8154,
      "num_input_tokens_seen": 41494128,
      "step": 71470
    },
    {
      "epoch": 10.645665773011618,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.015927442830095034,
      "loss": 0.8117,
      "num_input_tokens_seen": 41497072,
      "step": 71475
    },
    {
      "epoch": 10.646410485552577,
      "grad_norm": 0.04296875,
      "learning_rate": 0.015925496899615817,
      "loss": 0.8,
      "num_input_tokens_seen": 41500016,
      "step": 71480
    },
    {
      "epoch": 10.647155198093536,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01592355095350128,
      "loss": 0.7893,
      "num_input_tokens_seen": 41502640,
      "step": 71485
    },
    {
      "epoch": 10.647899910634495,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.01592160499178428,
      "loss": 0.7934,
      "num_input_tokens_seen": 41505584,
      "step": 71490
    },
    {
      "epoch": 10.648644623175453,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0159196590144977,
      "loss": 0.7841,
      "num_input_tokens_seen": 41508400,
      "step": 71495
    },
    {
      "epoch": 10.649389335716414,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.015917713021674423,
      "loss": 0.8373,
      "num_input_tokens_seen": 41511216,
      "step": 71500
    },
    {
      "epoch": 10.650134048257373,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.015915767013347307,
      "loss": 0.806,
      "num_input_tokens_seen": 41514224,
      "step": 71505
    },
    {
      "epoch": 10.650878760798332,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.015913820989549247,
      "loss": 0.7981,
      "num_input_tokens_seen": 41517200,
      "step": 71510
    },
    {
      "epoch": 10.65162347333929,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.015911874950313102,
      "loss": 0.8005,
      "num_input_tokens_seen": 41520304,
      "step": 71515
    },
    {
      "epoch": 10.65236818588025,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.015909928895671763,
      "loss": 0.7969,
      "num_input_tokens_seen": 41523024,
      "step": 71520
    },
    {
      "epoch": 10.65311289842121,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.015907982825658102,
      "loss": 0.7896,
      "num_input_tokens_seen": 41525744,
      "step": 71525
    },
    {
      "epoch": 10.653857610962168,
      "grad_norm": 0.0234375,
      "learning_rate": 0.01590603674030499,
      "loss": 0.8019,
      "num_input_tokens_seen": 41528400,
      "step": 71530
    },
    {
      "epoch": 10.654602323503127,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0159040906396453,
      "loss": 0.8135,
      "num_input_tokens_seen": 41531440,
      "step": 71535
    },
    {
      "epoch": 10.655347036044088,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.01590214452371193,
      "loss": 0.7954,
      "num_input_tokens_seen": 41534384,
      "step": 71540
    },
    {
      "epoch": 10.656091748585046,
      "grad_norm": 0.046875,
      "learning_rate": 0.015900198392537743,
      "loss": 0.8024,
      "num_input_tokens_seen": 41537552,
      "step": 71545
    },
    {
      "epoch": 10.656836461126005,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01589825224615562,
      "loss": 0.7906,
      "num_input_tokens_seen": 41540304,
      "step": 71550
    },
    {
      "epoch": 10.657581173666964,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.015896306084598435,
      "loss": 0.7994,
      "num_input_tokens_seen": 41543440,
      "step": 71555
    },
    {
      "epoch": 10.658325886207924,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.01589435990789907,
      "loss": 0.8,
      "num_input_tokens_seen": 41546128,
      "step": 71560
    },
    {
      "epoch": 10.659070598748883,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.0158924137160904,
      "loss": 0.8067,
      "num_input_tokens_seen": 41548848,
      "step": 71565
    },
    {
      "epoch": 10.659815311289842,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.015890467509205308,
      "loss": 0.7945,
      "num_input_tokens_seen": 41551504,
      "step": 71570
    },
    {
      "epoch": 10.6605600238308,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.015888521287276677,
      "loss": 0.808,
      "num_input_tokens_seen": 41554192,
      "step": 71575
    },
    {
      "epoch": 10.661304736371761,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01588657505033738,
      "loss": 0.803,
      "num_input_tokens_seen": 41557136,
      "step": 71580
    },
    {
      "epoch": 10.66204944891272,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.015884628798420296,
      "loss": 0.814,
      "num_input_tokens_seen": 41560016,
      "step": 71585
    },
    {
      "epoch": 10.662794161453679,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.015882682531558313,
      "loss": 0.7974,
      "num_input_tokens_seen": 41563056,
      "step": 71590
    },
    {
      "epoch": 10.663538873994638,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.015880736249784304,
      "loss": 0.7798,
      "num_input_tokens_seen": 41565648,
      "step": 71595
    },
    {
      "epoch": 10.664283586535598,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.015878789953131148,
      "loss": 0.7762,
      "num_input_tokens_seen": 41568624,
      "step": 71600
    },
    {
      "epoch": 10.665028299076557,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.01587684364163173,
      "loss": 0.8142,
      "num_input_tokens_seen": 41571728,
      "step": 71605
    },
    {
      "epoch": 10.665773011617516,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.015874897315318933,
      "loss": 0.7868,
      "num_input_tokens_seen": 41574864,
      "step": 71610
    },
    {
      "epoch": 10.666517724158474,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.01587295097422563,
      "loss": 0.8349,
      "num_input_tokens_seen": 41577840,
      "step": 71615
    },
    {
      "epoch": 10.667262436699435,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.015871004618384713,
      "loss": 0.7856,
      "num_input_tokens_seen": 41580848,
      "step": 71620
    },
    {
      "epoch": 10.668007149240394,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.015869058247829058,
      "loss": 0.8201,
      "num_input_tokens_seen": 41583824,
      "step": 71625
    },
    {
      "epoch": 10.668751861781352,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.015867111862591543,
      "loss": 0.8282,
      "num_input_tokens_seen": 41586544,
      "step": 71630
    },
    {
      "epoch": 10.669496574322311,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.015865165462705055,
      "loss": 0.7972,
      "num_input_tokens_seen": 41589232,
      "step": 71635
    },
    {
      "epoch": 10.67024128686327,
      "grad_norm": 0.04296875,
      "learning_rate": 0.015863219048202478,
      "loss": 0.8069,
      "num_input_tokens_seen": 41592080,
      "step": 71640
    },
    {
      "epoch": 10.67098599940423,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.015861272619116688,
      "loss": 0.8083,
      "num_input_tokens_seen": 41595056,
      "step": 71645
    },
    {
      "epoch": 10.67173071194519,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.01585932617548058,
      "loss": 0.8077,
      "num_input_tokens_seen": 41597776,
      "step": 71650
    },
    {
      "epoch": 10.672475424486148,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.015857379717327022,
      "loss": 0.8015,
      "num_input_tokens_seen": 41600496,
      "step": 71655
    },
    {
      "epoch": 10.673220137027108,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.015855433244688914,
      "loss": 0.8131,
      "num_input_tokens_seen": 41603472,
      "step": 71660
    },
    {
      "epoch": 10.673964849568067,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.015853486757599122,
      "loss": 0.7922,
      "num_input_tokens_seen": 41606352,
      "step": 71665
    },
    {
      "epoch": 10.674709562109026,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.015851540256090542,
      "loss": 0.7836,
      "num_input_tokens_seen": 41609296,
      "step": 71670
    },
    {
      "epoch": 10.675454274649985,
      "grad_norm": 0.03125,
      "learning_rate": 0.015849593740196056,
      "loss": 0.7987,
      "num_input_tokens_seen": 41612016,
      "step": 71675
    },
    {
      "epoch": 10.676198987190944,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.015847647209948544,
      "loss": 0.7786,
      "num_input_tokens_seen": 41614864,
      "step": 71680
    },
    {
      "epoch": 10.676943699731904,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.015845700665380897,
      "loss": 0.8137,
      "num_input_tokens_seen": 41617712,
      "step": 71685
    },
    {
      "epoch": 10.677688412272863,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.015843754106525998,
      "loss": 0.7702,
      "num_input_tokens_seen": 41620656,
      "step": 71690
    },
    {
      "epoch": 10.678433124813822,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.015841807533416727,
      "loss": 0.7863,
      "num_input_tokens_seen": 41623504,
      "step": 71695
    },
    {
      "epoch": 10.67917783735478,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.015839860946085978,
      "loss": 0.8091,
      "num_input_tokens_seen": 41626480,
      "step": 71700
    },
    {
      "epoch": 10.67992254989574,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.015837914344566627,
      "loss": 0.7839,
      "num_input_tokens_seen": 41629328,
      "step": 71705
    },
    {
      "epoch": 10.6806672624367,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.015835967728891565,
      "loss": 0.7872,
      "num_input_tokens_seen": 41632368,
      "step": 71710
    },
    {
      "epoch": 10.681411974977658,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.01583402109909368,
      "loss": 0.7619,
      "num_input_tokens_seen": 41635216,
      "step": 71715
    },
    {
      "epoch": 10.682156687518617,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.015832074455205857,
      "loss": 0.83,
      "num_input_tokens_seen": 41638032,
      "step": 71720
    },
    {
      "epoch": 10.682901400059578,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.015830127797260982,
      "loss": 0.8076,
      "num_input_tokens_seen": 41640848,
      "step": 71725
    },
    {
      "epoch": 10.683646112600536,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.01582818112529194,
      "loss": 0.7876,
      "num_input_tokens_seen": 41643760,
      "step": 71730
    },
    {
      "epoch": 10.684390825141495,
      "grad_norm": 0.05078125,
      "learning_rate": 0.015826234439331618,
      "loss": 0.8173,
      "num_input_tokens_seen": 41646416,
      "step": 71735
    },
    {
      "epoch": 10.685135537682454,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01582428773941291,
      "loss": 0.8372,
      "num_input_tokens_seen": 41649328,
      "step": 71740
    },
    {
      "epoch": 10.685880250223414,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.015822341025568693,
      "loss": 0.7678,
      "num_input_tokens_seen": 41652272,
      "step": 71745
    },
    {
      "epoch": 10.686624962764373,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.01582039429783186,
      "loss": 0.7979,
      "num_input_tokens_seen": 41655344,
      "step": 71750
    },
    {
      "epoch": 10.687369675305332,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.015818447556235304,
      "loss": 0.7778,
      "num_input_tokens_seen": 41658096,
      "step": 71755
    },
    {
      "epoch": 10.68811438784629,
      "grad_norm": 0.0390625,
      "learning_rate": 0.015816500800811907,
      "loss": 0.7874,
      "num_input_tokens_seen": 41660880,
      "step": 71760
    },
    {
      "epoch": 10.688859100387251,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.015814554031594555,
      "loss": 0.8157,
      "num_input_tokens_seen": 41663632,
      "step": 71765
    },
    {
      "epoch": 10.68960381292821,
      "grad_norm": 0.02734375,
      "learning_rate": 0.01581260724861614,
      "loss": 0.8025,
      "num_input_tokens_seen": 41666448,
      "step": 71770
    },
    {
      "epoch": 10.690348525469169,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.015810660451909556,
      "loss": 0.8001,
      "num_input_tokens_seen": 41669392,
      "step": 71775
    },
    {
      "epoch": 10.691093238010128,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.01580871364150769,
      "loss": 0.8031,
      "num_input_tokens_seen": 41672784,
      "step": 71780
    },
    {
      "epoch": 10.691837950551088,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.015806766817443422,
      "loss": 0.7951,
      "num_input_tokens_seen": 41675536,
      "step": 71785
    },
    {
      "epoch": 10.692582663092047,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.015804819979749654,
      "loss": 0.8172,
      "num_input_tokens_seen": 41678352,
      "step": 71790
    },
    {
      "epoch": 10.693327375633006,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.015802873128459266,
      "loss": 0.7811,
      "num_input_tokens_seen": 41680976,
      "step": 71795
    },
    {
      "epoch": 10.694072088173964,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.015800926263605155,
      "loss": 0.8066,
      "num_input_tokens_seen": 41683952,
      "step": 71800
    },
    {
      "epoch": 10.694816800714925,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.015798979385220208,
      "loss": 0.7834,
      "num_input_tokens_seen": 41686864,
      "step": 71805
    },
    {
      "epoch": 10.695561513255884,
      "grad_norm": 0.03125,
      "learning_rate": 0.015797032493337314,
      "loss": 0.8123,
      "num_input_tokens_seen": 41689840,
      "step": 71810
    },
    {
      "epoch": 10.696306225796842,
      "grad_norm": 0.03125,
      "learning_rate": 0.015795085587989373,
      "loss": 0.8174,
      "num_input_tokens_seen": 41692784,
      "step": 71815
    },
    {
      "epoch": 10.697050938337801,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.015793138669209263,
      "loss": 0.8209,
      "num_input_tokens_seen": 41695280,
      "step": 71820
    },
    {
      "epoch": 10.69779565087876,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.015791191737029885,
      "loss": 0.8011,
      "num_input_tokens_seen": 41698320,
      "step": 71825
    },
    {
      "epoch": 10.69854036341972,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.015789244791484126,
      "loss": 0.8053,
      "num_input_tokens_seen": 41701136,
      "step": 71830
    },
    {
      "epoch": 10.69928507596068,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.01578729783260488,
      "loss": 0.7985,
      "num_input_tokens_seen": 41703984,
      "step": 71835
    },
    {
      "epoch": 10.700029788501638,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.01578535086042503,
      "loss": 0.8025,
      "num_input_tokens_seen": 41707184,
      "step": 71840
    },
    {
      "epoch": 10.700774501042597,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.015783403874977477,
      "loss": 0.8188,
      "num_input_tokens_seen": 41710096,
      "step": 71845
    },
    {
      "epoch": 10.701519213583557,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.01578145687629512,
      "loss": 0.7814,
      "num_input_tokens_seen": 41713008,
      "step": 71850
    },
    {
      "epoch": 10.702263926124516,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.015779509864410835,
      "loss": 0.8214,
      "num_input_tokens_seen": 41715728,
      "step": 71855
    },
    {
      "epoch": 10.703008638665475,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01577756283935753,
      "loss": 0.7946,
      "num_input_tokens_seen": 41718448,
      "step": 71860
    },
    {
      "epoch": 10.703753351206434,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.015775615801168087,
      "loss": 0.7821,
      "num_input_tokens_seen": 41721232,
      "step": 71865
    },
    {
      "epoch": 10.704498063747394,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.015773668749875402,
      "loss": 0.7887,
      "num_input_tokens_seen": 41724016,
      "step": 71870
    },
    {
      "epoch": 10.705242776288353,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01577172168551237,
      "loss": 0.8008,
      "num_input_tokens_seen": 41726960,
      "step": 71875
    },
    {
      "epoch": 10.705987488829312,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.015769774608111884,
      "loss": 0.7879,
      "num_input_tokens_seen": 41730128,
      "step": 71880
    },
    {
      "epoch": 10.70673220137027,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.015767827517706844,
      "loss": 0.8109,
      "num_input_tokens_seen": 41733072,
      "step": 71885
    },
    {
      "epoch": 10.707476913911231,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.015765880414330136,
      "loss": 0.7805,
      "num_input_tokens_seen": 41735888,
      "step": 71890
    },
    {
      "epoch": 10.70822162645219,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.015763933298014652,
      "loss": 0.7983,
      "num_input_tokens_seen": 41738736,
      "step": 71895
    },
    {
      "epoch": 10.708966338993148,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.015761986168793297,
      "loss": 0.8127,
      "num_input_tokens_seen": 41741584,
      "step": 71900
    },
    {
      "epoch": 10.709711051534107,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.01576003902669896,
      "loss": 0.779,
      "num_input_tokens_seen": 41744880,
      "step": 71905
    },
    {
      "epoch": 10.710455764075068,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01575809187176453,
      "loss": 0.8009,
      "num_input_tokens_seen": 41747984,
      "step": 71910
    },
    {
      "epoch": 10.711200476616026,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.015756144704022913,
      "loss": 0.7908,
      "num_input_tokens_seen": 41751440,
      "step": 71915
    },
    {
      "epoch": 10.711945189156985,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.015754197523507,
      "loss": 0.7845,
      "num_input_tokens_seen": 41754512,
      "step": 71920
    },
    {
      "epoch": 10.712689901697944,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.015752250330249684,
      "loss": 0.7773,
      "num_input_tokens_seen": 41757232,
      "step": 71925
    },
    {
      "epoch": 10.713434614238905,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.015750303124283868,
      "loss": 0.7922,
      "num_input_tokens_seen": 41760176,
      "step": 71930
    },
    {
      "epoch": 10.714179326779863,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.015748355905642438,
      "loss": 0.7947,
      "num_input_tokens_seen": 41762992,
      "step": 71935
    },
    {
      "epoch": 10.714924039320822,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0157464086743583,
      "loss": 0.8013,
      "num_input_tokens_seen": 41766224,
      "step": 71940
    },
    {
      "epoch": 10.71566875186178,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.015744461430464343,
      "loss": 0.7933,
      "num_input_tokens_seen": 41768944,
      "step": 71945
    },
    {
      "epoch": 10.716413464402741,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.015742514173993468,
      "loss": 0.8127,
      "num_input_tokens_seen": 41772048,
      "step": 71950
    },
    {
      "epoch": 10.7171581769437,
      "grad_norm": 0.037109375,
      "learning_rate": 0.015740566904978572,
      "loss": 0.819,
      "num_input_tokens_seen": 41774864,
      "step": 71955
    },
    {
      "epoch": 10.717902889484659,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.015738619623452545,
      "loss": 0.8307,
      "num_input_tokens_seen": 41777712,
      "step": 71960
    },
    {
      "epoch": 10.718647602025618,
      "grad_norm": 0.04296875,
      "learning_rate": 0.01573667232944829,
      "loss": 0.8194,
      "num_input_tokens_seen": 41780752,
      "step": 71965
    },
    {
      "epoch": 10.719392314566576,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.015734725022998712,
      "loss": 0.8107,
      "num_input_tokens_seen": 41783536,
      "step": 71970
    },
    {
      "epoch": 10.720137027107537,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.015732777704136697,
      "loss": 0.7878,
      "num_input_tokens_seen": 41786160,
      "step": 71975
    },
    {
      "epoch": 10.720881739648496,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.015730830372895153,
      "loss": 0.8039,
      "num_input_tokens_seen": 41788816,
      "step": 71980
    },
    {
      "epoch": 10.721626452189454,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.01572888302930697,
      "loss": 0.7927,
      "num_input_tokens_seen": 41791632,
      "step": 71985
    },
    {
      "epoch": 10.722371164730415,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.015726935673405044,
      "loss": 0.8106,
      "num_input_tokens_seen": 41794480,
      "step": 71990
    },
    {
      "epoch": 10.723115877271374,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.015724988305222284,
      "loss": 0.804,
      "num_input_tokens_seen": 41797424,
      "step": 71995
    },
    {
      "epoch": 10.723860589812332,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.015723040924791577,
      "loss": 0.8162,
      "num_input_tokens_seen": 41800176,
      "step": 72000
    },
    {
      "epoch": 10.724605302353291,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.015721093532145836,
      "loss": 0.7936,
      "num_input_tokens_seen": 41802992,
      "step": 72005
    },
    {
      "epoch": 10.72535001489425,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.01571914612731795,
      "loss": 0.8042,
      "num_input_tokens_seen": 41805808,
      "step": 72010
    },
    {
      "epoch": 10.72609472743521,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.015717198710340823,
      "loss": 0.819,
      "num_input_tokens_seen": 41808880,
      "step": 72015
    },
    {
      "epoch": 10.72683943997617,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.01571525128124735,
      "loss": 0.7825,
      "num_input_tokens_seen": 41811888,
      "step": 72020
    },
    {
      "epoch": 10.727584152517128,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.015713303840070435,
      "loss": 0.8232,
      "num_input_tokens_seen": 41814640,
      "step": 72025
    },
    {
      "epoch": 10.728328865058087,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.015711356386842974,
      "loss": 0.807,
      "num_input_tokens_seen": 41817232,
      "step": 72030
    },
    {
      "epoch": 10.729073577599047,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.015709408921597872,
      "loss": 0.8073,
      "num_input_tokens_seen": 41820016,
      "step": 72035
    },
    {
      "epoch": 10.729818290140006,
      "grad_norm": 0.072265625,
      "learning_rate": 0.015707461444368024,
      "loss": 0.7879,
      "num_input_tokens_seen": 41822864,
      "step": 72040
    },
    {
      "epoch": 10.730563002680965,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.015705513955186334,
      "loss": 0.8086,
      "num_input_tokens_seen": 41826064,
      "step": 72045
    },
    {
      "epoch": 10.731307715221924,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.01570356645408571,
      "loss": 0.8331,
      "num_input_tokens_seen": 41828944,
      "step": 72050
    },
    {
      "epoch": 10.732052427762884,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.01570161894109904,
      "loss": 0.8027,
      "num_input_tokens_seen": 41831664,
      "step": 72055
    },
    {
      "epoch": 10.732797140303843,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.015699671416259234,
      "loss": 0.8153,
      "num_input_tokens_seen": 41834896,
      "step": 72060
    },
    {
      "epoch": 10.733541852844802,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.01569772387959919,
      "loss": 0.794,
      "num_input_tokens_seen": 41837680,
      "step": 72065
    },
    {
      "epoch": 10.73428656538576,
      "grad_norm": 0.0546875,
      "learning_rate": 0.015695776331151805,
      "loss": 0.7932,
      "num_input_tokens_seen": 41840496,
      "step": 72070
    },
    {
      "epoch": 10.735031277926721,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01569382877094999,
      "loss": 0.8187,
      "num_input_tokens_seen": 41843728,
      "step": 72075
    },
    {
      "epoch": 10.73577599046768,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.01569188119902664,
      "loss": 0.7974,
      "num_input_tokens_seen": 41846896,
      "step": 72080
    },
    {
      "epoch": 10.736520703008638,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.015689933615414667,
      "loss": 0.7816,
      "num_input_tokens_seen": 41850096,
      "step": 72085
    },
    {
      "epoch": 10.737265415549597,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.015687986020146965,
      "loss": 0.8092,
      "num_input_tokens_seen": 41853040,
      "step": 72090
    },
    {
      "epoch": 10.738010128090558,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.015686038413256432,
      "loss": 0.8217,
      "num_input_tokens_seen": 41856208,
      "step": 72095
    },
    {
      "epoch": 10.738754840631517,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.015684090794775986,
      "loss": 0.8157,
      "num_input_tokens_seen": 41859408,
      "step": 72100
    },
    {
      "epoch": 10.739499553172475,
      "grad_norm": 0.060546875,
      "learning_rate": 0.015682143164738517,
      "loss": 0.8109,
      "num_input_tokens_seen": 41862448,
      "step": 72105
    },
    {
      "epoch": 10.740244265713434,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.015680195523176933,
      "loss": 0.8119,
      "num_input_tokens_seen": 41865392,
      "step": 72110
    },
    {
      "epoch": 10.740988978254395,
      "grad_norm": 0.044921875,
      "learning_rate": 0.015678247870124135,
      "loss": 0.8008,
      "num_input_tokens_seen": 41868304,
      "step": 72115
    },
    {
      "epoch": 10.741733690795353,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.015676300205613034,
      "loss": 0.7917,
      "num_input_tokens_seen": 41871216,
      "step": 72120
    },
    {
      "epoch": 10.742478403336312,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.015674352529676524,
      "loss": 0.7931,
      "num_input_tokens_seen": 41874064,
      "step": 72125
    },
    {
      "epoch": 10.74322311587727,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.015672404842347512,
      "loss": 0.7933,
      "num_input_tokens_seen": 41876560,
      "step": 72130
    },
    {
      "epoch": 10.743967828418231,
      "grad_norm": 0.033203125,
      "learning_rate": 0.015670457143658903,
      "loss": 0.7873,
      "num_input_tokens_seen": 41879280,
      "step": 72135
    },
    {
      "epoch": 10.74471254095919,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.015668509433643604,
      "loss": 0.8287,
      "num_input_tokens_seen": 41882064,
      "step": 72140
    },
    {
      "epoch": 10.745457253500149,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.01566656171233452,
      "loss": 0.8276,
      "num_input_tokens_seen": 41885072,
      "step": 72145
    },
    {
      "epoch": 10.746201966041108,
      "grad_norm": 0.046875,
      "learning_rate": 0.01566461397976455,
      "loss": 0.8186,
      "num_input_tokens_seen": 41887952,
      "step": 72150
    },
    {
      "epoch": 10.746946678582066,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01566266623596661,
      "loss": 0.8188,
      "num_input_tokens_seen": 41890960,
      "step": 72155
    },
    {
      "epoch": 10.747691391123027,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.015660718480973588,
      "loss": 0.8156,
      "num_input_tokens_seen": 41893840,
      "step": 72160
    },
    {
      "epoch": 10.748436103663986,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.0156587707148184,
      "loss": 0.8086,
      "num_input_tokens_seen": 41896656,
      "step": 72165
    },
    {
      "epoch": 10.749180816204944,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.015656822937533954,
      "loss": 0.7939,
      "num_input_tokens_seen": 41899408,
      "step": 72170
    },
    {
      "epoch": 10.749925528745905,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.01565487514915315,
      "loss": 0.8064,
      "num_input_tokens_seen": 41902224,
      "step": 72175
    },
    {
      "epoch": 10.750670241286864,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.015652927349708898,
      "loss": 0.7985,
      "num_input_tokens_seen": 41905392,
      "step": 72180
    },
    {
      "epoch": 10.751414953827823,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.015650979539234104,
      "loss": 0.8098,
      "num_input_tokens_seen": 41908208,
      "step": 72185
    },
    {
      "epoch": 10.752159666368781,
      "grad_norm": 0.041015625,
      "learning_rate": 0.015649031717761663,
      "loss": 0.8091,
      "num_input_tokens_seen": 41910896,
      "step": 72190
    },
    {
      "epoch": 10.75290437890974,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.015647083885324497,
      "loss": 0.8101,
      "num_input_tokens_seen": 41913744,
      "step": 72195
    },
    {
      "epoch": 10.7536490914507,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.015645136041955506,
      "loss": 0.8056,
      "num_input_tokens_seen": 41916208,
      "step": 72200
    },
    {
      "epoch": 10.75439380399166,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.015643188187687597,
      "loss": 0.8048,
      "num_input_tokens_seen": 41919184,
      "step": 72205
    },
    {
      "epoch": 10.755138516532618,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.015641240322553675,
      "loss": 0.7958,
      "num_input_tokens_seen": 41922096,
      "step": 72210
    },
    {
      "epoch": 10.755883229073577,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.015639292446586656,
      "loss": 0.8046,
      "num_input_tokens_seen": 41925072,
      "step": 72215
    },
    {
      "epoch": 10.756627941614537,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.015637344559819435,
      "loss": 0.7839,
      "num_input_tokens_seen": 41927728,
      "step": 72220
    },
    {
      "epoch": 10.757372654155496,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.01563539666228493,
      "loss": 0.806,
      "num_input_tokens_seen": 41930640,
      "step": 72225
    },
    {
      "epoch": 10.758117366696455,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.01563344875401604,
      "loss": 0.791,
      "num_input_tokens_seen": 41933680,
      "step": 72230
    },
    {
      "epoch": 10.758862079237414,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.015631500835045678,
      "loss": 0.8108,
      "num_input_tokens_seen": 41936496,
      "step": 72235
    },
    {
      "epoch": 10.759606791778374,
      "grad_norm": 0.046875,
      "learning_rate": 0.01562955290540675,
      "loss": 0.8002,
      "num_input_tokens_seen": 41939472,
      "step": 72240
    },
    {
      "epoch": 10.760351504319333,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01562760496513217,
      "loss": 0.8082,
      "num_input_tokens_seen": 41942320,
      "step": 72245
    },
    {
      "epoch": 10.761096216860292,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.015625657014254838,
      "loss": 0.7969,
      "num_input_tokens_seen": 41945232,
      "step": 72250
    },
    {
      "epoch": 10.76184092940125,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01562370905280767,
      "loss": 0.795,
      "num_input_tokens_seen": 41948432,
      "step": 72255
    },
    {
      "epoch": 10.762585641942211,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.015621761080823568,
      "loss": 0.8262,
      "num_input_tokens_seen": 41951248,
      "step": 72260
    },
    {
      "epoch": 10.76333035448317,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.015619813098335445,
      "loss": 0.8058,
      "num_input_tokens_seen": 41954032,
      "step": 72265
    },
    {
      "epoch": 10.764075067024129,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.015617865105376213,
      "loss": 0.7926,
      "num_input_tokens_seen": 41956848,
      "step": 72270
    },
    {
      "epoch": 10.764819779565087,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.015615917101978774,
      "loss": 0.8071,
      "num_input_tokens_seen": 41959536,
      "step": 72275
    },
    {
      "epoch": 10.765564492106048,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.015613969088176041,
      "loss": 0.8354,
      "num_input_tokens_seen": 41962416,
      "step": 72280
    },
    {
      "epoch": 10.766309204647007,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.015612021064000928,
      "loss": 0.8043,
      "num_input_tokens_seen": 41965200,
      "step": 72285
    },
    {
      "epoch": 10.767053917187965,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.01561007302948634,
      "loss": 0.7953,
      "num_input_tokens_seen": 41968176,
      "step": 72290
    },
    {
      "epoch": 10.767798629728924,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.015608124984665187,
      "loss": 0.8004,
      "num_input_tokens_seen": 41971056,
      "step": 72295
    },
    {
      "epoch": 10.768543342269885,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.01560617692957038,
      "loss": 0.8002,
      "num_input_tokens_seen": 41973872,
      "step": 72300
    },
    {
      "epoch": 10.769288054810843,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.015604228864234833,
      "loss": 0.7963,
      "num_input_tokens_seen": 41976848,
      "step": 72305
    },
    {
      "epoch": 10.770032767351802,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.01560228078869145,
      "loss": 0.8097,
      "num_input_tokens_seen": 41979568,
      "step": 72310
    },
    {
      "epoch": 10.770777479892761,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.015600332702973148,
      "loss": 0.808,
      "num_input_tokens_seen": 41982576,
      "step": 72315
    },
    {
      "epoch": 10.771522192433721,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.015598384607112834,
      "loss": 0.8136,
      "num_input_tokens_seen": 41985584,
      "step": 72320
    },
    {
      "epoch": 10.77226690497468,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01559643650114342,
      "loss": 0.7958,
      "num_input_tokens_seen": 41989168,
      "step": 72325
    },
    {
      "epoch": 10.773011617515639,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.015594488385097817,
      "loss": 0.7994,
      "num_input_tokens_seen": 41992080,
      "step": 72330
    },
    {
      "epoch": 10.773756330056598,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.015592540259008938,
      "loss": 0.7895,
      "num_input_tokens_seen": 41994928,
      "step": 72335
    },
    {
      "epoch": 10.774501042597556,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.015590592122909691,
      "loss": 0.8154,
      "num_input_tokens_seen": 41997744,
      "step": 72340
    },
    {
      "epoch": 10.775245755138517,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.015588643976832994,
      "loss": 0.8113,
      "num_input_tokens_seen": 42000400,
      "step": 72345
    },
    {
      "epoch": 10.775990467679476,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.015586695820811756,
      "loss": 0.7982,
      "num_input_tokens_seen": 42003120,
      "step": 72350
    },
    {
      "epoch": 10.776735180220435,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.015584747654878887,
      "loss": 0.7809,
      "num_input_tokens_seen": 42006032,
      "step": 72355
    },
    {
      "epoch": 10.777479892761393,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.015582799479067298,
      "loss": 0.8253,
      "num_input_tokens_seen": 42009232,
      "step": 72360
    },
    {
      "epoch": 10.778224605302354,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.015580851293409904,
      "loss": 0.7996,
      "num_input_tokens_seen": 42012464,
      "step": 72365
    },
    {
      "epoch": 10.778969317843313,
      "grad_norm": 0.046875,
      "learning_rate": 0.01557890309793962,
      "loss": 0.8058,
      "num_input_tokens_seen": 42015440,
      "step": 72370
    },
    {
      "epoch": 10.779714030384271,
      "grad_norm": 0.04296875,
      "learning_rate": 0.015576954892689352,
      "loss": 0.8005,
      "num_input_tokens_seen": 42018320,
      "step": 72375
    },
    {
      "epoch": 10.78045874292523,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.015575006677692025,
      "loss": 0.8005,
      "num_input_tokens_seen": 42021136,
      "step": 72380
    },
    {
      "epoch": 10.78120345546619,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.015573058452980546,
      "loss": 0.7919,
      "num_input_tokens_seen": 42024272,
      "step": 72385
    },
    {
      "epoch": 10.78194816800715,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.01557111021858782,
      "loss": 0.8028,
      "num_input_tokens_seen": 42026960,
      "step": 72390
    },
    {
      "epoch": 10.782692880548108,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.015569161974546768,
      "loss": 0.804,
      "num_input_tokens_seen": 42030096,
      "step": 72395
    },
    {
      "epoch": 10.783437593089067,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.015567213720890302,
      "loss": 0.8003,
      "num_input_tokens_seen": 42032880,
      "step": 72400
    },
    {
      "epoch": 10.784182305630027,
      "grad_norm": 0.03515625,
      "learning_rate": 0.015565265457651337,
      "loss": 0.8009,
      "num_input_tokens_seen": 42035696,
      "step": 72405
    },
    {
      "epoch": 10.784927018170986,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.015563317184862788,
      "loss": 0.7961,
      "num_input_tokens_seen": 42038352,
      "step": 72410
    },
    {
      "epoch": 10.785671730711945,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.015561368902557573,
      "loss": 0.8041,
      "num_input_tokens_seen": 42041168,
      "step": 72415
    },
    {
      "epoch": 10.786416443252904,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.015559420610768596,
      "loss": 0.815,
      "num_input_tokens_seen": 42044016,
      "step": 72420
    },
    {
      "epoch": 10.787161155793864,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.015557472309528779,
      "loss": 0.8261,
      "num_input_tokens_seen": 42046800,
      "step": 72425
    },
    {
      "epoch": 10.787905868334823,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01555552399887103,
      "loss": 0.7764,
      "num_input_tokens_seen": 42049456,
      "step": 72430
    },
    {
      "epoch": 10.788650580875782,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.015553575678828266,
      "loss": 0.8125,
      "num_input_tokens_seen": 42052240,
      "step": 72435
    },
    {
      "epoch": 10.78939529341674,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.015551627349433408,
      "loss": 0.8121,
      "num_input_tokens_seen": 42055024,
      "step": 72440
    },
    {
      "epoch": 10.790140005957701,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.015549679010719363,
      "loss": 0.8014,
      "num_input_tokens_seen": 42057648,
      "step": 72445
    },
    {
      "epoch": 10.79088471849866,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.015547730662719054,
      "loss": 0.7925,
      "num_input_tokens_seen": 42060752,
      "step": 72450
    },
    {
      "epoch": 10.791629431039619,
      "grad_norm": 0.056640625,
      "learning_rate": 0.01554578230546539,
      "loss": 0.8144,
      "num_input_tokens_seen": 42063728,
      "step": 72455
    },
    {
      "epoch": 10.792374143580577,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.01554383393899129,
      "loss": 0.8075,
      "num_input_tokens_seen": 42066352,
      "step": 72460
    },
    {
      "epoch": 10.793118856121538,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.015541885563329666,
      "loss": 0.8041,
      "num_input_tokens_seen": 42069328,
      "step": 72465
    },
    {
      "epoch": 10.793863568662497,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.015539937178513437,
      "loss": 0.8103,
      "num_input_tokens_seen": 42071984,
      "step": 72470
    },
    {
      "epoch": 10.794608281203455,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.015537988784575518,
      "loss": 0.8082,
      "num_input_tokens_seen": 42075280,
      "step": 72475
    },
    {
      "epoch": 10.795352993744414,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.015536040381548825,
      "loss": 0.8307,
      "num_input_tokens_seen": 42078288,
      "step": 72480
    },
    {
      "epoch": 10.796097706285373,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.015534091969466276,
      "loss": 0.8089,
      "num_input_tokens_seen": 42081328,
      "step": 72485
    },
    {
      "epoch": 10.796842418826333,
      "grad_norm": 0.04296875,
      "learning_rate": 0.015532143548360786,
      "loss": 0.7942,
      "num_input_tokens_seen": 42084400,
      "step": 72490
    },
    {
      "epoch": 10.797587131367292,
      "grad_norm": 0.0234375,
      "learning_rate": 0.01553019511826527,
      "loss": 0.82,
      "num_input_tokens_seen": 42087440,
      "step": 72495
    },
    {
      "epoch": 10.798331843908251,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.015528246679212647,
      "loss": 0.8167,
      "num_input_tokens_seen": 42090416,
      "step": 72500
    },
    {
      "epoch": 10.799076556449211,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.015526298231235835,
      "loss": 0.8018,
      "num_input_tokens_seen": 42093200,
      "step": 72505
    },
    {
      "epoch": 10.79982126899017,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.015524349774367749,
      "loss": 0.7976,
      "num_input_tokens_seen": 42096304,
      "step": 72510
    },
    {
      "epoch": 10.800565981531129,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.015522401308641307,
      "loss": 0.8024,
      "num_input_tokens_seen": 42099472,
      "step": 72515
    },
    {
      "epoch": 10.801310694072088,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.015520452834089421,
      "loss": 0.7981,
      "num_input_tokens_seen": 42102352,
      "step": 72520
    },
    {
      "epoch": 10.802055406613047,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.015518504350745017,
      "loss": 0.7992,
      "num_input_tokens_seen": 42105200,
      "step": 72525
    },
    {
      "epoch": 10.802800119154007,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.015516555858641005,
      "loss": 0.8092,
      "num_input_tokens_seen": 42108240,
      "step": 72530
    },
    {
      "epoch": 10.803544831694966,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.01551460735781031,
      "loss": 0.8103,
      "num_input_tokens_seen": 42110928,
      "step": 72535
    },
    {
      "epoch": 10.804289544235925,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.015512658848285848,
      "loss": 0.7922,
      "num_input_tokens_seen": 42113776,
      "step": 72540
    },
    {
      "epoch": 10.805034256776883,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.015510710330100536,
      "loss": 0.8049,
      "num_input_tokens_seen": 42116592,
      "step": 72545
    },
    {
      "epoch": 10.805778969317844,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.015508761803287292,
      "loss": 0.8054,
      "num_input_tokens_seen": 42119248,
      "step": 72550
    },
    {
      "epoch": 10.806523681858803,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.015506813267879034,
      "loss": 0.8189,
      "num_input_tokens_seen": 42122256,
      "step": 72555
    },
    {
      "epoch": 10.807268394399761,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.015504864723908678,
      "loss": 0.8099,
      "num_input_tokens_seen": 42125168,
      "step": 72560
    },
    {
      "epoch": 10.80801310694072,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.01550291617140915,
      "loss": 0.8073,
      "num_input_tokens_seen": 42128016,
      "step": 72565
    },
    {
      "epoch": 10.80875781948168,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.015500967610413358,
      "loss": 0.8071,
      "num_input_tokens_seen": 42131152,
      "step": 72570
    },
    {
      "epoch": 10.80950253202264,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.01549901904095423,
      "loss": 0.8112,
      "num_input_tokens_seen": 42134128,
      "step": 72575
    },
    {
      "epoch": 10.810247244563598,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.015497070463064687,
      "loss": 0.7937,
      "num_input_tokens_seen": 42136976,
      "step": 72580
    },
    {
      "epoch": 10.810991957104557,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.015495121876777642,
      "loss": 0.8016,
      "num_input_tokens_seen": 42140080,
      "step": 72585
    },
    {
      "epoch": 10.811736669645517,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.015493173282126016,
      "loss": 0.8082,
      "num_input_tokens_seen": 42142608,
      "step": 72590
    },
    {
      "epoch": 10.812481382186476,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.015491224679142726,
      "loss": 0.7849,
      "num_input_tokens_seen": 42145392,
      "step": 72595
    },
    {
      "epoch": 10.813226094727435,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.015489276067860694,
      "loss": 0.8041,
      "num_input_tokens_seen": 42148560,
      "step": 72600
    },
    {
      "epoch": 10.813970807268394,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01548732744831284,
      "loss": 0.7985,
      "num_input_tokens_seen": 42151280,
      "step": 72605
    },
    {
      "epoch": 10.814715519809354,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.015485378820532086,
      "loss": 0.8014,
      "num_input_tokens_seen": 42154320,
      "step": 72610
    },
    {
      "epoch": 10.815460232350313,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.015483430184551355,
      "loss": 0.8007,
      "num_input_tokens_seen": 42157264,
      "step": 72615
    },
    {
      "epoch": 10.816204944891272,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.015481481540403555,
      "loss": 0.7891,
      "num_input_tokens_seen": 42160016,
      "step": 72620
    },
    {
      "epoch": 10.81694965743223,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.015479532888121619,
      "loss": 0.815,
      "num_input_tokens_seen": 42163088,
      "step": 72625
    },
    {
      "epoch": 10.817694369973191,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.015477584227738457,
      "loss": 0.8101,
      "num_input_tokens_seen": 42166000,
      "step": 72630
    },
    {
      "epoch": 10.81843908251415,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.015475635559286997,
      "loss": 0.7706,
      "num_input_tokens_seen": 42168656,
      "step": 72635
    },
    {
      "epoch": 10.819183795055109,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.015473686882800156,
      "loss": 0.7959,
      "num_input_tokens_seen": 42171472,
      "step": 72640
    },
    {
      "epoch": 10.819928507596067,
      "grad_norm": 0.04296875,
      "learning_rate": 0.015471738198310856,
      "loss": 0.811,
      "num_input_tokens_seen": 42174288,
      "step": 72645
    },
    {
      "epoch": 10.820673220137028,
      "grad_norm": 0.044921875,
      "learning_rate": 0.015469789505852026,
      "loss": 0.7873,
      "num_input_tokens_seen": 42177328,
      "step": 72650
    },
    {
      "epoch": 10.821417932677987,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.015467840805456573,
      "loss": 0.8047,
      "num_input_tokens_seen": 42180080,
      "step": 72655
    },
    {
      "epoch": 10.822162645218945,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.015465892097157426,
      "loss": 0.8073,
      "num_input_tokens_seen": 42182992,
      "step": 72660
    },
    {
      "epoch": 10.822907357759904,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.015463943380987506,
      "loss": 0.8092,
      "num_input_tokens_seen": 42185968,
      "step": 72665
    },
    {
      "epoch": 10.823652070300863,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.015461994656979736,
      "loss": 0.8001,
      "num_input_tokens_seen": 42188656,
      "step": 72670
    },
    {
      "epoch": 10.824396782841823,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.015460045925167035,
      "loss": 0.7991,
      "num_input_tokens_seen": 42191696,
      "step": 72675
    },
    {
      "epoch": 10.825141495382782,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.015458097185582328,
      "loss": 0.7717,
      "num_input_tokens_seen": 42194352,
      "step": 72680
    },
    {
      "epoch": 10.825886207923741,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.015456148438258531,
      "loss": 0.815,
      "num_input_tokens_seen": 42197520,
      "step": 72685
    },
    {
      "epoch": 10.826630920464702,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.015454199683228568,
      "loss": 0.7928,
      "num_input_tokens_seen": 42200176,
      "step": 72690
    },
    {
      "epoch": 10.82737563300566,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.015452250920525363,
      "loss": 0.7916,
      "num_input_tokens_seen": 42203024,
      "step": 72695
    },
    {
      "epoch": 10.828120345546619,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.015450302150181842,
      "loss": 0.8218,
      "num_input_tokens_seen": 42205840,
      "step": 72700
    },
    {
      "epoch": 10.828865058087578,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.01544835337223092,
      "loss": 0.7854,
      "num_input_tokens_seen": 42208720,
      "step": 72705
    },
    {
      "epoch": 10.829609770628537,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.015446404586705528,
      "loss": 0.7868,
      "num_input_tokens_seen": 42211600,
      "step": 72710
    },
    {
      "epoch": 10.830354483169497,
      "grad_norm": 0.0625,
      "learning_rate": 0.015444455793638586,
      "loss": 0.8063,
      "num_input_tokens_seen": 42214640,
      "step": 72715
    },
    {
      "epoch": 10.831099195710456,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.015442506993063009,
      "loss": 0.792,
      "num_input_tokens_seen": 42217520,
      "step": 72720
    },
    {
      "epoch": 10.831843908251415,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.01544055818501173,
      "loss": 0.7807,
      "num_input_tokens_seen": 42220464,
      "step": 72725
    },
    {
      "epoch": 10.832588620792373,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.015438609369517664,
      "loss": 0.8152,
      "num_input_tokens_seen": 42223568,
      "step": 72730
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.015436660546613742,
      "loss": 0.78,
      "num_input_tokens_seen": 42226288,
      "step": 72735
    },
    {
      "epoch": 10.834078045874293,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01543471171633288,
      "loss": 0.8044,
      "num_input_tokens_seen": 42228880,
      "step": 72740
    },
    {
      "epoch": 10.834822758415251,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01543276287870801,
      "loss": 0.7695,
      "num_input_tokens_seen": 42232016,
      "step": 72745
    },
    {
      "epoch": 10.83556747095621,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.015430814033772052,
      "loss": 0.7765,
      "num_input_tokens_seen": 42235024,
      "step": 72750
    },
    {
      "epoch": 10.83631218349717,
      "grad_norm": 0.0625,
      "learning_rate": 0.015428865181557925,
      "loss": 0.8029,
      "num_input_tokens_seen": 42237936,
      "step": 72755
    },
    {
      "epoch": 10.83705689603813,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.015426916322098558,
      "loss": 0.8048,
      "num_input_tokens_seen": 42240976,
      "step": 72760
    },
    {
      "epoch": 10.837801608579088,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.015424967455426873,
      "loss": 0.8255,
      "num_input_tokens_seen": 42243824,
      "step": 72765
    },
    {
      "epoch": 10.838546321120047,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.015423018581575795,
      "loss": 0.7978,
      "num_input_tokens_seen": 42246544,
      "step": 72770
    },
    {
      "epoch": 10.839291033661008,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.015421069700578246,
      "loss": 0.8019,
      "num_input_tokens_seen": 42249488,
      "step": 72775
    },
    {
      "epoch": 10.840035746201966,
      "grad_norm": 0.08203125,
      "learning_rate": 0.015419120812467157,
      "loss": 0.7917,
      "num_input_tokens_seen": 42252656,
      "step": 72780
    },
    {
      "epoch": 10.840780458742925,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.015417171917275445,
      "loss": 0.835,
      "num_input_tokens_seen": 42255408,
      "step": 72785
    },
    {
      "epoch": 10.841525171283884,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.015415223015036039,
      "loss": 0.7873,
      "num_input_tokens_seen": 42258288,
      "step": 72790
    },
    {
      "epoch": 10.842269883824844,
      "grad_norm": 0.06640625,
      "learning_rate": 0.01541327410578186,
      "loss": 0.7962,
      "num_input_tokens_seen": 42260880,
      "step": 72795
    },
    {
      "epoch": 10.843014596365803,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.015411325189545836,
      "loss": 0.817,
      "num_input_tokens_seen": 42263664,
      "step": 72800
    },
    {
      "epoch": 10.843759308906762,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.015409376266360889,
      "loss": 0.8199,
      "num_input_tokens_seen": 42266672,
      "step": 72805
    },
    {
      "epoch": 10.84450402144772,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.01540742733625995,
      "loss": 0.7933,
      "num_input_tokens_seen": 42269392,
      "step": 72810
    },
    {
      "epoch": 10.845248733988681,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.015405478399275939,
      "loss": 0.8157,
      "num_input_tokens_seen": 42272240,
      "step": 72815
    },
    {
      "epoch": 10.84599344652964,
      "grad_norm": 0.05859375,
      "learning_rate": 0.01540352945544178,
      "loss": 0.8072,
      "num_input_tokens_seen": 42274992,
      "step": 72820
    },
    {
      "epoch": 10.846738159070599,
      "grad_norm": 0.12890625,
      "learning_rate": 0.015401580504790402,
      "loss": 0.7861,
      "num_input_tokens_seen": 42278160,
      "step": 72825
    },
    {
      "epoch": 10.847482871611557,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.015399631547354735,
      "loss": 0.7932,
      "num_input_tokens_seen": 42280880,
      "step": 72830
    },
    {
      "epoch": 10.848227584152518,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.01539768258316769,
      "loss": 0.7671,
      "num_input_tokens_seen": 42283728,
      "step": 72835
    },
    {
      "epoch": 10.848972296693477,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.015395733612262208,
      "loss": 0.7929,
      "num_input_tokens_seen": 42286416,
      "step": 72840
    },
    {
      "epoch": 10.849717009234435,
      "grad_norm": 0.025390625,
      "learning_rate": 0.015393784634671209,
      "loss": 0.7882,
      "num_input_tokens_seen": 42289264,
      "step": 72845
    },
    {
      "epoch": 10.850461721775394,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.015391835650427614,
      "loss": 0.8079,
      "num_input_tokens_seen": 42291824,
      "step": 72850
    },
    {
      "epoch": 10.851206434316353,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01538988665956436,
      "loss": 0.7996,
      "num_input_tokens_seen": 42294448,
      "step": 72855
    },
    {
      "epoch": 10.851951146857314,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.015387937662114362,
      "loss": 0.8209,
      "num_input_tokens_seen": 42297424,
      "step": 72860
    },
    {
      "epoch": 10.852695859398272,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.015385988658110552,
      "loss": 0.7805,
      "num_input_tokens_seen": 42300272,
      "step": 72865
    },
    {
      "epoch": 10.853440571939231,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.015384039647585862,
      "loss": 0.8003,
      "num_input_tokens_seen": 42303120,
      "step": 72870
    },
    {
      "epoch": 10.85418528448019,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.015382090630573208,
      "loss": 0.8313,
      "num_input_tokens_seen": 42306128,
      "step": 72875
    },
    {
      "epoch": 10.85492999702115,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.015380141607105522,
      "loss": 0.7767,
      "num_input_tokens_seen": 42309104,
      "step": 72880
    },
    {
      "epoch": 10.85567470956211,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.015378192577215729,
      "loss": 0.8143,
      "num_input_tokens_seen": 42312336,
      "step": 72885
    },
    {
      "epoch": 10.856419422103068,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.015376243540936755,
      "loss": 0.8151,
      "num_input_tokens_seen": 42315312,
      "step": 72890
    },
    {
      "epoch": 10.857164134644027,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.01537429449830153,
      "loss": 0.7853,
      "num_input_tokens_seen": 42318512,
      "step": 72895
    },
    {
      "epoch": 10.857908847184987,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.01537234544934298,
      "loss": 0.8159,
      "num_input_tokens_seen": 42321584,
      "step": 72900
    },
    {
      "epoch": 10.858653559725946,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.015370396394094031,
      "loss": 0.824,
      "num_input_tokens_seen": 42324400,
      "step": 72905
    },
    {
      "epoch": 10.859398272266905,
      "grad_norm": 0.0703125,
      "learning_rate": 0.015368447332587614,
      "loss": 0.8051,
      "num_input_tokens_seen": 42327312,
      "step": 72910
    },
    {
      "epoch": 10.860142984807863,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.015366498264856657,
      "loss": 0.8161,
      "num_input_tokens_seen": 42330384,
      "step": 72915
    },
    {
      "epoch": 10.860887697348824,
      "grad_norm": 0.06640625,
      "learning_rate": 0.01536454919093408,
      "loss": 0.8121,
      "num_input_tokens_seen": 42333328,
      "step": 72920
    },
    {
      "epoch": 10.861632409889783,
      "grad_norm": 0.064453125,
      "learning_rate": 0.015362600110852814,
      "loss": 0.7974,
      "num_input_tokens_seen": 42336400,
      "step": 72925
    },
    {
      "epoch": 10.862377122430741,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.015360651024645785,
      "loss": 0.8159,
      "num_input_tokens_seen": 42339312,
      "step": 72930
    },
    {
      "epoch": 10.8631218349717,
      "grad_norm": 0.052734375,
      "learning_rate": 0.015358701932345929,
      "loss": 0.8163,
      "num_input_tokens_seen": 42342032,
      "step": 72935
    },
    {
      "epoch": 10.86386654751266,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.015356752833986164,
      "loss": 0.8065,
      "num_input_tokens_seen": 42345072,
      "step": 72940
    },
    {
      "epoch": 10.86461126005362,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.015354803729599428,
      "loss": 0.7869,
      "num_input_tokens_seen": 42347760,
      "step": 72945
    },
    {
      "epoch": 10.865355972594578,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.015352854619218644,
      "loss": 0.8036,
      "num_input_tokens_seen": 42350608,
      "step": 72950
    },
    {
      "epoch": 10.866100685135537,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.015350905502876735,
      "loss": 0.8089,
      "num_input_tokens_seen": 42353968,
      "step": 72955
    },
    {
      "epoch": 10.866845397676498,
      "grad_norm": 0.041015625,
      "learning_rate": 0.015348956380606633,
      "loss": 0.8043,
      "num_input_tokens_seen": 42356720,
      "step": 72960
    },
    {
      "epoch": 10.867590110217456,
      "grad_norm": 0.06640625,
      "learning_rate": 0.015347007252441269,
      "loss": 0.788,
      "num_input_tokens_seen": 42359440,
      "step": 72965
    },
    {
      "epoch": 10.868334822758415,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01534505811841357,
      "loss": 0.8066,
      "num_input_tokens_seen": 42362224,
      "step": 72970
    },
    {
      "epoch": 10.869079535299374,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.015343108978556468,
      "loss": 0.812,
      "num_input_tokens_seen": 42365040,
      "step": 72975
    },
    {
      "epoch": 10.869824247840334,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.015341159832902887,
      "loss": 0.7929,
      "num_input_tokens_seen": 42367952,
      "step": 72980
    },
    {
      "epoch": 10.870568960381293,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.015339210681485756,
      "loss": 0.7952,
      "num_input_tokens_seen": 42370960,
      "step": 72985
    },
    {
      "epoch": 10.871313672922252,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.015337261524338012,
      "loss": 0.7949,
      "num_input_tokens_seen": 42373968,
      "step": 72990
    },
    {
      "epoch": 10.87205838546321,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.015335312361492567,
      "loss": 0.7985,
      "num_input_tokens_seen": 42376752,
      "step": 72995
    },
    {
      "epoch": 10.872803098004171,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.015333363192982365,
      "loss": 0.7752,
      "num_input_tokens_seen": 42379280,
      "step": 73000
    },
    {
      "epoch": 10.87354781054513,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01533141401884033,
      "loss": 0.7752,
      "num_input_tokens_seen": 42381936,
      "step": 73005
    },
    {
      "epoch": 10.874292523086089,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.015329464839099395,
      "loss": 0.8194,
      "num_input_tokens_seen": 42384624,
      "step": 73010
    },
    {
      "epoch": 10.875037235627047,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.015327515653792485,
      "loss": 0.8481,
      "num_input_tokens_seen": 42387376,
      "step": 73015
    },
    {
      "epoch": 10.875781948168008,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.01532556646295253,
      "loss": 0.7949,
      "num_input_tokens_seen": 42389904,
      "step": 73020
    },
    {
      "epoch": 10.876526660708967,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.015323617266612457,
      "loss": 0.7943,
      "num_input_tokens_seen": 42392912,
      "step": 73025
    },
    {
      "epoch": 10.877271373249926,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.015321668064805206,
      "loss": 0.7837,
      "num_input_tokens_seen": 42395856,
      "step": 73030
    },
    {
      "epoch": 10.878016085790884,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.015319718857563697,
      "loss": 0.7962,
      "num_input_tokens_seen": 42398608,
      "step": 73035
    },
    {
      "epoch": 10.878760798331843,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.01531776964492086,
      "loss": 0.8127,
      "num_input_tokens_seen": 42401328,
      "step": 73040
    },
    {
      "epoch": 10.879505510872804,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.015315820426909633,
      "loss": 0.7839,
      "num_input_tokens_seen": 42404432,
      "step": 73045
    },
    {
      "epoch": 10.880250223413762,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.015313871203562939,
      "loss": 0.8196,
      "num_input_tokens_seen": 42407280,
      "step": 73050
    },
    {
      "epoch": 10.880994935954721,
      "grad_norm": 0.029296875,
      "learning_rate": 0.015311921974913706,
      "loss": 0.8066,
      "num_input_tokens_seen": 42410160,
      "step": 73055
    },
    {
      "epoch": 10.88173964849568,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.015309972740994873,
      "loss": 0.7919,
      "num_input_tokens_seen": 42413232,
      "step": 73060
    },
    {
      "epoch": 10.88248436103664,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.015308023501839364,
      "loss": 0.7927,
      "num_input_tokens_seen": 42416080,
      "step": 73065
    },
    {
      "epoch": 10.8832290735776,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.015306074257480111,
      "loss": 0.8089,
      "num_input_tokens_seen": 42418928,
      "step": 73070
    },
    {
      "epoch": 10.883973786118558,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.015304125007950043,
      "loss": 0.8047,
      "num_input_tokens_seen": 42421776,
      "step": 73075
    },
    {
      "epoch": 10.884718498659517,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.015302175753282098,
      "loss": 0.7963,
      "num_input_tokens_seen": 42424624,
      "step": 73080
    },
    {
      "epoch": 10.885463211200477,
      "grad_norm": 0.05859375,
      "learning_rate": 0.015300226493509194,
      "loss": 0.8157,
      "num_input_tokens_seen": 42427920,
      "step": 73085
    },
    {
      "epoch": 10.886207923741436,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.015298277228664272,
      "loss": 0.8052,
      "num_input_tokens_seen": 42430768,
      "step": 73090
    },
    {
      "epoch": 10.886952636282395,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.015296327958780257,
      "loss": 0.7642,
      "num_input_tokens_seen": 42433616,
      "step": 73095
    },
    {
      "epoch": 10.887697348823353,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.015294378683890083,
      "loss": 0.8061,
      "num_input_tokens_seen": 42436560,
      "step": 73100
    },
    {
      "epoch": 10.888442061364314,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01529242940402668,
      "loss": 0.7836,
      "num_input_tokens_seen": 42439216,
      "step": 73105
    },
    {
      "epoch": 10.889186773905273,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.015290480119222983,
      "loss": 0.8138,
      "num_input_tokens_seen": 42442704,
      "step": 73110
    },
    {
      "epoch": 10.889931486446232,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.015288530829511917,
      "loss": 0.8237,
      "num_input_tokens_seen": 42445680,
      "step": 73115
    },
    {
      "epoch": 10.89067619898719,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.015286581534926418,
      "loss": 0.8157,
      "num_input_tokens_seen": 42448432,
      "step": 73120
    },
    {
      "epoch": 10.89142091152815,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.015284632235499411,
      "loss": 0.8175,
      "num_input_tokens_seen": 42451248,
      "step": 73125
    },
    {
      "epoch": 10.89216562406911,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.015282682931263836,
      "loss": 0.7931,
      "num_input_tokens_seen": 42454000,
      "step": 73130
    },
    {
      "epoch": 10.892910336610068,
      "grad_norm": 0.02734375,
      "learning_rate": 0.015280733622252615,
      "loss": 0.7957,
      "num_input_tokens_seen": 42456816,
      "step": 73135
    },
    {
      "epoch": 10.893655049151027,
      "grad_norm": 0.041015625,
      "learning_rate": 0.015278784308498685,
      "loss": 0.8215,
      "num_input_tokens_seen": 42459536,
      "step": 73140
    },
    {
      "epoch": 10.894399761691988,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.015276834990034984,
      "loss": 0.8125,
      "num_input_tokens_seen": 42462416,
      "step": 73145
    },
    {
      "epoch": 10.895144474232946,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.015274885666894436,
      "loss": 0.7896,
      "num_input_tokens_seen": 42465328,
      "step": 73150
    },
    {
      "epoch": 10.895889186773905,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.01527293633910997,
      "loss": 0.7995,
      "num_input_tokens_seen": 42468080,
      "step": 73155
    },
    {
      "epoch": 10.896633899314864,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.015270987006714523,
      "loss": 0.7859,
      "num_input_tokens_seen": 42470864,
      "step": 73160
    },
    {
      "epoch": 10.897378611855824,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.015269037669741026,
      "loss": 0.8192,
      "num_input_tokens_seen": 42473968,
      "step": 73165
    },
    {
      "epoch": 10.898123324396783,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.01526708832822241,
      "loss": 0.833,
      "num_input_tokens_seen": 42476656,
      "step": 73170
    },
    {
      "epoch": 10.898868036937742,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.015265138982191612,
      "loss": 0.8322,
      "num_input_tokens_seen": 42479472,
      "step": 73175
    },
    {
      "epoch": 10.8996127494787,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.015263189631681558,
      "loss": 0.8136,
      "num_input_tokens_seen": 42482096,
      "step": 73180
    },
    {
      "epoch": 10.90035746201966,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.01526124027672518,
      "loss": 0.8084,
      "num_input_tokens_seen": 42485104,
      "step": 73185
    },
    {
      "epoch": 10.90110217456062,
      "grad_norm": 0.03125,
      "learning_rate": 0.015259290917355419,
      "loss": 0.8272,
      "num_input_tokens_seen": 42488240,
      "step": 73190
    },
    {
      "epoch": 10.901846887101579,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.015257341553605197,
      "loss": 0.8198,
      "num_input_tokens_seen": 42491312,
      "step": 73195
    },
    {
      "epoch": 10.902591599642538,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.015255392185507451,
      "loss": 0.8192,
      "num_input_tokens_seen": 42494288,
      "step": 73200
    },
    {
      "epoch": 10.903336312183498,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.015253442813095113,
      "loss": 0.8,
      "num_input_tokens_seen": 42497040,
      "step": 73205
    },
    {
      "epoch": 10.904081024724457,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.015251493436401121,
      "loss": 0.8078,
      "num_input_tokens_seen": 42499952,
      "step": 73210
    },
    {
      "epoch": 10.904825737265416,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0152495440554584,
      "loss": 0.8128,
      "num_input_tokens_seen": 42502640,
      "step": 73215
    },
    {
      "epoch": 10.905570449806374,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.015247594670299885,
      "loss": 0.7962,
      "num_input_tokens_seen": 42505584,
      "step": 73220
    },
    {
      "epoch": 10.906315162347333,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.015245645280958509,
      "loss": 0.8111,
      "num_input_tokens_seen": 42508432,
      "step": 73225
    },
    {
      "epoch": 10.907059874888294,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.015243695887467209,
      "loss": 0.7973,
      "num_input_tokens_seen": 42511440,
      "step": 73230
    },
    {
      "epoch": 10.907804587429252,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.015241746489858913,
      "loss": 0.7886,
      "num_input_tokens_seen": 42514192,
      "step": 73235
    },
    {
      "epoch": 10.908549299970211,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.015239797088166553,
      "loss": 0.7964,
      "num_input_tokens_seen": 42516848,
      "step": 73240
    },
    {
      "epoch": 10.90929401251117,
      "grad_norm": 0.046875,
      "learning_rate": 0.015237847682423072,
      "loss": 0.798,
      "num_input_tokens_seen": 42519920,
      "step": 73245
    },
    {
      "epoch": 10.91003872505213,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.015235898272661391,
      "loss": 0.7804,
      "num_input_tokens_seen": 42522576,
      "step": 73250
    },
    {
      "epoch": 10.91078343759309,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.015233948858914447,
      "loss": 0.7978,
      "num_input_tokens_seen": 42525424,
      "step": 73255
    },
    {
      "epoch": 10.911528150134048,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.015231999441215175,
      "loss": 0.8138,
      "num_input_tokens_seen": 42528240,
      "step": 73260
    },
    {
      "epoch": 10.912272862675007,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.015230050019596512,
      "loss": 0.8171,
      "num_input_tokens_seen": 42531152,
      "step": 73265
    },
    {
      "epoch": 10.913017575215967,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.015228100594091385,
      "loss": 0.7935,
      "num_input_tokens_seen": 42533808,
      "step": 73270
    },
    {
      "epoch": 10.913762287756926,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.015226151164732733,
      "loss": 0.7903,
      "num_input_tokens_seen": 42536624,
      "step": 73275
    },
    {
      "epoch": 10.914507000297885,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.015224201731553486,
      "loss": 0.802,
      "num_input_tokens_seen": 42539472,
      "step": 73280
    },
    {
      "epoch": 10.915251712838844,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.015222252294586578,
      "loss": 0.8132,
      "num_input_tokens_seen": 42542320,
      "step": 73285
    },
    {
      "epoch": 10.915996425379804,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.015220302853864942,
      "loss": 0.7851,
      "num_input_tokens_seen": 42545040,
      "step": 73290
    },
    {
      "epoch": 10.916741137920763,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.015218353409421513,
      "loss": 0.8012,
      "num_input_tokens_seen": 42548016,
      "step": 73295
    },
    {
      "epoch": 10.917485850461722,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.015216403961289225,
      "loss": 0.8031,
      "num_input_tokens_seen": 42550736,
      "step": 73300
    },
    {
      "epoch": 10.91823056300268,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.015214454509501011,
      "loss": 0.8067,
      "num_input_tokens_seen": 42553520,
      "step": 73305
    },
    {
      "epoch": 10.918975275543641,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.015212505054089812,
      "loss": 0.8039,
      "num_input_tokens_seen": 42556528,
      "step": 73310
    },
    {
      "epoch": 10.9197199880846,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.015210555595088555,
      "loss": 0.7875,
      "num_input_tokens_seen": 42559440,
      "step": 73315
    },
    {
      "epoch": 10.920464700625558,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01520860613253017,
      "loss": 0.8,
      "num_input_tokens_seen": 42562320,
      "step": 73320
    },
    {
      "epoch": 10.921209413166517,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.015206656666447601,
      "loss": 0.8008,
      "num_input_tokens_seen": 42564976,
      "step": 73325
    },
    {
      "epoch": 10.921954125707478,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.015204707196873776,
      "loss": 0.8047,
      "num_input_tokens_seen": 42568112,
      "step": 73330
    },
    {
      "epoch": 10.922698838248436,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.015202757723841628,
      "loss": 0.8184,
      "num_input_tokens_seen": 42571088,
      "step": 73335
    },
    {
      "epoch": 10.923443550789395,
      "grad_norm": 0.04296875,
      "learning_rate": 0.015200808247384099,
      "loss": 0.7911,
      "num_input_tokens_seen": 42573904,
      "step": 73340
    },
    {
      "epoch": 10.924188263330354,
      "grad_norm": 0.037109375,
      "learning_rate": 0.015198858767534115,
      "loss": 0.8139,
      "num_input_tokens_seen": 42577360,
      "step": 73345
    },
    {
      "epoch": 10.924932975871315,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.015196909284324616,
      "loss": 0.8185,
      "num_input_tokens_seen": 42580464,
      "step": 73350
    },
    {
      "epoch": 10.925677688412273,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.015194959797788538,
      "loss": 0.7962,
      "num_input_tokens_seen": 42583376,
      "step": 73355
    },
    {
      "epoch": 10.926422400953232,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.015193010307958808,
      "loss": 0.8055,
      "num_input_tokens_seen": 42586672,
      "step": 73360
    },
    {
      "epoch": 10.92716711349419,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.015191060814868367,
      "loss": 0.7775,
      "num_input_tokens_seen": 42589648,
      "step": 73365
    },
    {
      "epoch": 10.92791182603515,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.015189111318550143,
      "loss": 0.8,
      "num_input_tokens_seen": 42592496,
      "step": 73370
    },
    {
      "epoch": 10.92865653857611,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.015187161819037082,
      "loss": 0.7736,
      "num_input_tokens_seen": 42595504,
      "step": 73375
    },
    {
      "epoch": 10.929401251117069,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.01518521231636211,
      "loss": 0.8011,
      "num_input_tokens_seen": 42598640,
      "step": 73380
    },
    {
      "epoch": 10.930145963658028,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.015183262810558162,
      "loss": 0.7846,
      "num_input_tokens_seen": 42601424,
      "step": 73385
    },
    {
      "epoch": 10.930890676198988,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.015181313301658176,
      "loss": 0.788,
      "num_input_tokens_seen": 42603952,
      "step": 73390
    },
    {
      "epoch": 10.931635388739947,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.015179363789695088,
      "loss": 0.8111,
      "num_input_tokens_seen": 42606672,
      "step": 73395
    },
    {
      "epoch": 10.932380101280906,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01517741427470183,
      "loss": 0.7905,
      "num_input_tokens_seen": 42609968,
      "step": 73400
    },
    {
      "epoch": 10.933124813821864,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.015175464756711336,
      "loss": 0.81,
      "num_input_tokens_seen": 42612752,
      "step": 73405
    },
    {
      "epoch": 10.933869526362823,
      "grad_norm": 0.02734375,
      "learning_rate": 0.01517351523575655,
      "loss": 0.8157,
      "num_input_tokens_seen": 42615536,
      "step": 73410
    },
    {
      "epoch": 10.934614238903784,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.015171565711870396,
      "loss": 0.7898,
      "num_input_tokens_seen": 42618224,
      "step": 73415
    },
    {
      "epoch": 10.935358951444742,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.01516961618508581,
      "loss": 0.8133,
      "num_input_tokens_seen": 42621168,
      "step": 73420
    },
    {
      "epoch": 10.936103663985701,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.015167666655435737,
      "loss": 0.8069,
      "num_input_tokens_seen": 42623856,
      "step": 73425
    },
    {
      "epoch": 10.93684837652666,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.015165717122953101,
      "loss": 0.7993,
      "num_input_tokens_seen": 42626576,
      "step": 73430
    },
    {
      "epoch": 10.93759308906762,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.015163767587670846,
      "loss": 0.8169,
      "num_input_tokens_seen": 42629680,
      "step": 73435
    },
    {
      "epoch": 10.93833780160858,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.015161818049621904,
      "loss": 0.8045,
      "num_input_tokens_seen": 42632400,
      "step": 73440
    },
    {
      "epoch": 10.939082514149538,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.015159868508839212,
      "loss": 0.8207,
      "num_input_tokens_seen": 42635184,
      "step": 73445
    },
    {
      "epoch": 10.939827226690497,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.015157918965355701,
      "loss": 0.7952,
      "num_input_tokens_seen": 42638256,
      "step": 73450
    },
    {
      "epoch": 10.940571939231457,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01515596941920431,
      "loss": 0.7875,
      "num_input_tokens_seen": 42640976,
      "step": 73455
    },
    {
      "epoch": 10.941316651772416,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.015154019870417973,
      "loss": 0.8008,
      "num_input_tokens_seen": 42643792,
      "step": 73460
    },
    {
      "epoch": 10.942061364313375,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.01515207031902963,
      "loss": 0.8108,
      "num_input_tokens_seen": 42646768,
      "step": 73465
    },
    {
      "epoch": 10.942806076854334,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.015150120765072208,
      "loss": 0.8,
      "num_input_tokens_seen": 42649744,
      "step": 73470
    },
    {
      "epoch": 10.943550789395294,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.015148171208578655,
      "loss": 0.7994,
      "num_input_tokens_seen": 42652336,
      "step": 73475
    },
    {
      "epoch": 10.944295501936253,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.015146221649581902,
      "loss": 0.817,
      "num_input_tokens_seen": 42655280,
      "step": 73480
    },
    {
      "epoch": 10.945040214477212,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.015144272088114875,
      "loss": 0.7754,
      "num_input_tokens_seen": 42657840,
      "step": 73485
    },
    {
      "epoch": 10.94578492701817,
      "grad_norm": 0.046875,
      "learning_rate": 0.015142322524210518,
      "loss": 0.7998,
      "num_input_tokens_seen": 42660912,
      "step": 73490
    },
    {
      "epoch": 10.946529639559131,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.015140372957901773,
      "loss": 0.8037,
      "num_input_tokens_seen": 42664112,
      "step": 73495
    },
    {
      "epoch": 10.94727435210009,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.015138423389221563,
      "loss": 0.8113,
      "num_input_tokens_seen": 42666832,
      "step": 73500
    },
    {
      "epoch": 10.948019064641048,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.015136473818202838,
      "loss": 0.7763,
      "num_input_tokens_seen": 42669712,
      "step": 73505
    },
    {
      "epoch": 10.948763777182007,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.01513452424487852,
      "loss": 0.8044,
      "num_input_tokens_seen": 42672656,
      "step": 73510
    },
    {
      "epoch": 10.949508489722968,
      "grad_norm": 0.033203125,
      "learning_rate": 0.015132574669281557,
      "loss": 0.778,
      "num_input_tokens_seen": 42675408,
      "step": 73515
    },
    {
      "epoch": 10.950253202263927,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.015130625091444874,
      "loss": 0.8068,
      "num_input_tokens_seen": 42678512,
      "step": 73520
    },
    {
      "epoch": 10.950997914804885,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.015128675511401414,
      "loss": 0.8165,
      "num_input_tokens_seen": 42681200,
      "step": 73525
    },
    {
      "epoch": 10.951742627345844,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.015126725929184114,
      "loss": 0.7782,
      "num_input_tokens_seen": 42684880,
      "step": 73530
    },
    {
      "epoch": 10.952487339886805,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.015124776344825905,
      "loss": 0.7979,
      "num_input_tokens_seen": 42687888,
      "step": 73535
    },
    {
      "epoch": 10.953232052427763,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.015122826758359733,
      "loss": 0.8233,
      "num_input_tokens_seen": 42690768,
      "step": 73540
    },
    {
      "epoch": 10.953976764968722,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01512087716981852,
      "loss": 0.784,
      "num_input_tokens_seen": 42693616,
      "step": 73545
    },
    {
      "epoch": 10.95472147750968,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.015118927579235214,
      "loss": 0.8313,
      "num_input_tokens_seen": 42696592,
      "step": 73550
    },
    {
      "epoch": 10.95546619005064,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.01511697798664275,
      "loss": 0.8025,
      "num_input_tokens_seen": 42699472,
      "step": 73555
    },
    {
      "epoch": 10.9562109025916,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.015115028392074058,
      "loss": 0.7834,
      "num_input_tokens_seen": 42702416,
      "step": 73560
    },
    {
      "epoch": 10.956955615132559,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.015113078795562078,
      "loss": 0.8121,
      "num_input_tokens_seen": 42705520,
      "step": 73565
    },
    {
      "epoch": 10.957700327673518,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.015111129197139746,
      "loss": 0.79,
      "num_input_tokens_seen": 42708368,
      "step": 73570
    },
    {
      "epoch": 10.958445040214476,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.015109179596840003,
      "loss": 0.8115,
      "num_input_tokens_seen": 42711248,
      "step": 73575
    },
    {
      "epoch": 10.959189752755437,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.015107229994695779,
      "loss": 0.7891,
      "num_input_tokens_seen": 42714256,
      "step": 73580
    },
    {
      "epoch": 10.959934465296396,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.015105280390740014,
      "loss": 0.8309,
      "num_input_tokens_seen": 42717328,
      "step": 73585
    },
    {
      "epoch": 10.960679177837354,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.015103330785005645,
      "loss": 0.8154,
      "num_input_tokens_seen": 42720528,
      "step": 73590
    },
    {
      "epoch": 10.961423890378313,
      "grad_norm": 0.05078125,
      "learning_rate": 0.01510138117752561,
      "loss": 0.8072,
      "num_input_tokens_seen": 42723472,
      "step": 73595
    },
    {
      "epoch": 10.962168602919274,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.015099431568332836,
      "loss": 0.7858,
      "num_input_tokens_seen": 42726576,
      "step": 73600
    },
    {
      "epoch": 10.962913315460233,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01509748195746027,
      "loss": 0.7912,
      "num_input_tokens_seen": 42729872,
      "step": 73605
    },
    {
      "epoch": 10.963658028001191,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.015095532344940849,
      "loss": 0.8107,
      "num_input_tokens_seen": 42732720,
      "step": 73610
    },
    {
      "epoch": 10.96440274054215,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0150935827308075,
      "loss": 0.7913,
      "num_input_tokens_seen": 42735600,
      "step": 73615
    },
    {
      "epoch": 10.96514745308311,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.015091633115093168,
      "loss": 0.8272,
      "num_input_tokens_seen": 42738576,
      "step": 73620
    },
    {
      "epoch": 10.96589216562407,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.015089683497830787,
      "loss": 0.8207,
      "num_input_tokens_seen": 42741328,
      "step": 73625
    },
    {
      "epoch": 10.966636878165028,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.015087733879053294,
      "loss": 0.8036,
      "num_input_tokens_seen": 42744336,
      "step": 73630
    },
    {
      "epoch": 10.967381590705987,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.01508578425879363,
      "loss": 0.7944,
      "num_input_tokens_seen": 42747280,
      "step": 73635
    },
    {
      "epoch": 10.968126303246947,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.015083834637084723,
      "loss": 0.8086,
      "num_input_tokens_seen": 42750512,
      "step": 73640
    },
    {
      "epoch": 10.968871015787906,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01508188501395952,
      "loss": 0.7969,
      "num_input_tokens_seen": 42753296,
      "step": 73645
    },
    {
      "epoch": 10.969615728328865,
      "grad_norm": 0.06640625,
      "learning_rate": 0.01507993538945095,
      "loss": 0.8155,
      "num_input_tokens_seen": 42756016,
      "step": 73650
    },
    {
      "epoch": 10.970360440869824,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.015077985763591956,
      "loss": 0.7943,
      "num_input_tokens_seen": 42759088,
      "step": 73655
    },
    {
      "epoch": 10.971105153410784,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.015076036136415468,
      "loss": 0.8036,
      "num_input_tokens_seen": 42761968,
      "step": 73660
    },
    {
      "epoch": 10.971849865951743,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.015074086507954424,
      "loss": 0.778,
      "num_input_tokens_seen": 42764624,
      "step": 73665
    },
    {
      "epoch": 10.972594578492702,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.015072136878241773,
      "loss": 0.7996,
      "num_input_tokens_seen": 42767536,
      "step": 73670
    },
    {
      "epoch": 10.97333929103366,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.015070187247310435,
      "loss": 0.7938,
      "num_input_tokens_seen": 42770480,
      "step": 73675
    },
    {
      "epoch": 10.974084003574621,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.01506823761519336,
      "loss": 0.7996,
      "num_input_tokens_seen": 42773360,
      "step": 73680
    },
    {
      "epoch": 10.97482871611558,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.015066287981923478,
      "loss": 0.7893,
      "num_input_tokens_seen": 42776112,
      "step": 73685
    },
    {
      "epoch": 10.975573428656539,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.015064338347533726,
      "loss": 0.8234,
      "num_input_tokens_seen": 42779216,
      "step": 73690
    },
    {
      "epoch": 10.976318141197497,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.015062388712057041,
      "loss": 0.7991,
      "num_input_tokens_seen": 42782288,
      "step": 73695
    },
    {
      "epoch": 10.977062853738456,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.015060439075526367,
      "loss": 0.7837,
      "num_input_tokens_seen": 42785072,
      "step": 73700
    },
    {
      "epoch": 10.977807566279417,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.015058489437974635,
      "loss": 0.7828,
      "num_input_tokens_seen": 42787888,
      "step": 73705
    },
    {
      "epoch": 10.978552278820375,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.015056539799434785,
      "loss": 0.8325,
      "num_input_tokens_seen": 42791024,
      "step": 73710
    },
    {
      "epoch": 10.979296991361334,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01505459015993975,
      "loss": 0.7618,
      "num_input_tokens_seen": 42794032,
      "step": 73715
    },
    {
      "epoch": 10.980041703902295,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.015052640519522471,
      "loss": 0.7853,
      "num_input_tokens_seen": 42796688,
      "step": 73720
    },
    {
      "epoch": 10.980786416443253,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.01505069087821588,
      "loss": 0.8105,
      "num_input_tokens_seen": 42799760,
      "step": 73725
    },
    {
      "epoch": 10.981531128984212,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.015048741236052921,
      "loss": 0.7796,
      "num_input_tokens_seen": 42802576,
      "step": 73730
    },
    {
      "epoch": 10.982275841525171,
      "grad_norm": 0.044921875,
      "learning_rate": 0.01504679159306653,
      "loss": 0.7773,
      "num_input_tokens_seen": 42805456,
      "step": 73735
    },
    {
      "epoch": 10.98302055406613,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.015044841949289641,
      "loss": 0.7715,
      "num_input_tokens_seen": 42808336,
      "step": 73740
    },
    {
      "epoch": 10.98376526660709,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.015042892304755192,
      "loss": 0.7917,
      "num_input_tokens_seen": 42811376,
      "step": 73745
    },
    {
      "epoch": 10.984509979148049,
      "grad_norm": 0.037109375,
      "learning_rate": 0.015040942659496122,
      "loss": 0.762,
      "num_input_tokens_seen": 42814352,
      "step": 73750
    },
    {
      "epoch": 10.985254691689008,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01503899301354537,
      "loss": 0.764,
      "num_input_tokens_seen": 42817168,
      "step": 73755
    },
    {
      "epoch": 10.985999404229966,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.015037043366935867,
      "loss": 0.8561,
      "num_input_tokens_seen": 42819952,
      "step": 73760
    },
    {
      "epoch": 10.986744116770927,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.015035093719700555,
      "loss": 0.8002,
      "num_input_tokens_seen": 42823056,
      "step": 73765
    },
    {
      "epoch": 10.987488829311886,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.015033144071872371,
      "loss": 0.7928,
      "num_input_tokens_seen": 42826000,
      "step": 73770
    },
    {
      "epoch": 10.988233541852845,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.015031194423484254,
      "loss": 0.7373,
      "num_input_tokens_seen": 42828912,
      "step": 73775
    },
    {
      "epoch": 10.988978254393803,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.015029244774569133,
      "loss": 0.7872,
      "num_input_tokens_seen": 42832016,
      "step": 73780
    },
    {
      "epoch": 10.989722966934764,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.015027295125159953,
      "loss": 0.811,
      "num_input_tokens_seen": 42834736,
      "step": 73785
    },
    {
      "epoch": 10.990467679475723,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.015025345475289652,
      "loss": 0.7884,
      "num_input_tokens_seen": 42837776,
      "step": 73790
    },
    {
      "epoch": 10.991212392016681,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.015023395824991168,
      "loss": 0.8083,
      "num_input_tokens_seen": 42840400,
      "step": 73795
    },
    {
      "epoch": 10.99195710455764,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.015021446174297428,
      "loss": 0.8184,
      "num_input_tokens_seen": 42843408,
      "step": 73800
    },
    {
      "epoch": 10.9927018170986,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.01501949652324138,
      "loss": 0.7914,
      "num_input_tokens_seen": 42846384,
      "step": 73805
    },
    {
      "epoch": 10.99344652963956,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.01501754687185596,
      "loss": 0.7948,
      "num_input_tokens_seen": 42849264,
      "step": 73810
    },
    {
      "epoch": 10.994191242180518,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0150155972201741,
      "loss": 0.8442,
      "num_input_tokens_seen": 42852080,
      "step": 73815
    },
    {
      "epoch": 10.994935954721477,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.015013647568228742,
      "loss": 0.7689,
      "num_input_tokens_seen": 42855504,
      "step": 73820
    },
    {
      "epoch": 10.995680667262437,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.015011697916052821,
      "loss": 0.8165,
      "num_input_tokens_seen": 42858832,
      "step": 73825
    },
    {
      "epoch": 10.996425379803396,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.015009748263679276,
      "loss": 0.8111,
      "num_input_tokens_seen": 42861712,
      "step": 73830
    },
    {
      "epoch": 10.997170092344355,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.015007798611141047,
      "loss": 0.7984,
      "num_input_tokens_seen": 42864368,
      "step": 73835
    },
    {
      "epoch": 10.997914804885314,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.015005848958471065,
      "loss": 0.8032,
      "num_input_tokens_seen": 42867184,
      "step": 73840
    },
    {
      "epoch": 10.998659517426274,
      "grad_norm": 0.033203125,
      "learning_rate": 0.015003899305702275,
      "loss": 0.8066,
      "num_input_tokens_seen": 42870352,
      "step": 73845
    },
    {
      "epoch": 10.999404229967233,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.015001949652867606,
      "loss": 0.7992,
      "num_input_tokens_seen": 42873520,
      "step": 73850
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.8003315925598145,
      "eval_runtime": 70.6735,
      "eval_samples_per_second": 42.222,
      "eval_steps_per_second": 10.556,
      "num_input_tokens_seen": 42875264,
      "step": 73854
    },
    {
      "epoch": 11.000148942508192,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.015,
      "loss": 0.7945,
      "num_input_tokens_seen": 42875808,
      "step": 73855
    },
    {
      "epoch": 11.00089365504915,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.014998050347132398,
      "loss": 0.7884,
      "num_input_tokens_seen": 42878944,
      "step": 73860
    },
    {
      "epoch": 11.001638367590111,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.014996100694297727,
      "loss": 0.8115,
      "num_input_tokens_seen": 42881696,
      "step": 73865
    },
    {
      "epoch": 11.00238308013107,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.014994151041528936,
      "loss": 0.7803,
      "num_input_tokens_seen": 42884480,
      "step": 73870
    },
    {
      "epoch": 11.003127792672029,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.014992201388858954,
      "loss": 0.7922,
      "num_input_tokens_seen": 42887488,
      "step": 73875
    },
    {
      "epoch": 11.003872505212987,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.014990251736320724,
      "loss": 0.7991,
      "num_input_tokens_seen": 42890112,
      "step": 73880
    },
    {
      "epoch": 11.004617217753948,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.014988302083947178,
      "loss": 0.8083,
      "num_input_tokens_seen": 42893248,
      "step": 73885
    },
    {
      "epoch": 11.005361930294907,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.014986352431771259,
      "loss": 0.7776,
      "num_input_tokens_seen": 42895744,
      "step": 73890
    },
    {
      "epoch": 11.006106642835865,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.014984402779825899,
      "loss": 0.7946,
      "num_input_tokens_seen": 42898528,
      "step": 73895
    },
    {
      "epoch": 11.006851355376824,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.014982453128144042,
      "loss": 0.7931,
      "num_input_tokens_seen": 42901600,
      "step": 73900
    },
    {
      "epoch": 11.007596067917783,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.014980503476758621,
      "loss": 0.7777,
      "num_input_tokens_seen": 42904768,
      "step": 73905
    },
    {
      "epoch": 11.008340780458743,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.014978553825702573,
      "loss": 0.773,
      "num_input_tokens_seen": 42907584,
      "step": 73910
    },
    {
      "epoch": 11.009085492999702,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.014976604175008838,
      "loss": 0.8241,
      "num_input_tokens_seen": 42910848,
      "step": 73915
    },
    {
      "epoch": 11.009830205540661,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.014974654524710347,
      "loss": 0.8103,
      "num_input_tokens_seen": 42913984,
      "step": 73920
    },
    {
      "epoch": 11.01057491808162,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.014972704874840046,
      "loss": 0.776,
      "num_input_tokens_seen": 42917376,
      "step": 73925
    },
    {
      "epoch": 11.01131963062258,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.014970755225430862,
      "loss": 0.8131,
      "num_input_tokens_seen": 42920448,
      "step": 73930
    },
    {
      "epoch": 11.012064343163539,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.014968805576515747,
      "loss": 0.7924,
      "num_input_tokens_seen": 42923584,
      "step": 73935
    },
    {
      "epoch": 11.012809055704498,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01496685592812763,
      "loss": 0.7985,
      "num_input_tokens_seen": 42926144,
      "step": 73940
    },
    {
      "epoch": 11.013553768245457,
      "grad_norm": 0.07421875,
      "learning_rate": 0.014964906280299445,
      "loss": 0.8169,
      "num_input_tokens_seen": 42929216,
      "step": 73945
    },
    {
      "epoch": 11.014298480786417,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.014962956633064135,
      "loss": 0.8023,
      "num_input_tokens_seen": 42932192,
      "step": 73950
    },
    {
      "epoch": 11.015043193327376,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.014961006986454629,
      "loss": 0.8189,
      "num_input_tokens_seen": 42935104,
      "step": 73955
    },
    {
      "epoch": 11.015787905868335,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.014959057340503878,
      "loss": 0.8033,
      "num_input_tokens_seen": 42937920,
      "step": 73960
    },
    {
      "epoch": 11.016532618409293,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.014957107695244803,
      "loss": 0.7788,
      "num_input_tokens_seen": 42940608,
      "step": 73965
    },
    {
      "epoch": 11.017277330950254,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.014955158050710357,
      "loss": 0.7906,
      "num_input_tokens_seen": 42943840,
      "step": 73970
    },
    {
      "epoch": 11.018022043491213,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.014953208406933469,
      "loss": 0.7618,
      "num_input_tokens_seen": 42946592,
      "step": 73975
    },
    {
      "epoch": 11.018766756032171,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.014951258763947078,
      "loss": 0.7984,
      "num_input_tokens_seen": 42949760,
      "step": 73980
    },
    {
      "epoch": 11.01951146857313,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.014949309121784122,
      "loss": 0.814,
      "num_input_tokens_seen": 42952864,
      "step": 73985
    },
    {
      "epoch": 11.02025618111409,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.014947359480477531,
      "loss": 0.83,
      "num_input_tokens_seen": 42955616,
      "step": 73990
    },
    {
      "epoch": 11.02100089365505,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.014945409840060254,
      "loss": 0.7853,
      "num_input_tokens_seen": 42958560,
      "step": 73995
    },
    {
      "epoch": 11.021745606196008,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.014943460200565218,
      "loss": 0.7856,
      "num_input_tokens_seen": 42961344,
      "step": 74000
    },
    {
      "epoch": 11.022490318736967,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.014941510562025365,
      "loss": 0.8045,
      "num_input_tokens_seen": 42964192,
      "step": 74005
    },
    {
      "epoch": 11.023235031277927,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01493956092447363,
      "loss": 0.8054,
      "num_input_tokens_seen": 42967040,
      "step": 74010
    },
    {
      "epoch": 11.023979743818886,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.014937611287942958,
      "loss": 0.7795,
      "num_input_tokens_seen": 42970208,
      "step": 74015
    },
    {
      "epoch": 11.024724456359845,
      "grad_norm": 0.0390625,
      "learning_rate": 0.01493566165246628,
      "loss": 0.8187,
      "num_input_tokens_seen": 42973024,
      "step": 74020
    },
    {
      "epoch": 11.025469168900804,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.014933712018076523,
      "loss": 0.7955,
      "num_input_tokens_seen": 42976000,
      "step": 74025
    },
    {
      "epoch": 11.026213881441764,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.014931762384806644,
      "loss": 0.819,
      "num_input_tokens_seen": 42979040,
      "step": 74030
    },
    {
      "epoch": 11.026958593982723,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.014929812752689566,
      "loss": 0.7991,
      "num_input_tokens_seen": 42981888,
      "step": 74035
    },
    {
      "epoch": 11.027703306523682,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.014927863121758228,
      "loss": 0.7971,
      "num_input_tokens_seen": 42984672,
      "step": 74040
    },
    {
      "epoch": 11.02844801906464,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.014925913492045572,
      "loss": 0.8239,
      "num_input_tokens_seen": 42987616,
      "step": 74045
    },
    {
      "epoch": 11.029192731605601,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.014923963863584534,
      "loss": 0.7809,
      "num_input_tokens_seen": 42990496,
      "step": 74050
    },
    {
      "epoch": 11.02993744414656,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.014922014236408043,
      "loss": 0.7829,
      "num_input_tokens_seen": 42993280,
      "step": 74055
    },
    {
      "epoch": 11.030682156687519,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.014920064610549049,
      "loss": 0.8041,
      "num_input_tokens_seen": 42996352,
      "step": 74060
    },
    {
      "epoch": 11.031426869228477,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.014918114986040483,
      "loss": 0.7945,
      "num_input_tokens_seen": 42999072,
      "step": 74065
    },
    {
      "epoch": 11.032171581769436,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.014916165362915276,
      "loss": 0.8068,
      "num_input_tokens_seen": 43001824,
      "step": 74070
    },
    {
      "epoch": 11.032916294310397,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.014914215741206376,
      "loss": 0.7863,
      "num_input_tokens_seen": 43004544,
      "step": 74075
    },
    {
      "epoch": 11.033661006851355,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.014912266120946704,
      "loss": 0.8095,
      "num_input_tokens_seen": 43007584,
      "step": 74080
    },
    {
      "epoch": 11.034405719392314,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.014910316502169216,
      "loss": 0.785,
      "num_input_tokens_seen": 43010208,
      "step": 74085
    },
    {
      "epoch": 11.035150431933273,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.014908366884906831,
      "loss": 0.8176,
      "num_input_tokens_seen": 43013216,
      "step": 74090
    },
    {
      "epoch": 11.035895144474233,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.014906417269192502,
      "loss": 0.7998,
      "num_input_tokens_seen": 43016096,
      "step": 74095
    },
    {
      "epoch": 11.036639857015192,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.014904467655059157,
      "loss": 0.8031,
      "num_input_tokens_seen": 43018528,
      "step": 74100
    },
    {
      "epoch": 11.037384569556151,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.014902518042539732,
      "loss": 0.7853,
      "num_input_tokens_seen": 43021408,
      "step": 74105
    },
    {
      "epoch": 11.03812928209711,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.014900568431667168,
      "loss": 0.8338,
      "num_input_tokens_seen": 43024032,
      "step": 74110
    },
    {
      "epoch": 11.03887399463807,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.014898618822474393,
      "loss": 0.8111,
      "num_input_tokens_seen": 43026880,
      "step": 74115
    },
    {
      "epoch": 11.039618707179029,
      "grad_norm": 0.056640625,
      "learning_rate": 0.014896669214994357,
      "loss": 0.8019,
      "num_input_tokens_seen": 43029952,
      "step": 74120
    },
    {
      "epoch": 11.040363419719988,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.014894719609259983,
      "loss": 0.8546,
      "num_input_tokens_seen": 43032800,
      "step": 74125
    },
    {
      "epoch": 11.041108132260947,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.01489277000530422,
      "loss": 0.7957,
      "num_input_tokens_seen": 43035648,
      "step": 74130
    },
    {
      "epoch": 11.041852844801907,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.014890820403159997,
      "loss": 0.8182,
      "num_input_tokens_seen": 43038368,
      "step": 74135
    },
    {
      "epoch": 11.042597557342866,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.014888870802860254,
      "loss": 0.7925,
      "num_input_tokens_seen": 43041056,
      "step": 74140
    },
    {
      "epoch": 11.043342269883825,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.014886921204437926,
      "loss": 0.8006,
      "num_input_tokens_seen": 43043872,
      "step": 74145
    },
    {
      "epoch": 11.044086982424783,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.014884971607925943,
      "loss": 0.8064,
      "num_input_tokens_seen": 43046912,
      "step": 74150
    },
    {
      "epoch": 11.044831694965744,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.014883022013357254,
      "loss": 0.795,
      "num_input_tokens_seen": 43049600,
      "step": 74155
    },
    {
      "epoch": 11.045576407506703,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.014881072420764783,
      "loss": 0.799,
      "num_input_tokens_seen": 43052512,
      "step": 74160
    },
    {
      "epoch": 11.046321120047661,
      "grad_norm": 0.046875,
      "learning_rate": 0.014879122830181479,
      "loss": 0.8123,
      "num_input_tokens_seen": 43055744,
      "step": 74165
    },
    {
      "epoch": 11.04706583258862,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.014877173241640269,
      "loss": 0.796,
      "num_input_tokens_seen": 43058976,
      "step": 74170
    },
    {
      "epoch": 11.04781054512958,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.014875223655174095,
      "loss": 0.7973,
      "num_input_tokens_seen": 43062112,
      "step": 74175
    },
    {
      "epoch": 11.04855525767054,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.01487327407081589,
      "loss": 0.805,
      "num_input_tokens_seen": 43064992,
      "step": 74180
    },
    {
      "epoch": 11.049299970211498,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.014871324488598587,
      "loss": 0.7835,
      "num_input_tokens_seen": 43067872,
      "step": 74185
    },
    {
      "epoch": 11.050044682752457,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01486937490855513,
      "loss": 0.801,
      "num_input_tokens_seen": 43070752,
      "step": 74190
    },
    {
      "epoch": 11.050789395293418,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.014867425330718444,
      "loss": 0.8043,
      "num_input_tokens_seen": 43073568,
      "step": 74195
    },
    {
      "epoch": 11.051534107834376,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.01486547575512148,
      "loss": 0.7859,
      "num_input_tokens_seen": 43076320,
      "step": 74200
    },
    {
      "epoch": 11.052278820375335,
      "grad_norm": 0.033203125,
      "learning_rate": 0.014863526181797164,
      "loss": 0.8017,
      "num_input_tokens_seen": 43078976,
      "step": 74205
    },
    {
      "epoch": 11.053023532916294,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.014861576610778438,
      "loss": 0.7869,
      "num_input_tokens_seen": 43081696,
      "step": 74210
    },
    {
      "epoch": 11.053768245457254,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.014859627042098226,
      "loss": 0.8036,
      "num_input_tokens_seen": 43084768,
      "step": 74215
    },
    {
      "epoch": 11.054512957998213,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.01485767747578948,
      "loss": 0.8031,
      "num_input_tokens_seen": 43087520,
      "step": 74220
    },
    {
      "epoch": 11.055257670539172,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.014855727911885129,
      "loss": 0.8086,
      "num_input_tokens_seen": 43090080,
      "step": 74225
    },
    {
      "epoch": 11.05600238308013,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0148537783504181,
      "loss": 0.7937,
      "num_input_tokens_seen": 43092960,
      "step": 74230
    },
    {
      "epoch": 11.056747095621091,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.014851828791421344,
      "loss": 0.8177,
      "num_input_tokens_seen": 43095584,
      "step": 74235
    },
    {
      "epoch": 11.05749180816205,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.014849879234927789,
      "loss": 0.7888,
      "num_input_tokens_seen": 43098336,
      "step": 74240
    },
    {
      "epoch": 11.058236520703009,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.014847929680970373,
      "loss": 0.7802,
      "num_input_tokens_seen": 43101088,
      "step": 74245
    },
    {
      "epoch": 11.058981233243967,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.014845980129582024,
      "loss": 0.7888,
      "num_input_tokens_seen": 43103968,
      "step": 74250
    },
    {
      "epoch": 11.059725945784926,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01484403058079569,
      "loss": 0.7816,
      "num_input_tokens_seen": 43106592,
      "step": 74255
    },
    {
      "epoch": 11.060470658325887,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.014842081034644303,
      "loss": 0.7858,
      "num_input_tokens_seen": 43109280,
      "step": 74260
    },
    {
      "epoch": 11.061215370866845,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.014840131491160789,
      "loss": 0.8063,
      "num_input_tokens_seen": 43112000,
      "step": 74265
    },
    {
      "epoch": 11.061960083407804,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.014838181950378098,
      "loss": 0.8144,
      "num_input_tokens_seen": 43114848,
      "step": 74270
    },
    {
      "epoch": 11.062704795948763,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.014836232412329155,
      "loss": 0.8139,
      "num_input_tokens_seen": 43117920,
      "step": 74275
    },
    {
      "epoch": 11.063449508489724,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.014834282877046901,
      "loss": 0.7859,
      "num_input_tokens_seen": 43120544,
      "step": 74280
    },
    {
      "epoch": 11.064194221030682,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.014832333344564262,
      "loss": 0.8128,
      "num_input_tokens_seen": 43123520,
      "step": 74285
    },
    {
      "epoch": 11.064938933571641,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.014830383814914189,
      "loss": 0.8086,
      "num_input_tokens_seen": 43126592,
      "step": 74290
    },
    {
      "epoch": 11.0656836461126,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.014828434288129603,
      "loss": 0.8021,
      "num_input_tokens_seen": 43129600,
      "step": 74295
    },
    {
      "epoch": 11.06642835865356,
      "grad_norm": 0.087890625,
      "learning_rate": 0.014826484764243452,
      "loss": 0.7913,
      "num_input_tokens_seen": 43132384,
      "step": 74300
    },
    {
      "epoch": 11.067173071194519,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.014824535243288663,
      "loss": 0.7882,
      "num_input_tokens_seen": 43134944,
      "step": 74305
    },
    {
      "epoch": 11.067917783735478,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.014822585725298171,
      "loss": 0.7896,
      "num_input_tokens_seen": 43137856,
      "step": 74310
    },
    {
      "epoch": 11.068662496276437,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.014820636210304914,
      "loss": 0.7985,
      "num_input_tokens_seen": 43140512,
      "step": 74315
    },
    {
      "epoch": 11.069407208817397,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.014818686698341821,
      "loss": 0.8082,
      "num_input_tokens_seen": 43143200,
      "step": 74320
    },
    {
      "epoch": 11.070151921358356,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.014816737189441839,
      "loss": 0.7929,
      "num_input_tokens_seen": 43146208,
      "step": 74325
    },
    {
      "epoch": 11.070896633899315,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.014814787683637892,
      "loss": 0.7955,
      "num_input_tokens_seen": 43149056,
      "step": 74330
    },
    {
      "epoch": 11.071641346440273,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.014812838180962918,
      "loss": 0.8009,
      "num_input_tokens_seen": 43151744,
      "step": 74335
    },
    {
      "epoch": 11.072386058981234,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.014810888681449858,
      "loss": 0.7763,
      "num_input_tokens_seen": 43155040,
      "step": 74340
    },
    {
      "epoch": 11.073130771522193,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.014808939185131636,
      "loss": 0.7866,
      "num_input_tokens_seen": 43158144,
      "step": 74345
    },
    {
      "epoch": 11.073875484063151,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.014806989692041196,
      "loss": 0.8134,
      "num_input_tokens_seen": 43161056,
      "step": 74350
    },
    {
      "epoch": 11.07462019660411,
      "grad_norm": 0.033203125,
      "learning_rate": 0.014805040202211465,
      "loss": 0.792,
      "num_input_tokens_seen": 43164288,
      "step": 74355
    },
    {
      "epoch": 11.07536490914507,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.014803090715675385,
      "loss": 0.8011,
      "num_input_tokens_seen": 43166848,
      "step": 74360
    },
    {
      "epoch": 11.07610962168603,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.014801141232465885,
      "loss": 0.769,
      "num_input_tokens_seen": 43169600,
      "step": 74365
    },
    {
      "epoch": 11.076854334226988,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.014799191752615902,
      "loss": 0.8185,
      "num_input_tokens_seen": 43172288,
      "step": 74370
    },
    {
      "epoch": 11.077599046767947,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.01479724227615837,
      "loss": 0.7958,
      "num_input_tokens_seen": 43175168,
      "step": 74375
    },
    {
      "epoch": 11.078343759308908,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.014795292803126227,
      "loss": 0.805,
      "num_input_tokens_seen": 43178304,
      "step": 74380
    },
    {
      "epoch": 11.079088471849866,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.014793343333552403,
      "loss": 0.7946,
      "num_input_tokens_seen": 43181184,
      "step": 74385
    },
    {
      "epoch": 11.079833184390825,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.014791393867469829,
      "loss": 0.7889,
      "num_input_tokens_seen": 43184064,
      "step": 74390
    },
    {
      "epoch": 11.080577896931784,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.014789444404911449,
      "loss": 0.7931,
      "num_input_tokens_seen": 43186976,
      "step": 74395
    },
    {
      "epoch": 11.081322609472744,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.014787494945910189,
      "loss": 0.8035,
      "num_input_tokens_seen": 43189888,
      "step": 74400
    },
    {
      "epoch": 11.082067322013703,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.014785545490498988,
      "loss": 0.7839,
      "num_input_tokens_seen": 43192896,
      "step": 74405
    },
    {
      "epoch": 11.082812034554662,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.014783596038710772,
      "loss": 0.8367,
      "num_input_tokens_seen": 43195936,
      "step": 74410
    },
    {
      "epoch": 11.08355674709562,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.014781646590578488,
      "loss": 0.8122,
      "num_input_tokens_seen": 43198496,
      "step": 74415
    },
    {
      "epoch": 11.08430145963658,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.014779697146135062,
      "loss": 0.8236,
      "num_input_tokens_seen": 43201344,
      "step": 74420
    },
    {
      "epoch": 11.08504617217754,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.014777747705413423,
      "loss": 0.7974,
      "num_input_tokens_seen": 43204192,
      "step": 74425
    },
    {
      "epoch": 11.085790884718499,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.014775798268446516,
      "loss": 0.8084,
      "num_input_tokens_seen": 43207168,
      "step": 74430
    },
    {
      "epoch": 11.086535597259457,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.014773848835267268,
      "loss": 0.8024,
      "num_input_tokens_seen": 43210464,
      "step": 74435
    },
    {
      "epoch": 11.087280309800416,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.014771899405908616,
      "loss": 0.7992,
      "num_input_tokens_seen": 43213344,
      "step": 74440
    },
    {
      "epoch": 11.088025022341377,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.014769949980403487,
      "loss": 0.807,
      "num_input_tokens_seen": 43216096,
      "step": 74445
    },
    {
      "epoch": 11.088769734882336,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.014768000558784824,
      "loss": 0.8019,
      "num_input_tokens_seen": 43219136,
      "step": 74450
    },
    {
      "epoch": 11.089514447423294,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01476605114108555,
      "loss": 0.7995,
      "num_input_tokens_seen": 43222304,
      "step": 74455
    },
    {
      "epoch": 11.090259159964253,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.01476410172733861,
      "loss": 0.8046,
      "num_input_tokens_seen": 43224960,
      "step": 74460
    },
    {
      "epoch": 11.091003872505214,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.014762152317576932,
      "loss": 0.8237,
      "num_input_tokens_seen": 43227936,
      "step": 74465
    },
    {
      "epoch": 11.091748585046172,
      "grad_norm": 0.025390625,
      "learning_rate": 0.014760202911833447,
      "loss": 0.8071,
      "num_input_tokens_seen": 43230816,
      "step": 74470
    },
    {
      "epoch": 11.092493297587131,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.01475825351014109,
      "loss": 0.8079,
      "num_input_tokens_seen": 43233504,
      "step": 74475
    },
    {
      "epoch": 11.09323801012809,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.014756304112532792,
      "loss": 0.8134,
      "num_input_tokens_seen": 43236096,
      "step": 74480
    },
    {
      "epoch": 11.09398272266905,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.014754354719041492,
      "loss": 0.8211,
      "num_input_tokens_seen": 43239104,
      "step": 74485
    },
    {
      "epoch": 11.09472743521001,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.014752405329700112,
      "loss": 0.8026,
      "num_input_tokens_seen": 43242016,
      "step": 74490
    },
    {
      "epoch": 11.095472147750968,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0147504559445416,
      "loss": 0.8048,
      "num_input_tokens_seen": 43244864,
      "step": 74495
    },
    {
      "epoch": 11.096216860291927,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.014748506563598877,
      "loss": 0.7964,
      "num_input_tokens_seen": 43247904,
      "step": 74500
    },
    {
      "epoch": 11.096961572832887,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.014746557186904886,
      "loss": 0.8108,
      "num_input_tokens_seen": 43250848,
      "step": 74505
    },
    {
      "epoch": 11.097706285373846,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.014744607814492551,
      "loss": 0.7994,
      "num_input_tokens_seen": 43253504,
      "step": 74510
    },
    {
      "epoch": 11.098450997914805,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.014742658446394802,
      "loss": 0.7985,
      "num_input_tokens_seen": 43256384,
      "step": 74515
    },
    {
      "epoch": 11.099195710455763,
      "grad_norm": 0.0703125,
      "learning_rate": 0.014740709082644583,
      "loss": 0.8125,
      "num_input_tokens_seen": 43259040,
      "step": 74520
    },
    {
      "epoch": 11.099940422996724,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.014738759723274816,
      "loss": 0.8111,
      "num_input_tokens_seen": 43263104,
      "step": 74525
    },
    {
      "epoch": 11.100685135537683,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.014736810368318443,
      "loss": 0.815,
      "num_input_tokens_seen": 43265920,
      "step": 74530
    },
    {
      "epoch": 11.101429848078642,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.014734861017808389,
      "loss": 0.8144,
      "num_input_tokens_seen": 43269024,
      "step": 74535
    },
    {
      "epoch": 11.1021745606196,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01473291167177759,
      "loss": 0.7996,
      "num_input_tokens_seen": 43271744,
      "step": 74540
    },
    {
      "epoch": 11.10291927316056,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.014730962330258978,
      "loss": 0.8007,
      "num_input_tokens_seen": 43274624,
      "step": 74545
    },
    {
      "epoch": 11.10366398570152,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.014729012993285476,
      "loss": 0.8001,
      "num_input_tokens_seen": 43277472,
      "step": 74550
    },
    {
      "epoch": 11.104408698242478,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.014727063660890032,
      "loss": 0.7821,
      "num_input_tokens_seen": 43280704,
      "step": 74555
    },
    {
      "epoch": 11.105153410783437,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.014725114333105565,
      "loss": 0.7858,
      "num_input_tokens_seen": 43283456,
      "step": 74560
    },
    {
      "epoch": 11.105898123324398,
      "grad_norm": 0.04296875,
      "learning_rate": 0.014723165009965017,
      "loss": 0.7916,
      "num_input_tokens_seen": 43286208,
      "step": 74565
    },
    {
      "epoch": 11.106642835865356,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01472121569150131,
      "loss": 0.8014,
      "num_input_tokens_seen": 43288992,
      "step": 74570
    },
    {
      "epoch": 11.107387548406315,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.014719266377747385,
      "loss": 0.8086,
      "num_input_tokens_seen": 43292032,
      "step": 74575
    },
    {
      "epoch": 11.108132260947274,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.014717317068736163,
      "loss": 0.8109,
      "num_input_tokens_seen": 43295264,
      "step": 74580
    },
    {
      "epoch": 11.108876973488233,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.014715367764500588,
      "loss": 0.7885,
      "num_input_tokens_seen": 43297824,
      "step": 74585
    },
    {
      "epoch": 11.109621686029193,
      "grad_norm": 0.029296875,
      "learning_rate": 0.014713418465073588,
      "loss": 0.8035,
      "num_input_tokens_seen": 43300480,
      "step": 74590
    },
    {
      "epoch": 11.110366398570152,
      "grad_norm": 0.037109375,
      "learning_rate": 0.014711469170488082,
      "loss": 0.7878,
      "num_input_tokens_seen": 43303360,
      "step": 74595
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.014709519880777018,
      "loss": 0.7807,
      "num_input_tokens_seen": 43306176,
      "step": 74600
    },
    {
      "epoch": 11.11185582365207,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.014707570595973317,
      "loss": 0.7966,
      "num_input_tokens_seen": 43308992,
      "step": 74605
    },
    {
      "epoch": 11.11260053619303,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.014705621316109918,
      "loss": 0.8206,
      "num_input_tokens_seen": 43311808,
      "step": 74610
    },
    {
      "epoch": 11.113345248733989,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.01470367204121974,
      "loss": 0.8171,
      "num_input_tokens_seen": 43314368,
      "step": 74615
    },
    {
      "epoch": 11.114089961274948,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.014701722771335728,
      "loss": 0.8017,
      "num_input_tokens_seen": 43317568,
      "step": 74620
    },
    {
      "epoch": 11.114834673815906,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.014699773506490809,
      "loss": 0.8229,
      "num_input_tokens_seen": 43320416,
      "step": 74625
    },
    {
      "epoch": 11.115579386356867,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.014697824246717903,
      "loss": 0.7924,
      "num_input_tokens_seen": 43323296,
      "step": 74630
    },
    {
      "epoch": 11.116324098897826,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.014695874992049957,
      "loss": 0.8124,
      "num_input_tokens_seen": 43326400,
      "step": 74635
    },
    {
      "epoch": 11.117068811438784,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.014693925742519888,
      "loss": 0.7979,
      "num_input_tokens_seen": 43329248,
      "step": 74640
    },
    {
      "epoch": 11.117813523979743,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.014691976498160638,
      "loss": 0.7962,
      "num_input_tokens_seen": 43332320,
      "step": 74645
    },
    {
      "epoch": 11.118558236520704,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.014690027259005124,
      "loss": 0.8042,
      "num_input_tokens_seen": 43335392,
      "step": 74650
    },
    {
      "epoch": 11.119302949061662,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.014688078025086294,
      "loss": 0.7841,
      "num_input_tokens_seen": 43338464,
      "step": 74655
    },
    {
      "epoch": 11.120047661602621,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.014686128796437064,
      "loss": 0.8082,
      "num_input_tokens_seen": 43341152,
      "step": 74660
    },
    {
      "epoch": 11.12079237414358,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.014684179573090367,
      "loss": 0.8213,
      "num_input_tokens_seen": 43343968,
      "step": 74665
    },
    {
      "epoch": 11.12153708668454,
      "grad_norm": 0.0390625,
      "learning_rate": 0.014682230355079139,
      "loss": 0.8027,
      "num_input_tokens_seen": 43346752,
      "step": 74670
    },
    {
      "epoch": 11.1222817992255,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.014680281142436304,
      "loss": 0.8037,
      "num_input_tokens_seen": 43349600,
      "step": 74675
    },
    {
      "epoch": 11.123026511766458,
      "grad_norm": 0.017822265625,
      "learning_rate": 0.014678331935194798,
      "loss": 0.7858,
      "num_input_tokens_seen": 43352576,
      "step": 74680
    },
    {
      "epoch": 11.123771224307417,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.01467638273338754,
      "loss": 0.8079,
      "num_input_tokens_seen": 43355712,
      "step": 74685
    },
    {
      "epoch": 11.124515936848377,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.014674433537047471,
      "loss": 0.7989,
      "num_input_tokens_seen": 43358368,
      "step": 74690
    },
    {
      "epoch": 11.125260649389336,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.014672484346207517,
      "loss": 0.7883,
      "num_input_tokens_seen": 43361088,
      "step": 74695
    },
    {
      "epoch": 11.126005361930295,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.014670535160900606,
      "loss": 0.8195,
      "num_input_tokens_seen": 43364352,
      "step": 74700
    },
    {
      "epoch": 11.126750074471254,
      "grad_norm": 0.048828125,
      "learning_rate": 0.01466858598115967,
      "loss": 0.8031,
      "num_input_tokens_seen": 43367264,
      "step": 74705
    },
    {
      "epoch": 11.127494787012214,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.014666636807017635,
      "loss": 0.7974,
      "num_input_tokens_seen": 43370464,
      "step": 74710
    },
    {
      "epoch": 11.128239499553173,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.014664687638507435,
      "loss": 0.7883,
      "num_input_tokens_seen": 43373344,
      "step": 74715
    },
    {
      "epoch": 11.128984212094132,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.01466273847566199,
      "loss": 0.782,
      "num_input_tokens_seen": 43376224,
      "step": 74720
    },
    {
      "epoch": 11.12972892463509,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.014660789318514243,
      "loss": 0.7901,
      "num_input_tokens_seen": 43379232,
      "step": 74725
    },
    {
      "epoch": 11.13047363717605,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.014658840167097113,
      "loss": 0.7965,
      "num_input_tokens_seen": 43381984,
      "step": 74730
    },
    {
      "epoch": 11.13121834971701,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.014656891021443533,
      "loss": 0.8097,
      "num_input_tokens_seen": 43384928,
      "step": 74735
    },
    {
      "epoch": 11.131963062257968,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.014654941881586427,
      "loss": 0.8052,
      "num_input_tokens_seen": 43387840,
      "step": 74740
    },
    {
      "epoch": 11.132707774798927,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.01465299274755873,
      "loss": 0.8011,
      "num_input_tokens_seen": 43390656,
      "step": 74745
    },
    {
      "epoch": 11.133452487339888,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.01465104361939337,
      "loss": 0.8098,
      "num_input_tokens_seen": 43393856,
      "step": 74750
    },
    {
      "epoch": 11.134197199880846,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.014649094497123266,
      "loss": 0.8095,
      "num_input_tokens_seen": 43396800,
      "step": 74755
    },
    {
      "epoch": 11.134941912421805,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.014647145380781362,
      "loss": 0.8082,
      "num_input_tokens_seen": 43399488,
      "step": 74760
    },
    {
      "epoch": 11.135686624962764,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.014645196270400574,
      "loss": 0.799,
      "num_input_tokens_seen": 43402464,
      "step": 74765
    },
    {
      "epoch": 11.136431337503723,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.014643247166013836,
      "loss": 0.7936,
      "num_input_tokens_seen": 43405504,
      "step": 74770
    },
    {
      "epoch": 11.137176050044683,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.014641298067654069,
      "loss": 0.8022,
      "num_input_tokens_seen": 43408384,
      "step": 74775
    },
    {
      "epoch": 11.137920762585642,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.014639348975354216,
      "loss": 0.8063,
      "num_input_tokens_seen": 43411392,
      "step": 74780
    },
    {
      "epoch": 11.1386654751266,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.014637399889147192,
      "loss": 0.7915,
      "num_input_tokens_seen": 43414336,
      "step": 74785
    },
    {
      "epoch": 11.13941018766756,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.014635450809065921,
      "loss": 0.8262,
      "num_input_tokens_seen": 43417024,
      "step": 74790
    },
    {
      "epoch": 11.14015490020852,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.014633501735143347,
      "loss": 0.8018,
      "num_input_tokens_seen": 43419648,
      "step": 74795
    },
    {
      "epoch": 11.140899612749479,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.014631552667412387,
      "loss": 0.7983,
      "num_input_tokens_seen": 43422560,
      "step": 74800
    },
    {
      "epoch": 11.141644325290438,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.01462960360590597,
      "loss": 0.7892,
      "num_input_tokens_seen": 43425760,
      "step": 74805
    },
    {
      "epoch": 11.142389037831396,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.014627654550657019,
      "loss": 0.8077,
      "num_input_tokens_seen": 43428544,
      "step": 74810
    },
    {
      "epoch": 11.143133750372357,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.014625705501698472,
      "loss": 0.8009,
      "num_input_tokens_seen": 43431392,
      "step": 74815
    },
    {
      "epoch": 11.143878462913316,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.014623756459063244,
      "loss": 0.814,
      "num_input_tokens_seen": 43434496,
      "step": 74820
    },
    {
      "epoch": 11.144623175454274,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.014621807422784273,
      "loss": 0.799,
      "num_input_tokens_seen": 43437472,
      "step": 74825
    },
    {
      "epoch": 11.145367887995233,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.014619858392894484,
      "loss": 0.8093,
      "num_input_tokens_seen": 43440288,
      "step": 74830
    },
    {
      "epoch": 11.146112600536194,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.014617909369426795,
      "loss": 0.8102,
      "num_input_tokens_seen": 43443008,
      "step": 74835
    },
    {
      "epoch": 11.146857313077152,
      "grad_norm": 0.05078125,
      "learning_rate": 0.014615960352414144,
      "loss": 0.8028,
      "num_input_tokens_seen": 43446112,
      "step": 74840
    },
    {
      "epoch": 11.147602025618111,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.014614011341889445,
      "loss": 0.7924,
      "num_input_tokens_seen": 43448992,
      "step": 74845
    },
    {
      "epoch": 11.14834673815907,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.014612062337885638,
      "loss": 0.8135,
      "num_input_tokens_seen": 43451936,
      "step": 74850
    },
    {
      "epoch": 11.14909145070003,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.014610113340435639,
      "loss": 0.7834,
      "num_input_tokens_seen": 43454560,
      "step": 74855
    },
    {
      "epoch": 11.14983616324099,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.014608164349572385,
      "loss": 0.8139,
      "num_input_tokens_seen": 43457760,
      "step": 74860
    },
    {
      "epoch": 11.150580875781948,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.014606215365328796,
      "loss": 0.7838,
      "num_input_tokens_seen": 43460992,
      "step": 74865
    },
    {
      "epoch": 11.151325588322907,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.014604266387737793,
      "loss": 0.7917,
      "num_input_tokens_seen": 43463904,
      "step": 74870
    },
    {
      "epoch": 11.152070300863867,
      "grad_norm": 0.041015625,
      "learning_rate": 0.014602317416832312,
      "loss": 0.7942,
      "num_input_tokens_seen": 43466816,
      "step": 74875
    },
    {
      "epoch": 11.152815013404826,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.014600368452645267,
      "loss": 0.7955,
      "num_input_tokens_seen": 43469568,
      "step": 74880
    },
    {
      "epoch": 11.153559725945785,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.014598419495209599,
      "loss": 0.7871,
      "num_input_tokens_seen": 43472576,
      "step": 74885
    },
    {
      "epoch": 11.154304438486744,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.014596470544558217,
      "loss": 0.8004,
      "num_input_tokens_seen": 43475392,
      "step": 74890
    },
    {
      "epoch": 11.155049151027704,
      "grad_norm": 0.0234375,
      "learning_rate": 0.014594521600724062,
      "loss": 0.8155,
      "num_input_tokens_seen": 43478336,
      "step": 74895
    },
    {
      "epoch": 11.155793863568663,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.014592572663740049,
      "loss": 0.7987,
      "num_input_tokens_seen": 43481216,
      "step": 74900
    },
    {
      "epoch": 11.156538576109622,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.014590623733639112,
      "loss": 0.778,
      "num_input_tokens_seen": 43484416,
      "step": 74905
    },
    {
      "epoch": 11.15728328865058,
      "grad_norm": 0.041015625,
      "learning_rate": 0.014588674810454168,
      "loss": 0.7961,
      "num_input_tokens_seen": 43487040,
      "step": 74910
    },
    {
      "epoch": 11.158028001191541,
      "grad_norm": 0.03515625,
      "learning_rate": 0.014586725894218139,
      "loss": 0.8254,
      "num_input_tokens_seen": 43489952,
      "step": 74915
    },
    {
      "epoch": 11.1587727137325,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.014584776984963964,
      "loss": 0.8002,
      "num_input_tokens_seen": 43492832,
      "step": 74920
    },
    {
      "epoch": 11.159517426273458,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.014582828082724554,
      "loss": 0.7997,
      "num_input_tokens_seen": 43495584,
      "step": 74925
    },
    {
      "epoch": 11.160262138814417,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.014580879187532843,
      "loss": 0.7865,
      "num_input_tokens_seen": 43498496,
      "step": 74930
    },
    {
      "epoch": 11.161006851355376,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.014578930299421752,
      "loss": 0.8164,
      "num_input_tokens_seen": 43501632,
      "step": 74935
    },
    {
      "epoch": 11.161751563896336,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.014576981418424206,
      "loss": 0.7938,
      "num_input_tokens_seen": 43504544,
      "step": 74940
    },
    {
      "epoch": 11.162496276437295,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.014575032544573131,
      "loss": 0.8536,
      "num_input_tokens_seen": 43507392,
      "step": 74945
    },
    {
      "epoch": 11.163240988978254,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.014573083677901442,
      "loss": 0.783,
      "num_input_tokens_seen": 43510176,
      "step": 74950
    },
    {
      "epoch": 11.163985701519213,
      "grad_norm": 0.064453125,
      "learning_rate": 0.014571134818442077,
      "loss": 0.8138,
      "num_input_tokens_seen": 43512960,
      "step": 74955
    },
    {
      "epoch": 11.164730414060173,
      "grad_norm": 0.056640625,
      "learning_rate": 0.014569185966227947,
      "loss": 0.8171,
      "num_input_tokens_seen": 43515712,
      "step": 74960
    },
    {
      "epoch": 11.165475126601132,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.014567237121291989,
      "loss": 0.7908,
      "num_input_tokens_seen": 43518528,
      "step": 74965
    },
    {
      "epoch": 11.16621983914209,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.014565288283667117,
      "loss": 0.7988,
      "num_input_tokens_seen": 43521344,
      "step": 74970
    },
    {
      "epoch": 11.16696455168305,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.01456333945338626,
      "loss": 0.8089,
      "num_input_tokens_seen": 43524000,
      "step": 74975
    },
    {
      "epoch": 11.16770926422401,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.014561390630482333,
      "loss": 0.7888,
      "num_input_tokens_seen": 43526752,
      "step": 74980
    },
    {
      "epoch": 11.168453976764969,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.014559441814988272,
      "loss": 0.8021,
      "num_input_tokens_seen": 43529696,
      "step": 74985
    },
    {
      "epoch": 11.169198689305928,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.014557493006936994,
      "loss": 0.8042,
      "num_input_tokens_seen": 43532352,
      "step": 74990
    },
    {
      "epoch": 11.169943401846886,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.014555544206361414,
      "loss": 0.7923,
      "num_input_tokens_seen": 43535296,
      "step": 74995
    },
    {
      "epoch": 11.170688114387847,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.014553595413294472,
      "loss": 0.8247,
      "num_input_tokens_seen": 43538496,
      "step": 75000
    },
    {
      "epoch": 11.171432826928806,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.014551646627769078,
      "loss": 0.8245,
      "num_input_tokens_seen": 43541408,
      "step": 75005
    },
    {
      "epoch": 11.172177539469764,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.014549697849818159,
      "loss": 0.8099,
      "num_input_tokens_seen": 43544672,
      "step": 75010
    },
    {
      "epoch": 11.172922252010723,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.014547749079474634,
      "loss": 0.8075,
      "num_input_tokens_seen": 43547680,
      "step": 75015
    },
    {
      "epoch": 11.173666964551684,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.014545800316771433,
      "loss": 0.8077,
      "num_input_tokens_seen": 43550656,
      "step": 75020
    },
    {
      "epoch": 11.174411677092642,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.014543851561741475,
      "loss": 0.7993,
      "num_input_tokens_seen": 43553664,
      "step": 75025
    },
    {
      "epoch": 11.175156389633601,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.014541902814417675,
      "loss": 0.8192,
      "num_input_tokens_seen": 43556576,
      "step": 75030
    },
    {
      "epoch": 11.17590110217456,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.014539954074832965,
      "loss": 0.8105,
      "num_input_tokens_seen": 43559168,
      "step": 75035
    },
    {
      "epoch": 11.17664581471552,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.014538005343020263,
      "loss": 0.7984,
      "num_input_tokens_seen": 43561792,
      "step": 75040
    },
    {
      "epoch": 11.17739052725648,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.014536056619012496,
      "loss": 0.786,
      "num_input_tokens_seen": 43564768,
      "step": 75045
    },
    {
      "epoch": 11.178135239797438,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.014534107902842573,
      "loss": 0.7825,
      "num_input_tokens_seen": 43567648,
      "step": 75050
    },
    {
      "epoch": 11.178879952338397,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.014532159194543428,
      "loss": 0.7855,
      "num_input_tokens_seen": 43570528,
      "step": 75055
    },
    {
      "epoch": 11.179624664879357,
      "grad_norm": 0.025390625,
      "learning_rate": 0.014530210494147974,
      "loss": 0.7997,
      "num_input_tokens_seen": 43573440,
      "step": 75060
    },
    {
      "epoch": 11.180369377420316,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.014528261801689143,
      "loss": 0.8353,
      "num_input_tokens_seen": 43576448,
      "step": 75065
    },
    {
      "epoch": 11.181114089961275,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.014526313117199843,
      "loss": 0.8076,
      "num_input_tokens_seen": 43579264,
      "step": 75070
    },
    {
      "epoch": 11.181858802502234,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.014524364440713004,
      "loss": 0.7876,
      "num_input_tokens_seen": 43582176,
      "step": 75075
    },
    {
      "epoch": 11.182603515043194,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.014522415772261547,
      "loss": 0.8088,
      "num_input_tokens_seen": 43585088,
      "step": 75080
    },
    {
      "epoch": 11.183348227584153,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.014520467111878384,
      "loss": 0.7899,
      "num_input_tokens_seen": 43587744,
      "step": 75085
    },
    {
      "epoch": 11.184092940125112,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.014518518459596447,
      "loss": 0.792,
      "num_input_tokens_seen": 43590528,
      "step": 75090
    },
    {
      "epoch": 11.18483765266607,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.014516569815448648,
      "loss": 0.8091,
      "num_input_tokens_seen": 43593248,
      "step": 75095
    },
    {
      "epoch": 11.18558236520703,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.014514621179467915,
      "loss": 0.8096,
      "num_input_tokens_seen": 43596160,
      "step": 75100
    },
    {
      "epoch": 11.18632707774799,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.014512672551687155,
      "loss": 0.7951,
      "num_input_tokens_seen": 43598880,
      "step": 75105
    },
    {
      "epoch": 11.187071790288948,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.014510723932139306,
      "loss": 0.7931,
      "num_input_tokens_seen": 43601696,
      "step": 75110
    },
    {
      "epoch": 11.187816502829907,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.014508775320857278,
      "loss": 0.79,
      "num_input_tokens_seen": 43604672,
      "step": 75115
    },
    {
      "epoch": 11.188561215370866,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.014506826717873986,
      "loss": 0.7816,
      "num_input_tokens_seen": 43607424,
      "step": 75120
    },
    {
      "epoch": 11.189305927911827,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.01450487812322236,
      "loss": 0.8301,
      "num_input_tokens_seen": 43610528,
      "step": 75125
    },
    {
      "epoch": 11.190050640452785,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.014502929536935313,
      "loss": 0.79,
      "num_input_tokens_seen": 43613376,
      "step": 75130
    },
    {
      "epoch": 11.190795352993744,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.01450098095904577,
      "loss": 0.7869,
      "num_input_tokens_seen": 43616384,
      "step": 75135
    },
    {
      "epoch": 11.191540065534703,
      "grad_norm": 0.04296875,
      "learning_rate": 0.01449903238958664,
      "loss": 0.8021,
      "num_input_tokens_seen": 43619104,
      "step": 75140
    },
    {
      "epoch": 11.192284778075663,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.014497083828590853,
      "loss": 0.7964,
      "num_input_tokens_seen": 43622080,
      "step": 75145
    },
    {
      "epoch": 11.193029490616622,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.014495135276091326,
      "loss": 0.8167,
      "num_input_tokens_seen": 43625056,
      "step": 75150
    },
    {
      "epoch": 11.19377420315758,
      "grad_norm": 0.03515625,
      "learning_rate": 0.014493186732120968,
      "loss": 0.7992,
      "num_input_tokens_seen": 43628096,
      "step": 75155
    },
    {
      "epoch": 11.19451891569854,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.014491238196712712,
      "loss": 0.774,
      "num_input_tokens_seen": 43630912,
      "step": 75160
    },
    {
      "epoch": 11.1952636282395,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.014489289669899465,
      "loss": 0.811,
      "num_input_tokens_seen": 43633664,
      "step": 75165
    },
    {
      "epoch": 11.196008340780459,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.014487341151714154,
      "loss": 0.7914,
      "num_input_tokens_seen": 43636512,
      "step": 75170
    },
    {
      "epoch": 11.196753053321418,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.014485392642189687,
      "loss": 0.7744,
      "num_input_tokens_seen": 43639488,
      "step": 75175
    },
    {
      "epoch": 11.197497765862376,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.014483444141358994,
      "loss": 0.8096,
      "num_input_tokens_seen": 43642240,
      "step": 75180
    },
    {
      "epoch": 11.198242478403337,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.01448149564925498,
      "loss": 0.8142,
      "num_input_tokens_seen": 43644992,
      "step": 75185
    },
    {
      "epoch": 11.198987190944296,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.01447954716591058,
      "loss": 0.7903,
      "num_input_tokens_seen": 43647872,
      "step": 75190
    },
    {
      "epoch": 11.199731903485254,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.014477598691358699,
      "loss": 0.8146,
      "num_input_tokens_seen": 43650720,
      "step": 75195
    },
    {
      "epoch": 11.200476616026213,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.014475650225632254,
      "loss": 0.8254,
      "num_input_tokens_seen": 43653312,
      "step": 75200
    },
    {
      "epoch": 11.201221328567174,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.014473701768764169,
      "loss": 0.8034,
      "num_input_tokens_seen": 43656000,
      "step": 75205
    },
    {
      "epoch": 11.201966041108133,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.014471753320787352,
      "loss": 0.8118,
      "num_input_tokens_seen": 43658784,
      "step": 75210
    },
    {
      "epoch": 11.202710753649091,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.014469804881734731,
      "loss": 0.7905,
      "num_input_tokens_seen": 43661664,
      "step": 75215
    },
    {
      "epoch": 11.20345546619005,
      "grad_norm": 0.033203125,
      "learning_rate": 0.014467856451639212,
      "loss": 0.8139,
      "num_input_tokens_seen": 43664544,
      "step": 75220
    },
    {
      "epoch": 11.20420017873101,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.014465908030533723,
      "loss": 0.8162,
      "num_input_tokens_seen": 43667776,
      "step": 75225
    },
    {
      "epoch": 11.20494489127197,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.014463959618451176,
      "loss": 0.7933,
      "num_input_tokens_seen": 43670688,
      "step": 75230
    },
    {
      "epoch": 11.205689603812928,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.014462011215424482,
      "loss": 0.7993,
      "num_input_tokens_seen": 43673792,
      "step": 75235
    },
    {
      "epoch": 11.206434316353887,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.014460062821486567,
      "loss": 0.8068,
      "num_input_tokens_seen": 43676800,
      "step": 75240
    },
    {
      "epoch": 11.207179028894847,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.014458114436670333,
      "loss": 0.8072,
      "num_input_tokens_seen": 43679392,
      "step": 75245
    },
    {
      "epoch": 11.207923741435806,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.014456166061008713,
      "loss": 0.8018,
      "num_input_tokens_seen": 43681984,
      "step": 75250
    },
    {
      "epoch": 11.208668453976765,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.014454217694534607,
      "loss": 0.8181,
      "num_input_tokens_seen": 43684864,
      "step": 75255
    },
    {
      "epoch": 11.209413166517724,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.014452269337280946,
      "loss": 0.7872,
      "num_input_tokens_seen": 43687776,
      "step": 75260
    },
    {
      "epoch": 11.210157879058684,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.014450320989280634,
      "loss": 0.7936,
      "num_input_tokens_seen": 43690720,
      "step": 75265
    },
    {
      "epoch": 11.210902591599643,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.014448372650566593,
      "loss": 0.8153,
      "num_input_tokens_seen": 43693696,
      "step": 75270
    },
    {
      "epoch": 11.211647304140602,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.014446424321171736,
      "loss": 0.7986,
      "num_input_tokens_seen": 43696512,
      "step": 75275
    },
    {
      "epoch": 11.21239201668156,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.014444476001128971,
      "loss": 0.803,
      "num_input_tokens_seen": 43699424,
      "step": 75280
    },
    {
      "epoch": 11.21313672922252,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.014442527690471227,
      "loss": 0.8075,
      "num_input_tokens_seen": 43702176,
      "step": 75285
    },
    {
      "epoch": 11.21388144176348,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.014440579389231403,
      "loss": 0.7824,
      "num_input_tokens_seen": 43705216,
      "step": 75290
    },
    {
      "epoch": 11.214626154304439,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.014438631097442428,
      "loss": 0.8111,
      "num_input_tokens_seen": 43708256,
      "step": 75295
    },
    {
      "epoch": 11.215370866845397,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.014436682815137207,
      "loss": 0.812,
      "num_input_tokens_seen": 43711520,
      "step": 75300
    },
    {
      "epoch": 11.216115579386356,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.014434734542348662,
      "loss": 0.8134,
      "num_input_tokens_seen": 43714368,
      "step": 75305
    },
    {
      "epoch": 11.216860291927317,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.014432786279109702,
      "loss": 0.789,
      "num_input_tokens_seen": 43717216,
      "step": 75310
    },
    {
      "epoch": 11.217605004468275,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.014430838025453233,
      "loss": 0.7994,
      "num_input_tokens_seen": 43719936,
      "step": 75315
    },
    {
      "epoch": 11.218349717009234,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.014428889781412184,
      "loss": 0.7996,
      "num_input_tokens_seen": 43722784,
      "step": 75320
    },
    {
      "epoch": 11.219094429550193,
      "grad_norm": 0.033203125,
      "learning_rate": 0.014426941547019456,
      "loss": 0.7792,
      "num_input_tokens_seen": 43725792,
      "step": 75325
    },
    {
      "epoch": 11.219839142091153,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.014424993322307976,
      "loss": 0.805,
      "num_input_tokens_seen": 43728832,
      "step": 75330
    },
    {
      "epoch": 11.220583854632112,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.014423045107310644,
      "loss": 0.7969,
      "num_input_tokens_seen": 43731776,
      "step": 75335
    },
    {
      "epoch": 11.221328567173071,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.014421096902060383,
      "loss": 0.8199,
      "num_input_tokens_seen": 43734656,
      "step": 75340
    },
    {
      "epoch": 11.22207327971403,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.014419148706590093,
      "loss": 0.8082,
      "num_input_tokens_seen": 43737664,
      "step": 75345
    },
    {
      "epoch": 11.22281799225499,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.014417200520932703,
      "loss": 0.7878,
      "num_input_tokens_seen": 43740544,
      "step": 75350
    },
    {
      "epoch": 11.223562704795949,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.014415252345121119,
      "loss": 0.7999,
      "num_input_tokens_seen": 43743232,
      "step": 75355
    },
    {
      "epoch": 11.224307417336908,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.014413304179188244,
      "loss": 0.8211,
      "num_input_tokens_seen": 43746272,
      "step": 75360
    },
    {
      "epoch": 11.225052129877866,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.014411356023167007,
      "loss": 0.7918,
      "num_input_tokens_seen": 43749184,
      "step": 75365
    },
    {
      "epoch": 11.225796842418827,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.014409407877090307,
      "loss": 0.8219,
      "num_input_tokens_seen": 43752192,
      "step": 75370
    },
    {
      "epoch": 11.226541554959786,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.014407459740991065,
      "loss": 0.7883,
      "num_input_tokens_seen": 43754944,
      "step": 75375
    },
    {
      "epoch": 11.227286267500745,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.014405511614902182,
      "loss": 0.7967,
      "num_input_tokens_seen": 43758016,
      "step": 75380
    },
    {
      "epoch": 11.228030980041703,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.014403563498856582,
      "loss": 0.7942,
      "num_input_tokens_seen": 43760992,
      "step": 75385
    },
    {
      "epoch": 11.228775692582664,
      "grad_norm": 0.03125,
      "learning_rate": 0.014401615392887169,
      "loss": 0.7939,
      "num_input_tokens_seen": 43764192,
      "step": 75390
    },
    {
      "epoch": 11.229520405123623,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.014399667297026854,
      "loss": 0.8317,
      "num_input_tokens_seen": 43766880,
      "step": 75395
    },
    {
      "epoch": 11.230265117664581,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.01439771921130855,
      "loss": 0.8053,
      "num_input_tokens_seen": 43769696,
      "step": 75400
    },
    {
      "epoch": 11.23100983020554,
      "grad_norm": 0.06640625,
      "learning_rate": 0.014395771135765168,
      "loss": 0.7953,
      "num_input_tokens_seen": 43772640,
      "step": 75405
    },
    {
      "epoch": 11.2317545427465,
      "grad_norm": 0.046875,
      "learning_rate": 0.014393823070429622,
      "loss": 0.7947,
      "num_input_tokens_seen": 43775712,
      "step": 75410
    },
    {
      "epoch": 11.23249925528746,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.01439187501533481,
      "loss": 0.8015,
      "num_input_tokens_seen": 43778592,
      "step": 75415
    },
    {
      "epoch": 11.233243967828418,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01438992697051366,
      "loss": 0.8,
      "num_input_tokens_seen": 43781472,
      "step": 75420
    },
    {
      "epoch": 11.233988680369377,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.014387978935999073,
      "loss": 0.7835,
      "num_input_tokens_seen": 43784512,
      "step": 75425
    },
    {
      "epoch": 11.234733392910337,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.01438603091182396,
      "loss": 0.799,
      "num_input_tokens_seen": 43787264,
      "step": 75430
    },
    {
      "epoch": 11.235478105451296,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01438408289802123,
      "loss": 0.8038,
      "num_input_tokens_seen": 43789952,
      "step": 75435
    },
    {
      "epoch": 11.236222817992255,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.01438213489462379,
      "loss": 0.8045,
      "num_input_tokens_seen": 43792864,
      "step": 75440
    },
    {
      "epoch": 11.236967530533214,
      "grad_norm": 0.037109375,
      "learning_rate": 0.014380186901664557,
      "loss": 0.8156,
      "num_input_tokens_seen": 43795616,
      "step": 75445
    },
    {
      "epoch": 11.237712243074174,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.014378238919176431,
      "loss": 0.798,
      "num_input_tokens_seen": 43798528,
      "step": 75450
    },
    {
      "epoch": 11.238456955615133,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.014376290947192331,
      "loss": 0.8025,
      "num_input_tokens_seen": 43801440,
      "step": 75455
    },
    {
      "epoch": 11.239201668156092,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.01437434298574516,
      "loss": 0.8067,
      "num_input_tokens_seen": 43804320,
      "step": 75460
    },
    {
      "epoch": 11.23994638069705,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.014372395034867832,
      "loss": 0.7914,
      "num_input_tokens_seen": 43806944,
      "step": 75465
    },
    {
      "epoch": 11.24069109323801,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.014370447094593251,
      "loss": 0.8247,
      "num_input_tokens_seen": 43809632,
      "step": 75470
    },
    {
      "epoch": 11.24143580577897,
      "grad_norm": 0.05078125,
      "learning_rate": 0.014368499164954321,
      "loss": 0.8128,
      "num_input_tokens_seen": 43812096,
      "step": 75475
    },
    {
      "epoch": 11.242180518319929,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.014366551245983962,
      "loss": 0.8176,
      "num_input_tokens_seen": 43814688,
      "step": 75480
    },
    {
      "epoch": 11.242925230860887,
      "grad_norm": 0.064453125,
      "learning_rate": 0.01436460333771507,
      "loss": 0.7904,
      "num_input_tokens_seen": 43817632,
      "step": 75485
    },
    {
      "epoch": 11.243669943401846,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.014362655440180566,
      "loss": 0.8086,
      "num_input_tokens_seen": 43820288,
      "step": 75490
    },
    {
      "epoch": 11.244414655942807,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.014360707553413345,
      "loss": 0.7876,
      "num_input_tokens_seen": 43823008,
      "step": 75495
    },
    {
      "epoch": 11.245159368483765,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.014358759677446323,
      "loss": 0.7923,
      "num_input_tokens_seen": 43826336,
      "step": 75500
    },
    {
      "epoch": 11.245904081024724,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0143568118123124,
      "loss": 0.8103,
      "num_input_tokens_seen": 43829472,
      "step": 75505
    },
    {
      "epoch": 11.246648793565683,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.014354863958044493,
      "loss": 0.8105,
      "num_input_tokens_seen": 43832448,
      "step": 75510
    },
    {
      "epoch": 11.247393506106643,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.014352916114675504,
      "loss": 0.7891,
      "num_input_tokens_seen": 43835488,
      "step": 75515
    },
    {
      "epoch": 11.248138218647602,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.014350968282238334,
      "loss": 0.8075,
      "num_input_tokens_seen": 43838368,
      "step": 75520
    },
    {
      "epoch": 11.248882931188561,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.014349020460765902,
      "loss": 0.8005,
      "num_input_tokens_seen": 43841152,
      "step": 75525
    },
    {
      "epoch": 11.24962764372952,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.014347072650291103,
      "loss": 0.7974,
      "num_input_tokens_seen": 43843936,
      "step": 75530
    },
    {
      "epoch": 11.25037235627048,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01434512485084685,
      "loss": 0.8037,
      "num_input_tokens_seen": 43846720,
      "step": 75535
    },
    {
      "epoch": 11.251117068811439,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.014343177062466045,
      "loss": 0.7887,
      "num_input_tokens_seen": 43849728,
      "step": 75540
    },
    {
      "epoch": 11.251861781352398,
      "grad_norm": 0.08203125,
      "learning_rate": 0.014341229285181598,
      "loss": 0.8315,
      "num_input_tokens_seen": 43852672,
      "step": 75545
    },
    {
      "epoch": 11.252606493893357,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.014339281519026414,
      "loss": 0.7971,
      "num_input_tokens_seen": 43855360,
      "step": 75550
    },
    {
      "epoch": 11.253351206434317,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.014337333764033393,
      "loss": 0.7823,
      "num_input_tokens_seen": 43858144,
      "step": 75555
    },
    {
      "epoch": 11.254095918975276,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.014335386020235447,
      "loss": 0.7837,
      "num_input_tokens_seen": 43860992,
      "step": 75560
    },
    {
      "epoch": 11.254840631516235,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.014333438287665479,
      "loss": 0.7866,
      "num_input_tokens_seen": 43863936,
      "step": 75565
    },
    {
      "epoch": 11.255585344057193,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.014331490566356395,
      "loss": 0.8051,
      "num_input_tokens_seen": 43866912,
      "step": 75570
    },
    {
      "epoch": 11.256330056598154,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.014329542856341092,
      "loss": 0.8066,
      "num_input_tokens_seen": 43869696,
      "step": 75575
    },
    {
      "epoch": 11.257074769139113,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.01432759515765249,
      "loss": 0.7811,
      "num_input_tokens_seen": 43873056,
      "step": 75580
    },
    {
      "epoch": 11.257819481680071,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.014325647470323475,
      "loss": 0.7989,
      "num_input_tokens_seen": 43875872,
      "step": 75585
    },
    {
      "epoch": 11.25856419422103,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.014323699794386968,
      "loss": 0.8178,
      "num_input_tokens_seen": 43878784,
      "step": 75590
    },
    {
      "epoch": 11.25930890676199,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.014321752129875867,
      "loss": 0.821,
      "num_input_tokens_seen": 43881760,
      "step": 75595
    },
    {
      "epoch": 11.26005361930295,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.014319804476823069,
      "loss": 0.7889,
      "num_input_tokens_seen": 43884544,
      "step": 75600
    },
    {
      "epoch": 11.260798331843908,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.014317856835261487,
      "loss": 0.7918,
      "num_input_tokens_seen": 43887392,
      "step": 75605
    },
    {
      "epoch": 11.261543044384867,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.014315909205224013,
      "loss": 0.8009,
      "num_input_tokens_seen": 43890144,
      "step": 75610
    },
    {
      "epoch": 11.262287756925826,
      "grad_norm": 0.03515625,
      "learning_rate": 0.014313961586743565,
      "loss": 0.8018,
      "num_input_tokens_seen": 43892832,
      "step": 75615
    },
    {
      "epoch": 11.263032469466786,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.014312013979853034,
      "loss": 0.8134,
      "num_input_tokens_seen": 43895680,
      "step": 75620
    },
    {
      "epoch": 11.263777182007745,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.014310066384585332,
      "loss": 0.7946,
      "num_input_tokens_seen": 43898720,
      "step": 75625
    },
    {
      "epoch": 11.264521894548704,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.014308118800973359,
      "loss": 0.8009,
      "num_input_tokens_seen": 43901504,
      "step": 75630
    },
    {
      "epoch": 11.265266607089663,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01430617122905001,
      "loss": 0.7966,
      "num_input_tokens_seen": 43904704,
      "step": 75635
    },
    {
      "epoch": 11.266011319630623,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.014304223668848199,
      "loss": 0.795,
      "num_input_tokens_seen": 43907392,
      "step": 75640
    },
    {
      "epoch": 11.266756032171582,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.014302276120400812,
      "loss": 0.8145,
      "num_input_tokens_seen": 43910176,
      "step": 75645
    },
    {
      "epoch": 11.26750074471254,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.01430032858374077,
      "loss": 0.7782,
      "num_input_tokens_seen": 43912960,
      "step": 75650
    },
    {
      "epoch": 11.2682454572535,
      "grad_norm": 0.03125,
      "learning_rate": 0.01429838105890096,
      "loss": 0.7875,
      "num_input_tokens_seen": 43915552,
      "step": 75655
    },
    {
      "epoch": 11.26899016979446,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.01429643354591429,
      "loss": 0.7887,
      "num_input_tokens_seen": 43918464,
      "step": 75660
    },
    {
      "epoch": 11.269734882335419,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.014294486044813662,
      "loss": 0.7917,
      "num_input_tokens_seen": 43921792,
      "step": 75665
    },
    {
      "epoch": 11.270479594876377,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.014292538555631978,
      "loss": 0.8157,
      "num_input_tokens_seen": 43924448,
      "step": 75670
    },
    {
      "epoch": 11.271224307417336,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.014290591078402134,
      "loss": 0.8176,
      "num_input_tokens_seen": 43927456,
      "step": 75675
    },
    {
      "epoch": 11.271969019958297,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.014288643613157027,
      "loss": 0.7951,
      "num_input_tokens_seen": 43930592,
      "step": 75680
    },
    {
      "epoch": 11.272713732499255,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.014286696159929571,
      "loss": 0.802,
      "num_input_tokens_seen": 43933728,
      "step": 75685
    },
    {
      "epoch": 11.273458445040214,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01428474871875265,
      "loss": 0.8039,
      "num_input_tokens_seen": 43936736,
      "step": 75690
    },
    {
      "epoch": 11.274203157581173,
      "grad_norm": 0.09375,
      "learning_rate": 0.01428280128965918,
      "loss": 0.8165,
      "num_input_tokens_seen": 43939616,
      "step": 75695
    },
    {
      "epoch": 11.274947870122134,
      "grad_norm": 0.048828125,
      "learning_rate": 0.01428085387268205,
      "loss": 0.8071,
      "num_input_tokens_seen": 43942720,
      "step": 75700
    },
    {
      "epoch": 11.275692582663092,
      "grad_norm": 0.044921875,
      "learning_rate": 0.014278906467854166,
      "loss": 0.7854,
      "num_input_tokens_seen": 43945536,
      "step": 75705
    },
    {
      "epoch": 11.276437295204051,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.014276959075208416,
      "loss": 0.7966,
      "num_input_tokens_seen": 43948672,
      "step": 75710
    },
    {
      "epoch": 11.27718200774501,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.014275011694777717,
      "loss": 0.7891,
      "num_input_tokens_seen": 43951488,
      "step": 75715
    },
    {
      "epoch": 11.27792672028597,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.014273064326594956,
      "loss": 0.796,
      "num_input_tokens_seen": 43954432,
      "step": 75720
    },
    {
      "epoch": 11.278671432826929,
      "grad_norm": 0.048828125,
      "learning_rate": 0.014271116970693035,
      "loss": 0.7882,
      "num_input_tokens_seen": 43957280,
      "step": 75725
    },
    {
      "epoch": 11.279416145367888,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.014269169627104849,
      "loss": 0.8171,
      "num_input_tokens_seen": 43959872,
      "step": 75730
    },
    {
      "epoch": 11.280160857908847,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.014267222295863299,
      "loss": 0.7774,
      "num_input_tokens_seen": 43962560,
      "step": 75735
    },
    {
      "epoch": 11.280905570449807,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.014265274977001287,
      "loss": 0.8112,
      "num_input_tokens_seen": 43965600,
      "step": 75740
    },
    {
      "epoch": 11.281650282990766,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.014263327670551702,
      "loss": 0.7927,
      "num_input_tokens_seen": 43968512,
      "step": 75745
    },
    {
      "epoch": 11.282394995531725,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.014261380376547454,
      "loss": 0.8141,
      "num_input_tokens_seen": 43971808,
      "step": 75750
    },
    {
      "epoch": 11.283139708072683,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.014259433095021434,
      "loss": 0.7789,
      "num_input_tokens_seen": 43974816,
      "step": 75755
    },
    {
      "epoch": 11.283884420613644,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.014257485826006536,
      "loss": 0.8,
      "num_input_tokens_seen": 43978016,
      "step": 75760
    },
    {
      "epoch": 11.284629133154603,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.014255538569535662,
      "loss": 0.8084,
      "num_input_tokens_seen": 43981184,
      "step": 75765
    },
    {
      "epoch": 11.285373845695561,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.0142535913256417,
      "loss": 0.7962,
      "num_input_tokens_seen": 43984480,
      "step": 75770
    },
    {
      "epoch": 11.28611855823652,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.014251644094357563,
      "loss": 0.8069,
      "num_input_tokens_seen": 43987424,
      "step": 75775
    },
    {
      "epoch": 11.28686327077748,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.014249696875716132,
      "loss": 0.805,
      "num_input_tokens_seen": 43990144,
      "step": 75780
    },
    {
      "epoch": 11.28760798331844,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.014247749669750315,
      "loss": 0.8059,
      "num_input_tokens_seen": 43993248,
      "step": 75785
    },
    {
      "epoch": 11.288352695859398,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.014245802476492998,
      "loss": 0.796,
      "num_input_tokens_seen": 43997184,
      "step": 75790
    },
    {
      "epoch": 11.289097408400357,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.014243855295977086,
      "loss": 0.7873,
      "num_input_tokens_seen": 44000192,
      "step": 75795
    },
    {
      "epoch": 11.289842120941316,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.014241908128235472,
      "loss": 0.7772,
      "num_input_tokens_seen": 44003232,
      "step": 75800
    },
    {
      "epoch": 11.290586833482276,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.014239960973301042,
      "loss": 0.8148,
      "num_input_tokens_seen": 44006208,
      "step": 75805
    },
    {
      "epoch": 11.291331546023235,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.014238013831206707,
      "loss": 0.7836,
      "num_input_tokens_seen": 44008992,
      "step": 75810
    },
    {
      "epoch": 11.292076258564194,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.014236066701985345,
      "loss": 0.7895,
      "num_input_tokens_seen": 44011904,
      "step": 75815
    },
    {
      "epoch": 11.292820971105153,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.014234119585669867,
      "loss": 0.782,
      "num_input_tokens_seen": 44014656,
      "step": 75820
    },
    {
      "epoch": 11.293565683646113,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.014232172482293156,
      "loss": 0.8189,
      "num_input_tokens_seen": 44017472,
      "step": 75825
    },
    {
      "epoch": 11.294310396187072,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.014230225391888115,
      "loss": 0.8134,
      "num_input_tokens_seen": 44020640,
      "step": 75830
    },
    {
      "epoch": 11.29505510872803,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.014228278314487634,
      "loss": 0.8135,
      "num_input_tokens_seen": 44023648,
      "step": 75835
    },
    {
      "epoch": 11.29579982126899,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.014226331250124598,
      "loss": 0.8007,
      "num_input_tokens_seen": 44026592,
      "step": 75840
    },
    {
      "epoch": 11.29654453380995,
      "grad_norm": 0.048828125,
      "learning_rate": 0.014224384198831918,
      "loss": 0.7811,
      "num_input_tokens_seen": 44029440,
      "step": 75845
    },
    {
      "epoch": 11.297289246350909,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.01422243716064247,
      "loss": 0.7911,
      "num_input_tokens_seen": 44032256,
      "step": 75850
    },
    {
      "epoch": 11.298033958891867,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.014220490135589164,
      "loss": 0.8087,
      "num_input_tokens_seen": 44035456,
      "step": 75855
    },
    {
      "epoch": 11.298778671432826,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.014218543123704882,
      "loss": 0.7741,
      "num_input_tokens_seen": 44038368,
      "step": 75860
    },
    {
      "epoch": 11.299523383973787,
      "grad_norm": 0.03515625,
      "learning_rate": 0.014216596125022522,
      "loss": 0.8123,
      "num_input_tokens_seen": 44041600,
      "step": 75865
    },
    {
      "epoch": 11.300268096514746,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.014214649139574967,
      "loss": 0.7942,
      "num_input_tokens_seen": 44044736,
      "step": 75870
    },
    {
      "epoch": 11.301012809055704,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.014212702167395124,
      "loss": 0.8315,
      "num_input_tokens_seen": 44047744,
      "step": 75875
    },
    {
      "epoch": 11.301757521596663,
      "grad_norm": 0.048828125,
      "learning_rate": 0.014210755208515878,
      "loss": 0.7922,
      "num_input_tokens_seen": 44050560,
      "step": 75880
    },
    {
      "epoch": 11.302502234137624,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.014208808262970114,
      "loss": 0.7891,
      "num_input_tokens_seen": 44053344,
      "step": 75885
    },
    {
      "epoch": 11.303246946678582,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.014206861330790736,
      "loss": 0.7936,
      "num_input_tokens_seen": 44056000,
      "step": 75890
    },
    {
      "epoch": 11.303991659219541,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.014204914412010628,
      "loss": 0.8088,
      "num_input_tokens_seen": 44058976,
      "step": 75895
    },
    {
      "epoch": 11.3047363717605,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.014202967506662684,
      "loss": 0.8023,
      "num_input_tokens_seen": 44062048,
      "step": 75900
    },
    {
      "epoch": 11.30548108430146,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01420102061477979,
      "loss": 0.7964,
      "num_input_tokens_seen": 44065120,
      "step": 75905
    },
    {
      "epoch": 11.30622579684242,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.014199073736394844,
      "loss": 0.7893,
      "num_input_tokens_seen": 44067904,
      "step": 75910
    },
    {
      "epoch": 11.306970509383378,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.014197126871540736,
      "loss": 0.7947,
      "num_input_tokens_seen": 44071008,
      "step": 75915
    },
    {
      "epoch": 11.307715221924337,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.014195180020250346,
      "loss": 0.798,
      "num_input_tokens_seen": 44073888,
      "step": 75920
    },
    {
      "epoch": 11.308459934465297,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.014193233182556577,
      "loss": 0.8045,
      "num_input_tokens_seen": 44076544,
      "step": 75925
    },
    {
      "epoch": 11.309204647006256,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.014191286358492311,
      "loss": 0.8022,
      "num_input_tokens_seen": 44079456,
      "step": 75930
    },
    {
      "epoch": 11.309949359547215,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.014189339548090445,
      "loss": 0.7943,
      "num_input_tokens_seen": 44082272,
      "step": 75935
    },
    {
      "epoch": 11.310694072088173,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.014187392751383855,
      "loss": 0.7735,
      "num_input_tokens_seen": 44085088,
      "step": 75940
    },
    {
      "epoch": 11.311438784629134,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.014185445968405446,
      "loss": 0.7924,
      "num_input_tokens_seen": 44088064,
      "step": 75945
    },
    {
      "epoch": 11.312183497170093,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.014183499199188092,
      "loss": 0.7946,
      "num_input_tokens_seen": 44090816,
      "step": 75950
    },
    {
      "epoch": 11.312928209711052,
      "grad_norm": 0.0390625,
      "learning_rate": 0.014181552443764695,
      "loss": 0.7788,
      "num_input_tokens_seen": 44093696,
      "step": 75955
    },
    {
      "epoch": 11.31367292225201,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.01417960570216814,
      "loss": 0.7983,
      "num_input_tokens_seen": 44096512,
      "step": 75960
    },
    {
      "epoch": 11.31441763479297,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.014177658974431307,
      "loss": 0.7766,
      "num_input_tokens_seen": 44099712,
      "step": 75965
    },
    {
      "epoch": 11.31516234733393,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.014175712260587095,
      "loss": 0.8157,
      "num_input_tokens_seen": 44102560,
      "step": 75970
    },
    {
      "epoch": 11.315907059874888,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.01417376556066838,
      "loss": 0.775,
      "num_input_tokens_seen": 44105472,
      "step": 75975
    },
    {
      "epoch": 11.316651772415847,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.014171818874708062,
      "loss": 0.825,
      "num_input_tokens_seen": 44108384,
      "step": 75980
    },
    {
      "epoch": 11.317396484956806,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.014169872202739015,
      "loss": 0.8147,
      "num_input_tokens_seen": 44111424,
      "step": 75985
    },
    {
      "epoch": 11.318141197497766,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.014167925544794142,
      "loss": 0.7884,
      "num_input_tokens_seen": 44114496,
      "step": 75990
    },
    {
      "epoch": 11.318885910038725,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.014165978900906321,
      "loss": 0.8158,
      "num_input_tokens_seen": 44117696,
      "step": 75995
    },
    {
      "epoch": 11.319630622579684,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.014164032271108434,
      "loss": 0.7991,
      "num_input_tokens_seen": 44120608,
      "step": 76000
    },
    {
      "epoch": 11.320375335120643,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.014162085655433376,
      "loss": 0.7906,
      "num_input_tokens_seen": 44123392,
      "step": 76005
    },
    {
      "epoch": 11.321120047661603,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.014160139053914023,
      "loss": 0.7974,
      "num_input_tokens_seen": 44126624,
      "step": 76010
    },
    {
      "epoch": 11.321864760202562,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.014158192466583274,
      "loss": 0.8082,
      "num_input_tokens_seen": 44129600,
      "step": 76015
    },
    {
      "epoch": 11.32260947274352,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.014156245893474,
      "loss": 0.7971,
      "num_input_tokens_seen": 44132384,
      "step": 76020
    },
    {
      "epoch": 11.32335418528448,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0141542993346191,
      "loss": 0.8005,
      "num_input_tokens_seen": 44135264,
      "step": 76025
    },
    {
      "epoch": 11.32409889782544,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.014152352790051453,
      "loss": 0.7847,
      "num_input_tokens_seen": 44138176,
      "step": 76030
    },
    {
      "epoch": 11.324843610366399,
      "grad_norm": 0.033203125,
      "learning_rate": 0.014150406259803947,
      "loss": 0.7848,
      "num_input_tokens_seen": 44141280,
      "step": 76035
    },
    {
      "epoch": 11.325588322907358,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.01414845974390946,
      "loss": 0.8013,
      "num_input_tokens_seen": 44143904,
      "step": 76040
    },
    {
      "epoch": 11.326333035448316,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.014146513242400877,
      "loss": 0.7947,
      "num_input_tokens_seen": 44147136,
      "step": 76045
    },
    {
      "epoch": 11.327077747989277,
      "grad_norm": 0.03125,
      "learning_rate": 0.01414456675531109,
      "loss": 0.7946,
      "num_input_tokens_seen": 44149888,
      "step": 76050
    },
    {
      "epoch": 11.327822460530236,
      "grad_norm": 0.048828125,
      "learning_rate": 0.014142620282672975,
      "loss": 0.7984,
      "num_input_tokens_seen": 44152544,
      "step": 76055
    },
    {
      "epoch": 11.328567173071194,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.014140673824519421,
      "loss": 0.7858,
      "num_input_tokens_seen": 44155744,
      "step": 76060
    },
    {
      "epoch": 11.329311885612153,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.014138727380883308,
      "loss": 0.8045,
      "num_input_tokens_seen": 44158656,
      "step": 76065
    },
    {
      "epoch": 11.330056598153114,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.014136780951797524,
      "loss": 0.8038,
      "num_input_tokens_seen": 44161376,
      "step": 76070
    },
    {
      "epoch": 11.330801310694072,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.014134834537294947,
      "loss": 0.792,
      "num_input_tokens_seen": 44164096,
      "step": 76075
    },
    {
      "epoch": 11.331546023235031,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.014132888137408458,
      "loss": 0.8031,
      "num_input_tokens_seen": 44166816,
      "step": 76080
    },
    {
      "epoch": 11.33229073577599,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.014130941752170946,
      "loss": 0.8138,
      "num_input_tokens_seen": 44169632,
      "step": 76085
    },
    {
      "epoch": 11.33303544831695,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.014128995381615288,
      "loss": 0.7815,
      "num_input_tokens_seen": 44172352,
      "step": 76090
    },
    {
      "epoch": 11.33378016085791,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.014127049025774371,
      "loss": 0.79,
      "num_input_tokens_seen": 44175520,
      "step": 76095
    },
    {
      "epoch": 11.334524873398868,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.014125102684681068,
      "loss": 0.7991,
      "num_input_tokens_seen": 44178304,
      "step": 76100
    },
    {
      "epoch": 11.335269585939827,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.014123156358368269,
      "loss": 0.7827,
      "num_input_tokens_seen": 44181024,
      "step": 76105
    },
    {
      "epoch": 11.336014298480787,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.014121210046868848,
      "loss": 0.7988,
      "num_input_tokens_seen": 44183968,
      "step": 76110
    },
    {
      "epoch": 11.336759011021746,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.014119263750215697,
      "loss": 0.8305,
      "num_input_tokens_seen": 44186752,
      "step": 76115
    },
    {
      "epoch": 11.337503723562705,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.014117317468441689,
      "loss": 0.7763,
      "num_input_tokens_seen": 44189728,
      "step": 76120
    },
    {
      "epoch": 11.338248436103664,
      "grad_norm": 0.033203125,
      "learning_rate": 0.014115371201579702,
      "loss": 0.7888,
      "num_input_tokens_seen": 44192576,
      "step": 76125
    },
    {
      "epoch": 11.338993148644622,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.014113424949662624,
      "loss": 0.7983,
      "num_input_tokens_seen": 44195136,
      "step": 76130
    },
    {
      "epoch": 11.339737861185583,
      "grad_norm": 0.03125,
      "learning_rate": 0.014111478712723323,
      "loss": 0.8205,
      "num_input_tokens_seen": 44198048,
      "step": 76135
    },
    {
      "epoch": 11.340482573726542,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.014109532490794692,
      "loss": 0.8111,
      "num_input_tokens_seen": 44201024,
      "step": 76140
    },
    {
      "epoch": 11.3412272862675,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.014107586283909599,
      "loss": 0.7996,
      "num_input_tokens_seen": 44204224,
      "step": 76145
    },
    {
      "epoch": 11.341971998808459,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.014105640092100934,
      "loss": 0.7935,
      "num_input_tokens_seen": 44207072,
      "step": 76150
    },
    {
      "epoch": 11.34271671134942,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.014103693915401571,
      "loss": 0.792,
      "num_input_tokens_seen": 44209984,
      "step": 76155
    },
    {
      "epoch": 11.343461423890378,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.014101747753844386,
      "loss": 0.7994,
      "num_input_tokens_seen": 44212800,
      "step": 76160
    },
    {
      "epoch": 11.344206136431337,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.014099801607462263,
      "loss": 0.8092,
      "num_input_tokens_seen": 44215776,
      "step": 76165
    },
    {
      "epoch": 11.344950848972296,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.01409785547628807,
      "loss": 0.7828,
      "num_input_tokens_seen": 44218624,
      "step": 76170
    },
    {
      "epoch": 11.345695561513256,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.014095909360354696,
      "loss": 0.8047,
      "num_input_tokens_seen": 44221568,
      "step": 76175
    },
    {
      "epoch": 11.346440274054215,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.014093963259695011,
      "loss": 0.804,
      "num_input_tokens_seen": 44224704,
      "step": 76180
    },
    {
      "epoch": 11.347184986595174,
      "grad_norm": 0.0390625,
      "learning_rate": 0.014092017174341902,
      "loss": 0.7952,
      "num_input_tokens_seen": 44227584,
      "step": 76185
    },
    {
      "epoch": 11.347929699136133,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.014090071104328234,
      "loss": 0.8043,
      "num_input_tokens_seen": 44230944,
      "step": 76190
    },
    {
      "epoch": 11.348674411677093,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.014088125049686897,
      "loss": 0.8065,
      "num_input_tokens_seen": 44233792,
      "step": 76195
    },
    {
      "epoch": 11.349419124218052,
      "grad_norm": 0.041015625,
      "learning_rate": 0.014086179010450759,
      "loss": 0.7767,
      "num_input_tokens_seen": 44236800,
      "step": 76200
    },
    {
      "epoch": 11.35016383675901,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.01408423298665269,
      "loss": 0.7765,
      "num_input_tokens_seen": 44239776,
      "step": 76205
    },
    {
      "epoch": 11.35090854929997,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.014082286978325581,
      "loss": 0.8164,
      "num_input_tokens_seen": 44242912,
      "step": 76210
    },
    {
      "epoch": 11.35165326184093,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.014080340985502296,
      "loss": 0.7676,
      "num_input_tokens_seen": 44245856,
      "step": 76215
    },
    {
      "epoch": 11.352397974381889,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.01407839500821572,
      "loss": 0.7924,
      "num_input_tokens_seen": 44248928,
      "step": 76220
    },
    {
      "epoch": 11.353142686922848,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.014076449046498723,
      "loss": 0.7626,
      "num_input_tokens_seen": 44251904,
      "step": 76225
    },
    {
      "epoch": 11.353887399463806,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.014074503100384182,
      "loss": 0.8048,
      "num_input_tokens_seen": 44254592,
      "step": 76230
    },
    {
      "epoch": 11.354632112004767,
      "grad_norm": 0.044921875,
      "learning_rate": 0.014072557169904971,
      "loss": 0.8674,
      "num_input_tokens_seen": 44257760,
      "step": 76235
    },
    {
      "epoch": 11.355376824545726,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.014070611255093958,
      "loss": 0.796,
      "num_input_tokens_seen": 44260544,
      "step": 76240
    },
    {
      "epoch": 11.356121537086684,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.014068665355984028,
      "loss": 0.8098,
      "num_input_tokens_seen": 44263168,
      "step": 76245
    },
    {
      "epoch": 11.356866249627643,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.014066719472608045,
      "loss": 0.8158,
      "num_input_tokens_seen": 44265824,
      "step": 76250
    },
    {
      "epoch": 11.357610962168604,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.014064773604998894,
      "loss": 0.7971,
      "num_input_tokens_seen": 44268928,
      "step": 76255
    },
    {
      "epoch": 11.358355674709562,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01406282775318944,
      "loss": 0.8379,
      "num_input_tokens_seen": 44271712,
      "step": 76260
    },
    {
      "epoch": 11.359100387250521,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01406088191721256,
      "loss": 0.8012,
      "num_input_tokens_seen": 44274592,
      "step": 76265
    },
    {
      "epoch": 11.35984509979148,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.014058936097101122,
      "loss": 0.7918,
      "num_input_tokens_seen": 44277696,
      "step": 76270
    },
    {
      "epoch": 11.36058981233244,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.014056990292888006,
      "loss": 0.8012,
      "num_input_tokens_seen": 44280832,
      "step": 76275
    },
    {
      "epoch": 11.3613345248734,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.01405504450460608,
      "loss": 0.8153,
      "num_input_tokens_seen": 44283680,
      "step": 76280
    },
    {
      "epoch": 11.362079237414358,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.014053098732288213,
      "loss": 0.7976,
      "num_input_tokens_seen": 44286784,
      "step": 76285
    },
    {
      "epoch": 11.362823949955317,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.014051152975967285,
      "loss": 0.8215,
      "num_input_tokens_seen": 44289440,
      "step": 76290
    },
    {
      "epoch": 11.363568662496277,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.01404920723567616,
      "loss": 0.7978,
      "num_input_tokens_seen": 44292352,
      "step": 76295
    },
    {
      "epoch": 11.364313375037236,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.014047261511447718,
      "loss": 0.7924,
      "num_input_tokens_seen": 44295360,
      "step": 76300
    },
    {
      "epoch": 11.365058087578195,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.014045315803314815,
      "loss": 0.7867,
      "num_input_tokens_seen": 44298272,
      "step": 76305
    },
    {
      "epoch": 11.365802800119154,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.014043370111310339,
      "loss": 0.816,
      "num_input_tokens_seen": 44301472,
      "step": 76310
    },
    {
      "epoch": 11.366547512660112,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.014041424435467145,
      "loss": 0.8344,
      "num_input_tokens_seen": 44304800,
      "step": 76315
    },
    {
      "epoch": 11.367292225201073,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.014039478775818121,
      "loss": 0.7913,
      "num_input_tokens_seen": 44307520,
      "step": 76320
    },
    {
      "epoch": 11.368036937742032,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.014037533132396121,
      "loss": 0.781,
      "num_input_tokens_seen": 44310528,
      "step": 76325
    },
    {
      "epoch": 11.36878165028299,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.014035587505234023,
      "loss": 0.8025,
      "num_input_tokens_seen": 44313344,
      "step": 76330
    },
    {
      "epoch": 11.36952636282395,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.014033641894364695,
      "loss": 0.807,
      "num_input_tokens_seen": 44316096,
      "step": 76335
    },
    {
      "epoch": 11.37027107536491,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.014031696299820998,
      "loss": 0.8215,
      "num_input_tokens_seen": 44318816,
      "step": 76340
    },
    {
      "epoch": 11.371015787905868,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.014029750721635814,
      "loss": 0.7751,
      "num_input_tokens_seen": 44321440,
      "step": 76345
    },
    {
      "epoch": 11.371760500446827,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.014027805159842,
      "loss": 0.8065,
      "num_input_tokens_seen": 44324160,
      "step": 76350
    },
    {
      "epoch": 11.372505212987786,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.014025859614472436,
      "loss": 0.8,
      "num_input_tokens_seen": 44326944,
      "step": 76355
    },
    {
      "epoch": 11.373249925528746,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.014023914085559983,
      "loss": 0.7944,
      "num_input_tokens_seen": 44329824,
      "step": 76360
    },
    {
      "epoch": 11.373994638069705,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.014021968573137507,
      "loss": 0.7875,
      "num_input_tokens_seen": 44332544,
      "step": 76365
    },
    {
      "epoch": 11.374739350610664,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.014020023077237882,
      "loss": 0.8291,
      "num_input_tokens_seen": 44335424,
      "step": 76370
    },
    {
      "epoch": 11.375484063151623,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.014018077597893962,
      "loss": 0.817,
      "num_input_tokens_seen": 44338272,
      "step": 76375
    },
    {
      "epoch": 11.376228775692583,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.014016132135138632,
      "loss": 0.8186,
      "num_input_tokens_seen": 44341152,
      "step": 76380
    },
    {
      "epoch": 11.376973488233542,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.014014186689004745,
      "loss": 0.797,
      "num_input_tokens_seen": 44344192,
      "step": 76385
    },
    {
      "epoch": 11.3777182007745,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.014012241259525172,
      "loss": 0.8171,
      "num_input_tokens_seen": 44347264,
      "step": 76390
    },
    {
      "epoch": 11.37846291331546,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.01401029584673278,
      "loss": 0.7789,
      "num_input_tokens_seen": 44350304,
      "step": 76395
    },
    {
      "epoch": 11.37920762585642,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.014008350450660437,
      "loss": 0.8084,
      "num_input_tokens_seen": 44353344,
      "step": 76400
    },
    {
      "epoch": 11.379952338397379,
      "grad_norm": 0.037109375,
      "learning_rate": 0.014006405071341004,
      "loss": 0.81,
      "num_input_tokens_seen": 44356064,
      "step": 76405
    },
    {
      "epoch": 11.380697050938338,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.014004459708807341,
      "loss": 0.7923,
      "num_input_tokens_seen": 44358720,
      "step": 76410
    },
    {
      "epoch": 11.381441763479296,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.014002514363092326,
      "loss": 0.7837,
      "num_input_tokens_seen": 44361696,
      "step": 76415
    },
    {
      "epoch": 11.382186476020257,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.014000569034228815,
      "loss": 0.7961,
      "num_input_tokens_seen": 44364480,
      "step": 76420
    },
    {
      "epoch": 11.382931188561216,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.013998623722249677,
      "loss": 0.8071,
      "num_input_tokens_seen": 44367264,
      "step": 76425
    },
    {
      "epoch": 11.383675901102174,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.013996678427187768,
      "loss": 0.7879,
      "num_input_tokens_seen": 44370304,
      "step": 76430
    },
    {
      "epoch": 11.384420613643133,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.013994733149075962,
      "loss": 0.7895,
      "num_input_tokens_seen": 44373088,
      "step": 76435
    },
    {
      "epoch": 11.385165326184094,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.013992787887947118,
      "loss": 0.8181,
      "num_input_tokens_seen": 44375936,
      "step": 76440
    },
    {
      "epoch": 11.385910038725052,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.013990842643834094,
      "loss": 0.7998,
      "num_input_tokens_seen": 44378848,
      "step": 76445
    },
    {
      "epoch": 11.386654751266011,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.013988897416769763,
      "loss": 0.8412,
      "num_input_tokens_seen": 44381568,
      "step": 76450
    },
    {
      "epoch": 11.38739946380697,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.013986952206786979,
      "loss": 0.7881,
      "num_input_tokens_seen": 44384448,
      "step": 76455
    },
    {
      "epoch": 11.38814417634793,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.013985007013918612,
      "loss": 0.7944,
      "num_input_tokens_seen": 44387360,
      "step": 76460
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.013983061838197513,
      "loss": 0.8036,
      "num_input_tokens_seen": 44390112,
      "step": 76465
    },
    {
      "epoch": 11.389633601429848,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.013981116679656557,
      "loss": 0.781,
      "num_input_tokens_seen": 44393408,
      "step": 76470
    },
    {
      "epoch": 11.390378313970807,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.013979171538328592,
      "loss": 0.8153,
      "num_input_tokens_seen": 44396768,
      "step": 76475
    },
    {
      "epoch": 11.391123026511767,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.013977226414246494,
      "loss": 0.7902,
      "num_input_tokens_seen": 44399584,
      "step": 76480
    },
    {
      "epoch": 11.391867739052726,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.013975281307443117,
      "loss": 0.8184,
      "num_input_tokens_seen": 44402528,
      "step": 76485
    },
    {
      "epoch": 11.392612451593685,
      "grad_norm": 0.044921875,
      "learning_rate": 0.013973336217951316,
      "loss": 0.8187,
      "num_input_tokens_seen": 44405568,
      "step": 76490
    },
    {
      "epoch": 11.393357164134644,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.013971391145803959,
      "loss": 0.8221,
      "num_input_tokens_seen": 44408448,
      "step": 76495
    },
    {
      "epoch": 11.394101876675602,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.013969446091033899,
      "loss": 0.8066,
      "num_input_tokens_seen": 44411552,
      "step": 76500
    },
    {
      "epoch": 11.394846589216563,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.013967501053674005,
      "loss": 0.8041,
      "num_input_tokens_seen": 44414496,
      "step": 76505
    },
    {
      "epoch": 11.395591301757522,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.013965556033757124,
      "loss": 0.7951,
      "num_input_tokens_seen": 44417280,
      "step": 76510
    },
    {
      "epoch": 11.39633601429848,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.01396361103131613,
      "loss": 0.7959,
      "num_input_tokens_seen": 44420160,
      "step": 76515
    },
    {
      "epoch": 11.39708072683944,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.013961666046383873,
      "loss": 0.805,
      "num_input_tokens_seen": 44423136,
      "step": 76520
    },
    {
      "epoch": 11.3978254393804,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.01395972107899321,
      "loss": 0.8057,
      "num_input_tokens_seen": 44426112,
      "step": 76525
    },
    {
      "epoch": 11.398570151921358,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.013957776129177007,
      "loss": 0.7951,
      "num_input_tokens_seen": 44429056,
      "step": 76530
    },
    {
      "epoch": 11.399314864462317,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.013955831196968109,
      "loss": 0.7928,
      "num_input_tokens_seen": 44432032,
      "step": 76535
    },
    {
      "epoch": 11.400059577003276,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.013953886282399389,
      "loss": 0.7876,
      "num_input_tokens_seen": 44435008,
      "step": 76540
    },
    {
      "epoch": 11.400804289544237,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01395194138550369,
      "loss": 0.8037,
      "num_input_tokens_seen": 44437632,
      "step": 76545
    },
    {
      "epoch": 11.401549002085195,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.013949996506313882,
      "loss": 0.7889,
      "num_input_tokens_seen": 44440352,
      "step": 76550
    },
    {
      "epoch": 11.402293714626154,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.013948051644862815,
      "loss": 0.7864,
      "num_input_tokens_seen": 44443232,
      "step": 76555
    },
    {
      "epoch": 11.403038427167113,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.013946106801183349,
      "loss": 0.8128,
      "num_input_tokens_seen": 44445856,
      "step": 76560
    },
    {
      "epoch": 11.403783139708073,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.013944161975308336,
      "loss": 0.7964,
      "num_input_tokens_seen": 44448672,
      "step": 76565
    },
    {
      "epoch": 11.404527852249032,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.013942217167270626,
      "loss": 0.7915,
      "num_input_tokens_seen": 44451360,
      "step": 76570
    },
    {
      "epoch": 11.40527256478999,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.013940272377103088,
      "loss": 0.8162,
      "num_input_tokens_seen": 44454112,
      "step": 76575
    },
    {
      "epoch": 11.40601727733095,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.013938327604838567,
      "loss": 0.8131,
      "num_input_tokens_seen": 44456832,
      "step": 76580
    },
    {
      "epoch": 11.40676198987191,
      "grad_norm": 0.19140625,
      "learning_rate": 0.013936382850509929,
      "loss": 0.8283,
      "num_input_tokens_seen": 44459680,
      "step": 76585
    },
    {
      "epoch": 11.407506702412869,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.013934438114150016,
      "loss": 0.7991,
      "num_input_tokens_seen": 44462784,
      "step": 76590
    },
    {
      "epoch": 11.408251414953828,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.013932493395791693,
      "loss": 0.7955,
      "num_input_tokens_seen": 44465664,
      "step": 76595
    },
    {
      "epoch": 11.408996127494786,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.01393054869546781,
      "loss": 0.8046,
      "num_input_tokens_seen": 44468416,
      "step": 76600
    },
    {
      "epoch": 11.409740840035747,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.013928604013211212,
      "loss": 0.8157,
      "num_input_tokens_seen": 44471008,
      "step": 76605
    },
    {
      "epoch": 11.410485552576706,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.013926659349054765,
      "loss": 0.8142,
      "num_input_tokens_seen": 44474016,
      "step": 76610
    },
    {
      "epoch": 11.411230265117664,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.013924714703031313,
      "loss": 0.787,
      "num_input_tokens_seen": 44476832,
      "step": 76615
    },
    {
      "epoch": 11.411974977658623,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.01392277007517372,
      "loss": 0.8037,
      "num_input_tokens_seen": 44479840,
      "step": 76620
    },
    {
      "epoch": 11.412719690199584,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.013920825465514828,
      "loss": 0.7975,
      "num_input_tokens_seen": 44482848,
      "step": 76625
    },
    {
      "epoch": 11.413464402740543,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.013918880874087497,
      "loss": 0.7848,
      "num_input_tokens_seen": 44485696,
      "step": 76630
    },
    {
      "epoch": 11.414209115281501,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.013916936300924566,
      "loss": 0.7832,
      "num_input_tokens_seen": 44488608,
      "step": 76635
    },
    {
      "epoch": 11.41495382782246,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.013914991746058904,
      "loss": 0.8175,
      "num_input_tokens_seen": 44491424,
      "step": 76640
    },
    {
      "epoch": 11.41569854036342,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.013913047209523353,
      "loss": 0.7965,
      "num_input_tokens_seen": 44494144,
      "step": 76645
    },
    {
      "epoch": 11.41644325290438,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.013911102691350758,
      "loss": 0.7777,
      "num_input_tokens_seen": 44496992,
      "step": 76650
    },
    {
      "epoch": 11.417187965445338,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.013909158191573984,
      "loss": 0.8069,
      "num_input_tokens_seen": 44499872,
      "step": 76655
    },
    {
      "epoch": 11.417932677986297,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.013907213710225868,
      "loss": 0.8372,
      "num_input_tokens_seen": 44502816,
      "step": 76660
    },
    {
      "epoch": 11.418677390527257,
      "grad_norm": 0.03125,
      "learning_rate": 0.01390526924733927,
      "loss": 0.7973,
      "num_input_tokens_seen": 44505536,
      "step": 76665
    },
    {
      "epoch": 11.419422103068216,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.01390332480294703,
      "loss": 0.7968,
      "num_input_tokens_seen": 44508992,
      "step": 76670
    },
    {
      "epoch": 11.420166815609175,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.01390138037708201,
      "loss": 0.8066,
      "num_input_tokens_seen": 44511776,
      "step": 76675
    },
    {
      "epoch": 11.420911528150134,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.013899435969777049,
      "loss": 0.7816,
      "num_input_tokens_seen": 44514528,
      "step": 76680
    },
    {
      "epoch": 11.421656240691092,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.013897491581064994,
      "loss": 0.7954,
      "num_input_tokens_seen": 44517600,
      "step": 76685
    },
    {
      "epoch": 11.422400953232053,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.013895547210978704,
      "loss": 0.834,
      "num_input_tokens_seen": 44520576,
      "step": 76690
    },
    {
      "epoch": 11.423145665773012,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.01389360285955102,
      "loss": 0.8024,
      "num_input_tokens_seen": 44523360,
      "step": 76695
    },
    {
      "epoch": 11.42389037831397,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.013891658526814793,
      "loss": 0.7986,
      "num_input_tokens_seen": 44526144,
      "step": 76700
    },
    {
      "epoch": 11.42463509085493,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.013889714212802861,
      "loss": 0.7973,
      "num_input_tokens_seen": 44528960,
      "step": 76705
    },
    {
      "epoch": 11.42537980339589,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.013887769917548087,
      "loss": 0.8055,
      "num_input_tokens_seen": 44532128,
      "step": 76710
    },
    {
      "epoch": 11.426124515936849,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.013885825641083307,
      "loss": 0.787,
      "num_input_tokens_seen": 44535136,
      "step": 76715
    },
    {
      "epoch": 11.426869228477807,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.01388388138344137,
      "loss": 0.7905,
      "num_input_tokens_seen": 44538176,
      "step": 76720
    },
    {
      "epoch": 11.427613941018766,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.013881937144655125,
      "loss": 0.7855,
      "num_input_tokens_seen": 44541024,
      "step": 76725
    },
    {
      "epoch": 11.428358653559727,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.013879992924757412,
      "loss": 0.8005,
      "num_input_tokens_seen": 44543840,
      "step": 76730
    },
    {
      "epoch": 11.429103366100685,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.013878048723781086,
      "loss": 0.8011,
      "num_input_tokens_seen": 44547008,
      "step": 76735
    },
    {
      "epoch": 11.429848078641644,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.013876104541758978,
      "loss": 0.793,
      "num_input_tokens_seen": 44549632,
      "step": 76740
    },
    {
      "epoch": 11.430592791182603,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.013874160378723948,
      "loss": 0.8185,
      "num_input_tokens_seen": 44552608,
      "step": 76745
    },
    {
      "epoch": 11.431337503723563,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.013872216234708832,
      "loss": 0.8216,
      "num_input_tokens_seen": 44555392,
      "step": 76750
    },
    {
      "epoch": 11.432082216264522,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.013870272109746479,
      "loss": 0.7903,
      "num_input_tokens_seen": 44558400,
      "step": 76755
    },
    {
      "epoch": 11.432826928805481,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.013868328003869728,
      "loss": 0.7744,
      "num_input_tokens_seen": 44561216,
      "step": 76760
    },
    {
      "epoch": 11.43357164134644,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.013866383917111423,
      "loss": 0.8139,
      "num_input_tokens_seen": 44563936,
      "step": 76765
    },
    {
      "epoch": 11.4343163538874,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.013864439849504416,
      "loss": 0.8089,
      "num_input_tokens_seen": 44566816,
      "step": 76770
    },
    {
      "epoch": 11.435061066428359,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.013862495801081535,
      "loss": 0.7728,
      "num_input_tokens_seen": 44569536,
      "step": 76775
    },
    {
      "epoch": 11.435805778969318,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.013860551771875639,
      "loss": 0.7907,
      "num_input_tokens_seen": 44572448,
      "step": 76780
    },
    {
      "epoch": 11.436550491510276,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.013858607761919559,
      "loss": 0.7979,
      "num_input_tokens_seen": 44575264,
      "step": 76785
    },
    {
      "epoch": 11.437295204051237,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.013856663771246142,
      "loss": 0.8,
      "num_input_tokens_seen": 44578112,
      "step": 76790
    },
    {
      "epoch": 11.438039916592196,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.013854719799888225,
      "loss": 0.7793,
      "num_input_tokens_seen": 44580896,
      "step": 76795
    },
    {
      "epoch": 11.438784629133155,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.013852775847878659,
      "loss": 0.7821,
      "num_input_tokens_seen": 44583936,
      "step": 76800
    },
    {
      "epoch": 11.439529341674113,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.013850831915250278,
      "loss": 0.8029,
      "num_input_tokens_seen": 44586944,
      "step": 76805
    },
    {
      "epoch": 11.440274054215074,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.013848888002035917,
      "loss": 0.7888,
      "num_input_tokens_seen": 44589920,
      "step": 76810
    },
    {
      "epoch": 11.441018766756033,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01384694410826843,
      "loss": 0.8109,
      "num_input_tokens_seen": 44592992,
      "step": 76815
    },
    {
      "epoch": 11.441763479296991,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.013845000233980649,
      "loss": 0.782,
      "num_input_tokens_seen": 44596000,
      "step": 76820
    },
    {
      "epoch": 11.44250819183795,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.013843056379205416,
      "loss": 0.7894,
      "num_input_tokens_seen": 44598720,
      "step": 76825
    },
    {
      "epoch": 11.443252904378909,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.013841112543975565,
      "loss": 0.7875,
      "num_input_tokens_seen": 44601440,
      "step": 76830
    },
    {
      "epoch": 11.44399761691987,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.013839168728323943,
      "loss": 0.8038,
      "num_input_tokens_seen": 44604608,
      "step": 76835
    },
    {
      "epoch": 11.444742329460828,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.013837224932283388,
      "loss": 0.8102,
      "num_input_tokens_seen": 44607744,
      "step": 76840
    },
    {
      "epoch": 11.445487042001787,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.01383528115588673,
      "loss": 0.7999,
      "num_input_tokens_seen": 44610528,
      "step": 76845
    },
    {
      "epoch": 11.446231754542746,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.01383333739916682,
      "loss": 0.8104,
      "num_input_tokens_seen": 44613184,
      "step": 76850
    },
    {
      "epoch": 11.446976467083706,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.013831393662156485,
      "loss": 0.7917,
      "num_input_tokens_seen": 44616288,
      "step": 76855
    },
    {
      "epoch": 11.447721179624665,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.013829449944888567,
      "loss": 0.7794,
      "num_input_tokens_seen": 44619104,
      "step": 76860
    },
    {
      "epoch": 11.448465892165624,
      "grad_norm": 0.05859375,
      "learning_rate": 0.013827506247395899,
      "loss": 0.8103,
      "num_input_tokens_seen": 44621792,
      "step": 76865
    },
    {
      "epoch": 11.449210604706582,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.013825562569711328,
      "loss": 0.7891,
      "num_input_tokens_seen": 44624832,
      "step": 76870
    },
    {
      "epoch": 11.449955317247543,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.013823618911867674,
      "loss": 0.7873,
      "num_input_tokens_seen": 44627552,
      "step": 76875
    },
    {
      "epoch": 11.450700029788502,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.013821675273897794,
      "loss": 0.7948,
      "num_input_tokens_seen": 44630592,
      "step": 76880
    },
    {
      "epoch": 11.45144474232946,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.01381973165583451,
      "loss": 0.7932,
      "num_input_tokens_seen": 44633472,
      "step": 76885
    },
    {
      "epoch": 11.45218945487042,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.013817788057710657,
      "loss": 0.7982,
      "num_input_tokens_seen": 44636224,
      "step": 76890
    },
    {
      "epoch": 11.45293416741138,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.013815844479559075,
      "loss": 0.7779,
      "num_input_tokens_seen": 44639104,
      "step": 76895
    },
    {
      "epoch": 11.453678879952339,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.013813900921412593,
      "loss": 0.7677,
      "num_input_tokens_seen": 44642112,
      "step": 76900
    },
    {
      "epoch": 11.454423592493297,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.013811957383304055,
      "loss": 0.7873,
      "num_input_tokens_seen": 44644896,
      "step": 76905
    },
    {
      "epoch": 11.455168305034256,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.013810013865266282,
      "loss": 0.7784,
      "num_input_tokens_seen": 44647776,
      "step": 76910
    },
    {
      "epoch": 11.455913017575217,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.013808070367332123,
      "loss": 0.774,
      "num_input_tokens_seen": 44650912,
      "step": 76915
    },
    {
      "epoch": 11.456657730116175,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.013806126889534401,
      "loss": 0.8082,
      "num_input_tokens_seen": 44653728,
      "step": 76920
    },
    {
      "epoch": 11.457402442657134,
      "grad_norm": 0.046875,
      "learning_rate": 0.013804183431905955,
      "loss": 0.7927,
      "num_input_tokens_seen": 44656832,
      "step": 76925
    },
    {
      "epoch": 11.458147155198093,
      "grad_norm": 0.072265625,
      "learning_rate": 0.013802239994479614,
      "loss": 0.7869,
      "num_input_tokens_seen": 44659936,
      "step": 76930
    },
    {
      "epoch": 11.458891867739053,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.013800296577288203,
      "loss": 0.8048,
      "num_input_tokens_seen": 44662656,
      "step": 76935
    },
    {
      "epoch": 11.459636580280012,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.013798353180364572,
      "loss": 0.8232,
      "num_input_tokens_seen": 44665248,
      "step": 76940
    },
    {
      "epoch": 11.460381292820971,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.013796409803741534,
      "loss": 0.8117,
      "num_input_tokens_seen": 44668128,
      "step": 76945
    },
    {
      "epoch": 11.46112600536193,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.013794466447451935,
      "loss": 0.806,
      "num_input_tokens_seen": 44671072,
      "step": 76950
    },
    {
      "epoch": 11.46187071790289,
      "grad_norm": 0.029296875,
      "learning_rate": 0.013792523111528596,
      "loss": 0.7931,
      "num_input_tokens_seen": 44674240,
      "step": 76955
    },
    {
      "epoch": 11.462615430443849,
      "grad_norm": 0.05859375,
      "learning_rate": 0.013790579796004357,
      "loss": 0.7995,
      "num_input_tokens_seen": 44677280,
      "step": 76960
    },
    {
      "epoch": 11.463360142984808,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.013788636500912043,
      "loss": 0.8111,
      "num_input_tokens_seen": 44680288,
      "step": 76965
    },
    {
      "epoch": 11.464104855525767,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.013786693226284478,
      "loss": 0.8037,
      "num_input_tokens_seen": 44683040,
      "step": 76970
    },
    {
      "epoch": 11.464849568066727,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.013784749972154503,
      "loss": 0.7995,
      "num_input_tokens_seen": 44686080,
      "step": 76975
    },
    {
      "epoch": 11.465594280607686,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.013782806738554936,
      "loss": 0.7899,
      "num_input_tokens_seen": 44689088,
      "step": 76980
    },
    {
      "epoch": 11.466338993148645,
      "grad_norm": 0.029296875,
      "learning_rate": 0.013780863525518619,
      "loss": 0.8062,
      "num_input_tokens_seen": 44691840,
      "step": 76985
    },
    {
      "epoch": 11.467083705689603,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.01377892033307837,
      "loss": 0.8012,
      "num_input_tokens_seen": 44694944,
      "step": 76990
    },
    {
      "epoch": 11.467828418230564,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.013776977161267025,
      "loss": 0.7825,
      "num_input_tokens_seen": 44697664,
      "step": 76995
    },
    {
      "epoch": 11.468573130771523,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.0137750340101174,
      "loss": 0.8027,
      "num_input_tokens_seen": 44700672,
      "step": 77000
    },
    {
      "epoch": 11.469317843312481,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.01377309087966234,
      "loss": 0.8041,
      "num_input_tokens_seen": 44703488,
      "step": 77005
    },
    {
      "epoch": 11.47006255585344,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.013771147769934659,
      "loss": 0.7725,
      "num_input_tokens_seen": 44706624,
      "step": 77010
    },
    {
      "epoch": 11.470807268394399,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.013769204680967182,
      "loss": 0.7772,
      "num_input_tokens_seen": 44709440,
      "step": 77015
    },
    {
      "epoch": 11.47155198093536,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.013767261612792748,
      "loss": 0.7775,
      "num_input_tokens_seen": 44712192,
      "step": 77020
    },
    {
      "epoch": 11.472296693476318,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.013765318565444173,
      "loss": 0.7717,
      "num_input_tokens_seen": 44715424,
      "step": 77025
    },
    {
      "epoch": 11.473041406017277,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.013763375538954287,
      "loss": 0.7709,
      "num_input_tokens_seen": 44718400,
      "step": 77030
    },
    {
      "epoch": 11.473786118558236,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.013761432533355912,
      "loss": 0.7991,
      "num_input_tokens_seen": 44721504,
      "step": 77035
    },
    {
      "epoch": 11.474530831099196,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01375948954868188,
      "loss": 0.7795,
      "num_input_tokens_seen": 44724352,
      "step": 77040
    },
    {
      "epoch": 11.475275543640155,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.013757546584965011,
      "loss": 0.7838,
      "num_input_tokens_seen": 44727296,
      "step": 77045
    },
    {
      "epoch": 11.476020256181114,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.013755603642238122,
      "loss": 0.8436,
      "num_input_tokens_seen": 44730176,
      "step": 77050
    },
    {
      "epoch": 11.476764968722073,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.013753660720534052,
      "loss": 0.8155,
      "num_input_tokens_seen": 44733184,
      "step": 77055
    },
    {
      "epoch": 11.477509681263033,
      "grad_norm": 0.078125,
      "learning_rate": 0.013751717819885616,
      "loss": 0.7943,
      "num_input_tokens_seen": 44735872,
      "step": 77060
    },
    {
      "epoch": 11.478254393803992,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.01374977494032564,
      "loss": 0.798,
      "num_input_tokens_seen": 44738784,
      "step": 77065
    },
    {
      "epoch": 11.47899910634495,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.01374783208188694,
      "loss": 0.8018,
      "num_input_tokens_seen": 44741888,
      "step": 77070
    },
    {
      "epoch": 11.47974381888591,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.01374588924460235,
      "loss": 0.7801,
      "num_input_tokens_seen": 44745024,
      "step": 77075
    },
    {
      "epoch": 11.48048853142687,
      "grad_norm": 0.03515625,
      "learning_rate": 0.013743946428504685,
      "loss": 0.8315,
      "num_input_tokens_seen": 44748288,
      "step": 77080
    },
    {
      "epoch": 11.481233243967829,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.01374200363362677,
      "loss": 0.7817,
      "num_input_tokens_seen": 44751168,
      "step": 77085
    },
    {
      "epoch": 11.481977956508787,
      "grad_norm": 0.068359375,
      "learning_rate": 0.013740060860001423,
      "loss": 0.7568,
      "num_input_tokens_seen": 44754240,
      "step": 77090
    },
    {
      "epoch": 11.482722669049746,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.013738118107661465,
      "loss": 0.7832,
      "num_input_tokens_seen": 44756736,
      "step": 77095
    },
    {
      "epoch": 11.483467381590707,
      "grad_norm": 0.087890625,
      "learning_rate": 0.013736175376639724,
      "loss": 0.7988,
      "num_input_tokens_seen": 44759744,
      "step": 77100
    },
    {
      "epoch": 11.484212094131665,
      "grad_norm": 0.046875,
      "learning_rate": 0.01373423266696901,
      "loss": 0.7313,
      "num_input_tokens_seen": 44762688,
      "step": 77105
    },
    {
      "epoch": 11.484956806672624,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.013732289978682154,
      "loss": 0.8334,
      "num_input_tokens_seen": 44765440,
      "step": 77110
    },
    {
      "epoch": 11.485701519213583,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.013730347311811967,
      "loss": 0.7753,
      "num_input_tokens_seen": 44768256,
      "step": 77115
    },
    {
      "epoch": 11.486446231754543,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.013728404666391275,
      "loss": 0.8086,
      "num_input_tokens_seen": 44771136,
      "step": 77120
    },
    {
      "epoch": 11.487190944295502,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.013726462042452894,
      "loss": 0.8238,
      "num_input_tokens_seen": 44774240,
      "step": 77125
    },
    {
      "epoch": 11.487935656836461,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.013724519440029637,
      "loss": 0.8103,
      "num_input_tokens_seen": 44776832,
      "step": 77130
    },
    {
      "epoch": 11.48868036937742,
      "grad_norm": 0.044921875,
      "learning_rate": 0.013722576859154333,
      "loss": 0.7913,
      "num_input_tokens_seen": 44779424,
      "step": 77135
    },
    {
      "epoch": 11.48942508191838,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.013720634299859789,
      "loss": 0.8009,
      "num_input_tokens_seen": 44782272,
      "step": 77140
    },
    {
      "epoch": 11.490169794459339,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.013718691762178834,
      "loss": 0.8054,
      "num_input_tokens_seen": 44785280,
      "step": 77145
    },
    {
      "epoch": 11.490914507000298,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.013716749246144275,
      "loss": 0.8209,
      "num_input_tokens_seen": 44788160,
      "step": 77150
    },
    {
      "epoch": 11.491659219541257,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.013714806751788938,
      "loss": 0.7918,
      "num_input_tokens_seen": 44790848,
      "step": 77155
    },
    {
      "epoch": 11.492403932082217,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.013712864279145628,
      "loss": 0.8047,
      "num_input_tokens_seen": 44793952,
      "step": 77160
    },
    {
      "epoch": 11.493148644623176,
      "grad_norm": 0.0546875,
      "learning_rate": 0.013710921828247175,
      "loss": 0.7929,
      "num_input_tokens_seen": 44796864,
      "step": 77165
    },
    {
      "epoch": 11.493893357164135,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.013708979399126386,
      "loss": 0.8008,
      "num_input_tokens_seen": 44799872,
      "step": 77170
    },
    {
      "epoch": 11.494638069705093,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.013707036991816073,
      "loss": 0.8202,
      "num_input_tokens_seen": 44802688,
      "step": 77175
    },
    {
      "epoch": 11.495382782246054,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01370509460634906,
      "loss": 0.8334,
      "num_input_tokens_seen": 44805440,
      "step": 77180
    },
    {
      "epoch": 11.496127494787013,
      "grad_norm": 0.05078125,
      "learning_rate": 0.013703152242758157,
      "loss": 0.8087,
      "num_input_tokens_seen": 44808352,
      "step": 77185
    },
    {
      "epoch": 11.496872207327971,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.013701209901076181,
      "loss": 0.7697,
      "num_input_tokens_seen": 44811136,
      "step": 77190
    },
    {
      "epoch": 11.49761691986893,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.013699267581335937,
      "loss": 0.7962,
      "num_input_tokens_seen": 44813984,
      "step": 77195
    },
    {
      "epoch": 11.498361632409889,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.013697325283570252,
      "loss": 0.8146,
      "num_input_tokens_seen": 44816544,
      "step": 77200
    },
    {
      "epoch": 11.49910634495085,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.013695383007811932,
      "loss": 0.8195,
      "num_input_tokens_seen": 44819488,
      "step": 77205
    },
    {
      "epoch": 11.499851057491808,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.013693440754093783,
      "loss": 0.8417,
      "num_input_tokens_seen": 44822432,
      "step": 77210
    },
    {
      "epoch": 11.500595770032767,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.013691498522448633,
      "loss": 0.7886,
      "num_input_tokens_seen": 44825408,
      "step": 77215
    },
    {
      "epoch": 11.501340482573726,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.013689556312909282,
      "loss": 0.8027,
      "num_input_tokens_seen": 44828256,
      "step": 77220
    },
    {
      "epoch": 11.502085195114686,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.013687614125508547,
      "loss": 0.8123,
      "num_input_tokens_seen": 44831360,
      "step": 77225
    },
    {
      "epoch": 11.502829907655645,
      "grad_norm": 0.0390625,
      "learning_rate": 0.013685671960279233,
      "loss": 0.8061,
      "num_input_tokens_seen": 44834176,
      "step": 77230
    },
    {
      "epoch": 11.503574620196604,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.013683729817254159,
      "loss": 0.8049,
      "num_input_tokens_seen": 44836864,
      "step": 77235
    },
    {
      "epoch": 11.504319332737563,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.013681787696466127,
      "loss": 0.7941,
      "num_input_tokens_seen": 44839680,
      "step": 77240
    },
    {
      "epoch": 11.505064045278523,
      "grad_norm": 0.056640625,
      "learning_rate": 0.01367984559794796,
      "loss": 0.7933,
      "num_input_tokens_seen": 44842912,
      "step": 77245
    },
    {
      "epoch": 11.505808757819482,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.013677903521732458,
      "loss": 0.8177,
      "num_input_tokens_seen": 44846016,
      "step": 77250
    },
    {
      "epoch": 11.50655347036044,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.013675961467852429,
      "loss": 0.7988,
      "num_input_tokens_seen": 44849120,
      "step": 77255
    },
    {
      "epoch": 11.5072981829014,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.01367401943634069,
      "loss": 0.7953,
      "num_input_tokens_seen": 44852000,
      "step": 77260
    },
    {
      "epoch": 11.50804289544236,
      "grad_norm": 0.044921875,
      "learning_rate": 0.013672077427230037,
      "loss": 0.819,
      "num_input_tokens_seen": 44854912,
      "step": 77265
    },
    {
      "epoch": 11.508787607983319,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.013670135440553292,
      "loss": 0.7809,
      "num_input_tokens_seen": 44857760,
      "step": 77270
    },
    {
      "epoch": 11.509532320524277,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.013668193476343253,
      "loss": 0.807,
      "num_input_tokens_seen": 44860512,
      "step": 77275
    },
    {
      "epoch": 11.510277033065236,
      "grad_norm": 0.0703125,
      "learning_rate": 0.013666251534632738,
      "loss": 0.8254,
      "num_input_tokens_seen": 44863680,
      "step": 77280
    },
    {
      "epoch": 11.511021745606197,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.013664309615454544,
      "loss": 0.7958,
      "num_input_tokens_seen": 44866528,
      "step": 77285
    },
    {
      "epoch": 11.511766458147155,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.013662367718841483,
      "loss": 0.7988,
      "num_input_tokens_seen": 44869568,
      "step": 77290
    },
    {
      "epoch": 11.512511170688114,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.01366042584482636,
      "loss": 0.8013,
      "num_input_tokens_seen": 44872608,
      "step": 77295
    },
    {
      "epoch": 11.513255883229073,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.013658483993441977,
      "loss": 0.7943,
      "num_input_tokens_seen": 44875680,
      "step": 77300
    },
    {
      "epoch": 11.514000595770034,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.013656542164721147,
      "loss": 0.8062,
      "num_input_tokens_seen": 44878400,
      "step": 77305
    },
    {
      "epoch": 11.514745308310992,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.013654600358696667,
      "loss": 0.7806,
      "num_input_tokens_seen": 44881184,
      "step": 77310
    },
    {
      "epoch": 11.515490020851951,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.013652658575401353,
      "loss": 0.8193,
      "num_input_tokens_seen": 44884384,
      "step": 77315
    },
    {
      "epoch": 11.51623473339291,
      "grad_norm": 0.03125,
      "learning_rate": 0.013650716814868,
      "loss": 0.8002,
      "num_input_tokens_seen": 44887264,
      "step": 77320
    },
    {
      "epoch": 11.51697944593387,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.013648775077129418,
      "loss": 0.7743,
      "num_input_tokens_seen": 44889952,
      "step": 77325
    },
    {
      "epoch": 11.517724158474829,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.013646833362218407,
      "loss": 0.8143,
      "num_input_tokens_seen": 44893152,
      "step": 77330
    },
    {
      "epoch": 11.518468871015788,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.013644891670167766,
      "loss": 0.7922,
      "num_input_tokens_seen": 44896128,
      "step": 77335
    },
    {
      "epoch": 11.519213583556747,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.013642950001010307,
      "loss": 0.7963,
      "num_input_tokens_seen": 44898976,
      "step": 77340
    },
    {
      "epoch": 11.519958296097705,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.013641008354778824,
      "loss": 0.7991,
      "num_input_tokens_seen": 44901536,
      "step": 77345
    },
    {
      "epoch": 11.520703008638666,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.01363906673150613,
      "loss": 0.8039,
      "num_input_tokens_seen": 44904256,
      "step": 77350
    },
    {
      "epoch": 11.521447721179625,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.013637125131225017,
      "loss": 0.8029,
      "num_input_tokens_seen": 44907104,
      "step": 77355
    },
    {
      "epoch": 11.522192433720583,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.013635183553968293,
      "loss": 0.814,
      "num_input_tokens_seen": 44909856,
      "step": 77360
    },
    {
      "epoch": 11.522937146261544,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.013633241999768755,
      "loss": 0.8168,
      "num_input_tokens_seen": 44912928,
      "step": 77365
    },
    {
      "epoch": 11.523681858802503,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.013631300468659199,
      "loss": 0.8023,
      "num_input_tokens_seen": 44915648,
      "step": 77370
    },
    {
      "epoch": 11.524426571343461,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.013629358960672435,
      "loss": 0.7954,
      "num_input_tokens_seen": 44918496,
      "step": 77375
    },
    {
      "epoch": 11.52517128388442,
      "grad_norm": 0.09765625,
      "learning_rate": 0.013627417475841255,
      "loss": 0.8089,
      "num_input_tokens_seen": 44921312,
      "step": 77380
    },
    {
      "epoch": 11.525915996425379,
      "grad_norm": 0.03515625,
      "learning_rate": 0.013625476014198466,
      "loss": 0.7726,
      "num_input_tokens_seen": 44924160,
      "step": 77385
    },
    {
      "epoch": 11.52666070896634,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01362353457577686,
      "loss": 0.8057,
      "num_input_tokens_seen": 44926944,
      "step": 77390
    },
    {
      "epoch": 11.527405421507298,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01362159316060924,
      "loss": 0.8038,
      "num_input_tokens_seen": 44929856,
      "step": 77395
    },
    {
      "epoch": 11.528150134048257,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.013619651768728399,
      "loss": 0.8013,
      "num_input_tokens_seen": 44932832,
      "step": 77400
    },
    {
      "epoch": 11.528894846589216,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.013617710400167142,
      "loss": 0.8093,
      "num_input_tokens_seen": 44936064,
      "step": 77405
    },
    {
      "epoch": 11.529639559130176,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.013615769054958265,
      "loss": 0.8067,
      "num_input_tokens_seen": 44938944,
      "step": 77410
    },
    {
      "epoch": 11.530384271671135,
      "grad_norm": 0.044921875,
      "learning_rate": 0.01361382773313456,
      "loss": 0.8162,
      "num_input_tokens_seen": 44941504,
      "step": 77415
    },
    {
      "epoch": 11.531128984212094,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.013611886434728826,
      "loss": 0.7948,
      "num_input_tokens_seen": 44944352,
      "step": 77420
    },
    {
      "epoch": 11.531873696753053,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.013609945159773859,
      "loss": 0.8009,
      "num_input_tokens_seen": 44947072,
      "step": 77425
    },
    {
      "epoch": 11.532618409294013,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.01360800390830246,
      "loss": 0.7899,
      "num_input_tokens_seen": 44950208,
      "step": 77430
    },
    {
      "epoch": 11.533363121834972,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.013606062680347412,
      "loss": 0.7903,
      "num_input_tokens_seen": 44953024,
      "step": 77435
    },
    {
      "epoch": 11.53410783437593,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.013604121475941526,
      "loss": 0.8096,
      "num_input_tokens_seen": 44955872,
      "step": 77440
    },
    {
      "epoch": 11.53485254691689,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.013602180295117585,
      "loss": 0.7806,
      "num_input_tokens_seen": 44958720,
      "step": 77445
    },
    {
      "epoch": 11.53559725945785,
      "grad_norm": 0.029296875,
      "learning_rate": 0.013600239137908393,
      "loss": 0.8182,
      "num_input_tokens_seen": 44961408,
      "step": 77450
    },
    {
      "epoch": 11.536341971998809,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.013598298004346735,
      "loss": 0.7954,
      "num_input_tokens_seen": 44964288,
      "step": 77455
    },
    {
      "epoch": 11.537086684539767,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.013596356894465403,
      "loss": 0.7734,
      "num_input_tokens_seen": 44967040,
      "step": 77460
    },
    {
      "epoch": 11.537831397080726,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0135944158082972,
      "loss": 0.7872,
      "num_input_tokens_seen": 44969952,
      "step": 77465
    },
    {
      "epoch": 11.538576109621687,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.013592474745874909,
      "loss": 0.8199,
      "num_input_tokens_seen": 44972896,
      "step": 77470
    },
    {
      "epoch": 11.539320822162646,
      "grad_norm": 0.033203125,
      "learning_rate": 0.013590533707231331,
      "loss": 0.7851,
      "num_input_tokens_seen": 44975840,
      "step": 77475
    },
    {
      "epoch": 11.540065534703604,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01358859269239925,
      "loss": 0.7837,
      "num_input_tokens_seen": 44978688,
      "step": 77480
    },
    {
      "epoch": 11.540810247244563,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.013586651701411465,
      "loss": 0.8027,
      "num_input_tokens_seen": 44981600,
      "step": 77485
    },
    {
      "epoch": 11.541554959785524,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.013584710734300764,
      "loss": 0.7834,
      "num_input_tokens_seen": 44984544,
      "step": 77490
    },
    {
      "epoch": 11.542299672326482,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.01358276979109993,
      "loss": 0.7921,
      "num_input_tokens_seen": 44987232,
      "step": 77495
    },
    {
      "epoch": 11.543044384867441,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.013580828871841765,
      "loss": 0.8055,
      "num_input_tokens_seen": 44990272,
      "step": 77500
    },
    {
      "epoch": 11.5437890974084,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.013578887976559048,
      "loss": 0.781,
      "num_input_tokens_seen": 44993472,
      "step": 77505
    },
    {
      "epoch": 11.54453380994936,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.01357694710528458,
      "loss": 0.7948,
      "num_input_tokens_seen": 44996256,
      "step": 77510
    },
    {
      "epoch": 11.54527852249032,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.013575006258051142,
      "loss": 0.7672,
      "num_input_tokens_seen": 44999232,
      "step": 77515
    },
    {
      "epoch": 11.546023235031278,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.01357306543489153,
      "loss": 0.8476,
      "num_input_tokens_seen": 45001952,
      "step": 77520
    },
    {
      "epoch": 11.546767947572237,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.013571124635838518,
      "loss": 0.8025,
      "num_input_tokens_seen": 45004864,
      "step": 77525
    },
    {
      "epoch": 11.547512660113195,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.01356918386092491,
      "loss": 0.8201,
      "num_input_tokens_seen": 45007584,
      "step": 77530
    },
    {
      "epoch": 11.548257372654156,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.013567243110183487,
      "loss": 0.8313,
      "num_input_tokens_seen": 45010464,
      "step": 77535
    },
    {
      "epoch": 11.549002085195115,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.013565302383647029,
      "loss": 0.811,
      "num_input_tokens_seen": 45013504,
      "step": 77540
    },
    {
      "epoch": 11.549746797736073,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.013563361681348335,
      "loss": 0.8172,
      "num_input_tokens_seen": 45016544,
      "step": 77545
    },
    {
      "epoch": 11.550491510277032,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.013561421003320182,
      "loss": 0.8292,
      "num_input_tokens_seen": 45019456,
      "step": 77550
    },
    {
      "epoch": 11.551236222817993,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.013559480349595364,
      "loss": 0.7876,
      "num_input_tokens_seen": 45022464,
      "step": 77555
    },
    {
      "epoch": 11.551980935358952,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.013557539720206652,
      "loss": 0.7802,
      "num_input_tokens_seen": 45025664,
      "step": 77560
    },
    {
      "epoch": 11.55272564789991,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.01355559911518685,
      "loss": 0.7941,
      "num_input_tokens_seen": 45028512,
      "step": 77565
    },
    {
      "epoch": 11.553470360440869,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.01355365853456873,
      "loss": 0.8078,
      "num_input_tokens_seen": 45031584,
      "step": 77570
    },
    {
      "epoch": 11.55421507298183,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.013551717978385074,
      "loss": 0.8016,
      "num_input_tokens_seen": 45034560,
      "step": 77575
    },
    {
      "epoch": 11.554959785522788,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.013549777446668677,
      "loss": 0.795,
      "num_input_tokens_seen": 45037504,
      "step": 77580
    },
    {
      "epoch": 11.555704498063747,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.013547836939452313,
      "loss": 0.7797,
      "num_input_tokens_seen": 45040352,
      "step": 77585
    },
    {
      "epoch": 11.556449210604706,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.013545896456768772,
      "loss": 0.789,
      "num_input_tokens_seen": 45043328,
      "step": 77590
    },
    {
      "epoch": 11.557193923145666,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.013543955998650825,
      "loss": 0.8163,
      "num_input_tokens_seen": 45047488,
      "step": 77595
    },
    {
      "epoch": 11.557938635686625,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01354201556513127,
      "loss": 0.7829,
      "num_input_tokens_seen": 45050336,
      "step": 77600
    },
    {
      "epoch": 11.558683348227584,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.013540075156242871,
      "loss": 0.8147,
      "num_input_tokens_seen": 45053472,
      "step": 77605
    },
    {
      "epoch": 11.559428060768543,
      "grad_norm": 0.0390625,
      "learning_rate": 0.013538134772018425,
      "loss": 0.8067,
      "num_input_tokens_seen": 45056192,
      "step": 77610
    },
    {
      "epoch": 11.560172773309503,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.013536194412490708,
      "loss": 0.8291,
      "num_input_tokens_seen": 45058880,
      "step": 77615
    },
    {
      "epoch": 11.560917485850462,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.013534254077692496,
      "loss": 0.8019,
      "num_input_tokens_seen": 45061664,
      "step": 77620
    },
    {
      "epoch": 11.56166219839142,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.013532313767656573,
      "loss": 0.7745,
      "num_input_tokens_seen": 45064416,
      "step": 77625
    },
    {
      "epoch": 11.56240691093238,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.013530373482415712,
      "loss": 0.7802,
      "num_input_tokens_seen": 45067232,
      "step": 77630
    },
    {
      "epoch": 11.56315162347334,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.013528433222002702,
      "loss": 0.7994,
      "num_input_tokens_seen": 45070112,
      "step": 77635
    },
    {
      "epoch": 11.563896336014299,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.013526492986450311,
      "loss": 0.7904,
      "num_input_tokens_seen": 45073440,
      "step": 77640
    },
    {
      "epoch": 11.564641048555258,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.013524552775791331,
      "loss": 0.7975,
      "num_input_tokens_seen": 45076416,
      "step": 77645
    },
    {
      "epoch": 11.565385761096216,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.013522612590058531,
      "loss": 0.8313,
      "num_input_tokens_seen": 45079424,
      "step": 77650
    },
    {
      "epoch": 11.566130473637177,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.013520672429284689,
      "loss": 0.8096,
      "num_input_tokens_seen": 45082368,
      "step": 77655
    },
    {
      "epoch": 11.566875186178136,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.013518732293502584,
      "loss": 0.787,
      "num_input_tokens_seen": 45085280,
      "step": 77660
    },
    {
      "epoch": 11.567619898719094,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.013516792182744986,
      "loss": 0.791,
      "num_input_tokens_seen": 45088416,
      "step": 77665
    },
    {
      "epoch": 11.568364611260053,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.013514852097044682,
      "loss": 0.788,
      "num_input_tokens_seen": 45091520,
      "step": 77670
    },
    {
      "epoch": 11.569109323801012,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.013512912036434436,
      "loss": 0.8124,
      "num_input_tokens_seen": 45094368,
      "step": 77675
    },
    {
      "epoch": 11.569854036341972,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.013510972000947035,
      "loss": 0.8074,
      "num_input_tokens_seen": 45097184,
      "step": 77680
    },
    {
      "epoch": 11.570598748882931,
      "grad_norm": 0.05078125,
      "learning_rate": 0.013509031990615247,
      "loss": 0.8177,
      "num_input_tokens_seen": 45100000,
      "step": 77685
    },
    {
      "epoch": 11.57134346142389,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.013507092005471852,
      "loss": 0.8191,
      "num_input_tokens_seen": 45102848,
      "step": 77690
    },
    {
      "epoch": 11.57208817396485,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.01350515204554962,
      "loss": 0.8085,
      "num_input_tokens_seen": 45105472,
      "step": 77695
    },
    {
      "epoch": 11.57283288650581,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.013503212110881319,
      "loss": 0.8305,
      "num_input_tokens_seen": 45108384,
      "step": 77700
    },
    {
      "epoch": 11.573577599046768,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.013501272201499733,
      "loss": 0.7909,
      "num_input_tokens_seen": 45111680,
      "step": 77705
    },
    {
      "epoch": 11.574322311587727,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.013499332317437622,
      "loss": 0.8307,
      "num_input_tokens_seen": 45114304,
      "step": 77710
    },
    {
      "epoch": 11.575067024128685,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.013497392458727777,
      "loss": 0.7944,
      "num_input_tokens_seen": 45117088,
      "step": 77715
    },
    {
      "epoch": 11.575811736669646,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.013495452625402953,
      "loss": 0.7959,
      "num_input_tokens_seen": 45120000,
      "step": 77720
    },
    {
      "epoch": 11.576556449210605,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.013493512817495931,
      "loss": 0.7954,
      "num_input_tokens_seen": 45122976,
      "step": 77725
    },
    {
      "epoch": 11.577301161751564,
      "grad_norm": 0.04296875,
      "learning_rate": 0.01349157303503948,
      "loss": 0.7854,
      "num_input_tokens_seen": 45125952,
      "step": 77730
    },
    {
      "epoch": 11.578045874292522,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.013489633278066362,
      "loss": 0.7889,
      "num_input_tokens_seen": 45128736,
      "step": 77735
    },
    {
      "epoch": 11.578790586833483,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.013487693546609362,
      "loss": 0.8166,
      "num_input_tokens_seen": 45131840,
      "step": 77740
    },
    {
      "epoch": 11.579535299374442,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.013485753840701241,
      "loss": 0.8179,
      "num_input_tokens_seen": 45135296,
      "step": 77745
    },
    {
      "epoch": 11.5802800119154,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.013483814160374766,
      "loss": 0.8009,
      "num_input_tokens_seen": 45138240,
      "step": 77750
    },
    {
      "epoch": 11.581024724456359,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.013481874505662713,
      "loss": 0.7825,
      "num_input_tokens_seen": 45141088,
      "step": 77755
    },
    {
      "epoch": 11.58176943699732,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.013479934876597849,
      "loss": 0.8003,
      "num_input_tokens_seen": 45144288,
      "step": 77760
    },
    {
      "epoch": 11.582514149538278,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.013477995273212934,
      "loss": 0.8016,
      "num_input_tokens_seen": 45147264,
      "step": 77765
    },
    {
      "epoch": 11.583258862079237,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.013476055695540749,
      "loss": 0.7975,
      "num_input_tokens_seen": 45150176,
      "step": 77770
    },
    {
      "epoch": 11.584003574620196,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.013474116143614052,
      "loss": 0.803,
      "num_input_tokens_seen": 45152992,
      "step": 77775
    },
    {
      "epoch": 11.584748287161156,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.01347217661746561,
      "loss": 0.7883,
      "num_input_tokens_seen": 45155808,
      "step": 77780
    },
    {
      "epoch": 11.585492999702115,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.013470237117128196,
      "loss": 0.8383,
      "num_input_tokens_seen": 45158528,
      "step": 77785
    },
    {
      "epoch": 11.586237712243074,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.013468297642634564,
      "loss": 0.8074,
      "num_input_tokens_seen": 45161280,
      "step": 77790
    },
    {
      "epoch": 11.586982424784033,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.013466358194017492,
      "loss": 0.7814,
      "num_input_tokens_seen": 45164160,
      "step": 77795
    },
    {
      "epoch": 11.587727137324993,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.013464418771309734,
      "loss": 0.8017,
      "num_input_tokens_seen": 45167200,
      "step": 77800
    },
    {
      "epoch": 11.588471849865952,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.013462479374544065,
      "loss": 0.7949,
      "num_input_tokens_seen": 45169856,
      "step": 77805
    },
    {
      "epoch": 11.58921656240691,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.013460540003753246,
      "loss": 0.8018,
      "num_input_tokens_seen": 45172544,
      "step": 77810
    },
    {
      "epoch": 11.58996127494787,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.013458600658970036,
      "loss": 0.7868,
      "num_input_tokens_seen": 45175616,
      "step": 77815
    },
    {
      "epoch": 11.59070598748883,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.013456661340227202,
      "loss": 0.7893,
      "num_input_tokens_seen": 45178560,
      "step": 77820
    },
    {
      "epoch": 11.591450700029789,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.013454722047557502,
      "loss": 0.791,
      "num_input_tokens_seen": 45181376,
      "step": 77825
    },
    {
      "epoch": 11.592195412570748,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.013452782780993709,
      "loss": 0.8014,
      "num_input_tokens_seen": 45184160,
      "step": 77830
    },
    {
      "epoch": 11.592940125111706,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.013450843540568571,
      "loss": 0.7805,
      "num_input_tokens_seen": 45187104,
      "step": 77835
    },
    {
      "epoch": 11.593684837652667,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.013448904326314865,
      "loss": 0.796,
      "num_input_tokens_seen": 45189984,
      "step": 77840
    },
    {
      "epoch": 11.594429550193626,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.013446965138265338,
      "loss": 0.7733,
      "num_input_tokens_seen": 45192832,
      "step": 77845
    },
    {
      "epoch": 11.595174262734584,
      "grad_norm": 0.029296875,
      "learning_rate": 0.013445025976452762,
      "loss": 0.7717,
      "num_input_tokens_seen": 45195680,
      "step": 77850
    },
    {
      "epoch": 11.595918975275543,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01344308684090989,
      "loss": 0.8062,
      "num_input_tokens_seen": 45198432,
      "step": 77855
    },
    {
      "epoch": 11.596663687816502,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.013441147731669477,
      "loss": 0.8216,
      "num_input_tokens_seen": 45201152,
      "step": 77860
    },
    {
      "epoch": 11.597408400357462,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.013439208648764297,
      "loss": 0.8002,
      "num_input_tokens_seen": 45204288,
      "step": 77865
    },
    {
      "epoch": 11.598153112898421,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.013437269592227092,
      "loss": 0.8368,
      "num_input_tokens_seen": 45207360,
      "step": 77870
    },
    {
      "epoch": 11.59889782543938,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.013435330562090637,
      "loss": 0.7866,
      "num_input_tokens_seen": 45210528,
      "step": 77875
    },
    {
      "epoch": 11.59964253798034,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.01343339155838768,
      "loss": 0.7894,
      "num_input_tokens_seen": 45213024,
      "step": 77880
    },
    {
      "epoch": 11.6003872505213,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.01343145258115098,
      "loss": 0.7985,
      "num_input_tokens_seen": 45215776,
      "step": 77885
    },
    {
      "epoch": 11.601131963062258,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.013429513630413297,
      "loss": 0.8237,
      "num_input_tokens_seen": 45218880,
      "step": 77890
    },
    {
      "epoch": 11.601876675603217,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.013427574706207377,
      "loss": 0.8283,
      "num_input_tokens_seen": 45222112,
      "step": 77895
    },
    {
      "epoch": 11.602621388144176,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.013425635808565992,
      "loss": 0.7754,
      "num_input_tokens_seen": 45225024,
      "step": 77900
    },
    {
      "epoch": 11.603366100685136,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.013423696937521884,
      "loss": 0.7953,
      "num_input_tokens_seen": 45227776,
      "step": 77905
    },
    {
      "epoch": 11.604110813226095,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.013421758093107819,
      "loss": 0.8326,
      "num_input_tokens_seen": 45230656,
      "step": 77910
    },
    {
      "epoch": 11.604855525767054,
      "grad_norm": 0.0390625,
      "learning_rate": 0.013419819275356543,
      "loss": 0.7831,
      "num_input_tokens_seen": 45233440,
      "step": 77915
    },
    {
      "epoch": 11.605600238308012,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.01341788048430082,
      "loss": 0.7827,
      "num_input_tokens_seen": 45236416,
      "step": 77920
    },
    {
      "epoch": 11.606344950848973,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.013415941719973389,
      "loss": 0.806,
      "num_input_tokens_seen": 45239424,
      "step": 77925
    },
    {
      "epoch": 11.607089663389932,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.013414002982407022,
      "loss": 0.8031,
      "num_input_tokens_seen": 45242336,
      "step": 77930
    },
    {
      "epoch": 11.60783437593089,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.013412064271634461,
      "loss": 0.8024,
      "num_input_tokens_seen": 45245824,
      "step": 77935
    },
    {
      "epoch": 11.60857908847185,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.013410125587688454,
      "loss": 0.7902,
      "num_input_tokens_seen": 45248864,
      "step": 77940
    },
    {
      "epoch": 11.60932380101281,
      "grad_norm": 0.02734375,
      "learning_rate": 0.013408186930601767,
      "loss": 0.7907,
      "num_input_tokens_seen": 45251584,
      "step": 77945
    },
    {
      "epoch": 11.610068513553768,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.01340624830040714,
      "loss": 0.7759,
      "num_input_tokens_seen": 45254624,
      "step": 77950
    },
    {
      "epoch": 11.610813226094727,
      "grad_norm": 0.046875,
      "learning_rate": 0.01340430969713733,
      "loss": 0.7925,
      "num_input_tokens_seen": 45257600,
      "step": 77955
    },
    {
      "epoch": 11.611557938635686,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.01340237112082508,
      "loss": 0.7965,
      "num_input_tokens_seen": 45260320,
      "step": 77960
    },
    {
      "epoch": 11.612302651176647,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.013400432571503152,
      "loss": 0.8293,
      "num_input_tokens_seen": 45263136,
      "step": 77965
    },
    {
      "epoch": 11.613047363717605,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.013398494049204294,
      "loss": 0.7929,
      "num_input_tokens_seen": 45266112,
      "step": 77970
    },
    {
      "epoch": 11.613792076258564,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01339655555396124,
      "loss": 0.7871,
      "num_input_tokens_seen": 45269248,
      "step": 77975
    },
    {
      "epoch": 11.614536788799523,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.013394617085806759,
      "loss": 0.7932,
      "num_input_tokens_seen": 45272224,
      "step": 77980
    },
    {
      "epoch": 11.615281501340483,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.013392678644773587,
      "loss": 0.7842,
      "num_input_tokens_seen": 45275040,
      "step": 77985
    },
    {
      "epoch": 11.616026213881442,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.013390740230894479,
      "loss": 0.7883,
      "num_input_tokens_seen": 45277920,
      "step": 77990
    },
    {
      "epoch": 11.6167709264224,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.013388801844202174,
      "loss": 0.7823,
      "num_input_tokens_seen": 45280960,
      "step": 77995
    },
    {
      "epoch": 11.61751563896336,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.013386863484729427,
      "loss": 0.8119,
      "num_input_tokens_seen": 45283936,
      "step": 78000
    },
    {
      "epoch": 11.61826035150432,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.013384925152508978,
      "loss": 0.8119,
      "num_input_tokens_seen": 45286752,
      "step": 78005
    },
    {
      "epoch": 11.619005064045279,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.013382986847573585,
      "loss": 0.7945,
      "num_input_tokens_seen": 45289856,
      "step": 78010
    },
    {
      "epoch": 11.619749776586238,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.013381048569955983,
      "loss": 0.7869,
      "num_input_tokens_seen": 45292704,
      "step": 78015
    },
    {
      "epoch": 11.620494489127196,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.013379110319688919,
      "loss": 0.8146,
      "num_input_tokens_seen": 45295616,
      "step": 78020
    },
    {
      "epoch": 11.621239201668157,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.013377172096805142,
      "loss": 0.8289,
      "num_input_tokens_seen": 45298560,
      "step": 78025
    },
    {
      "epoch": 11.621983914209116,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.013375233901337383,
      "loss": 0.809,
      "num_input_tokens_seen": 45301472,
      "step": 78030
    },
    {
      "epoch": 11.622728626750074,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.013373295733318406,
      "loss": 0.8003,
      "num_input_tokens_seen": 45304224,
      "step": 78035
    },
    {
      "epoch": 11.623473339291033,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.013371357592780938,
      "loss": 0.795,
      "num_input_tokens_seen": 45307296,
      "step": 78040
    },
    {
      "epoch": 11.624218051831992,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.013369419479757732,
      "loss": 0.798,
      "num_input_tokens_seen": 45310176,
      "step": 78045
    },
    {
      "epoch": 11.624962764372953,
      "grad_norm": 0.046875,
      "learning_rate": 0.013367481394281525,
      "loss": 0.8188,
      "num_input_tokens_seen": 45313280,
      "step": 78050
    },
    {
      "epoch": 11.625707476913911,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.013365543336385063,
      "loss": 0.7791,
      "num_input_tokens_seen": 45316224,
      "step": 78055
    },
    {
      "epoch": 11.62645218945487,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.013363605306101088,
      "loss": 0.8007,
      "num_input_tokens_seen": 45319040,
      "step": 78060
    },
    {
      "epoch": 11.627196901995829,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.013361667303462329,
      "loss": 0.8292,
      "num_input_tokens_seen": 45321696,
      "step": 78065
    },
    {
      "epoch": 11.62794161453679,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.013359729328501543,
      "loss": 0.817,
      "num_input_tokens_seen": 45324608,
      "step": 78070
    },
    {
      "epoch": 11.628686327077748,
      "grad_norm": 0.0703125,
      "learning_rate": 0.01335779138125146,
      "loss": 0.7798,
      "num_input_tokens_seen": 45327904,
      "step": 78075
    },
    {
      "epoch": 11.629431039618707,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.01335585346174482,
      "loss": 0.7879,
      "num_input_tokens_seen": 45330720,
      "step": 78080
    },
    {
      "epoch": 11.630175752159666,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.013353915570014368,
      "loss": 0.7866,
      "num_input_tokens_seen": 45333568,
      "step": 78085
    },
    {
      "epoch": 11.630920464700626,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.013351977706092841,
      "loss": 0.7888,
      "num_input_tokens_seen": 45336256,
      "step": 78090
    },
    {
      "epoch": 11.631665177241585,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.013350039870012976,
      "loss": 0.807,
      "num_input_tokens_seen": 45338976,
      "step": 78095
    },
    {
      "epoch": 11.632409889782544,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.013348102061807504,
      "loss": 0.7819,
      "num_input_tokens_seen": 45341728,
      "step": 78100
    },
    {
      "epoch": 11.633154602323502,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.013346164281509174,
      "loss": 0.7962,
      "num_input_tokens_seen": 45344704,
      "step": 78105
    },
    {
      "epoch": 11.633899314864463,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.013344226529150719,
      "loss": 0.8053,
      "num_input_tokens_seen": 45347648,
      "step": 78110
    },
    {
      "epoch": 11.634644027405422,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.013342288804764872,
      "loss": 0.8104,
      "num_input_tokens_seen": 45350688,
      "step": 78115
    },
    {
      "epoch": 11.63538873994638,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.013340351108384366,
      "loss": 0.8163,
      "num_input_tokens_seen": 45353632,
      "step": 78120
    },
    {
      "epoch": 11.63613345248734,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.013338413440041948,
      "loss": 0.8333,
      "num_input_tokens_seen": 45356320,
      "step": 78125
    },
    {
      "epoch": 11.6368781650283,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.013336475799770339,
      "loss": 0.7926,
      "num_input_tokens_seen": 45358880,
      "step": 78130
    },
    {
      "epoch": 11.637622877569259,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.013334538187602288,
      "loss": 0.7877,
      "num_input_tokens_seen": 45361760,
      "step": 78135
    },
    {
      "epoch": 11.638367590110217,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.013332600603570522,
      "loss": 0.7779,
      "num_input_tokens_seen": 45364544,
      "step": 78140
    },
    {
      "epoch": 11.639112302651176,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.01333066304770777,
      "loss": 0.7942,
      "num_input_tokens_seen": 45367616,
      "step": 78145
    },
    {
      "epoch": 11.639857015192137,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.013328725520046773,
      "loss": 0.8097,
      "num_input_tokens_seen": 45370400,
      "step": 78150
    },
    {
      "epoch": 11.640601727733095,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.013326788020620253,
      "loss": 0.7815,
      "num_input_tokens_seen": 45373152,
      "step": 78155
    },
    {
      "epoch": 11.641346440274054,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.013324850549460955,
      "loss": 0.7966,
      "num_input_tokens_seen": 45376064,
      "step": 78160
    },
    {
      "epoch": 11.642091152815013,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.013322913106601598,
      "loss": 0.8283,
      "num_input_tokens_seen": 45379072,
      "step": 78165
    },
    {
      "epoch": 11.642835865355973,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.013320975692074926,
      "loss": 0.7795,
      "num_input_tokens_seen": 45382016,
      "step": 78170
    },
    {
      "epoch": 11.643580577896932,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.013319038305913663,
      "loss": 0.8195,
      "num_input_tokens_seen": 45384960,
      "step": 78175
    },
    {
      "epoch": 11.64432529043789,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.013317100948150537,
      "loss": 0.7956,
      "num_input_tokens_seen": 45387744,
      "step": 78180
    },
    {
      "epoch": 11.64507000297885,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.013315163618818283,
      "loss": 0.7769,
      "num_input_tokens_seen": 45390496,
      "step": 78185
    },
    {
      "epoch": 11.64581471551981,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.013313226317949619,
      "loss": 0.827,
      "num_input_tokens_seen": 45393568,
      "step": 78190
    },
    {
      "epoch": 11.646559428060769,
      "grad_norm": 0.0390625,
      "learning_rate": 0.01331128904557729,
      "loss": 0.7874,
      "num_input_tokens_seen": 45396480,
      "step": 78195
    },
    {
      "epoch": 11.647304140601728,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01330935180173401,
      "loss": 0.7985,
      "num_input_tokens_seen": 45399360,
      "step": 78200
    },
    {
      "epoch": 11.648048853142686,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.01330741458645252,
      "loss": 0.8044,
      "num_input_tokens_seen": 45402464,
      "step": 78205
    },
    {
      "epoch": 11.648793565683647,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.013305477399765536,
      "loss": 0.7936,
      "num_input_tokens_seen": 45405472,
      "step": 78210
    },
    {
      "epoch": 11.649538278224606,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.013303540241705792,
      "loss": 0.7898,
      "num_input_tokens_seen": 45408416,
      "step": 78215
    },
    {
      "epoch": 11.650282990765565,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.013301603112306014,
      "loss": 0.8021,
      "num_input_tokens_seen": 45411328,
      "step": 78220
    },
    {
      "epoch": 11.651027703306523,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.013299666011598914,
      "loss": 0.7808,
      "num_input_tokens_seen": 45414240,
      "step": 78225
    },
    {
      "epoch": 11.651772415847482,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01329772893961724,
      "loss": 0.8229,
      "num_input_tokens_seen": 45417568,
      "step": 78230
    },
    {
      "epoch": 11.652517128388443,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.013295791896393697,
      "loss": 0.7963,
      "num_input_tokens_seen": 45420384,
      "step": 78235
    },
    {
      "epoch": 11.653261840929401,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.013293854881961024,
      "loss": 0.819,
      "num_input_tokens_seen": 45423328,
      "step": 78240
    },
    {
      "epoch": 11.65400655347036,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.013291917896351936,
      "loss": 0.7957,
      "num_input_tokens_seen": 45426336,
      "step": 78245
    },
    {
      "epoch": 11.654751266011319,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.01328998093959916,
      "loss": 0.8036,
      "num_input_tokens_seen": 45429216,
      "step": 78250
    },
    {
      "epoch": 11.65549597855228,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01328804401173542,
      "loss": 0.7823,
      "num_input_tokens_seen": 45432096,
      "step": 78255
    },
    {
      "epoch": 11.656240691093238,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.013286107112793431,
      "loss": 0.7936,
      "num_input_tokens_seen": 45435072,
      "step": 78260
    },
    {
      "epoch": 11.656985403634197,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.013284170242805926,
      "loss": 0.8139,
      "num_input_tokens_seen": 45437632,
      "step": 78265
    },
    {
      "epoch": 11.657730116175156,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.013282233401805616,
      "loss": 0.7875,
      "num_input_tokens_seen": 45440736,
      "step": 78270
    },
    {
      "epoch": 11.658474828716116,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.01328029658982523,
      "loss": 0.7999,
      "num_input_tokens_seen": 45443744,
      "step": 78275
    },
    {
      "epoch": 11.659219541257075,
      "grad_norm": 0.041015625,
      "learning_rate": 0.013278359806897485,
      "loss": 0.8032,
      "num_input_tokens_seen": 45446560,
      "step": 78280
    },
    {
      "epoch": 11.659964253798034,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.013276423053055103,
      "loss": 0.7938,
      "num_input_tokens_seen": 45449344,
      "step": 78285
    },
    {
      "epoch": 11.660708966338992,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.013274486328330795,
      "loss": 0.804,
      "num_input_tokens_seen": 45452544,
      "step": 78290
    },
    {
      "epoch": 11.661453678879953,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.013272549632757294,
      "loss": 0.7858,
      "num_input_tokens_seen": 45455328,
      "step": 78295
    },
    {
      "epoch": 11.662198391420912,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.01327061296636731,
      "loss": 0.7978,
      "num_input_tokens_seen": 45458752,
      "step": 78300
    },
    {
      "epoch": 11.66294310396187,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.013268676329193556,
      "loss": 0.7922,
      "num_input_tokens_seen": 45461440,
      "step": 78305
    },
    {
      "epoch": 11.66368781650283,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.013266739721268765,
      "loss": 0.8141,
      "num_input_tokens_seen": 45464448,
      "step": 78310
    },
    {
      "epoch": 11.66443252904379,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.013264803142625639,
      "loss": 0.7908,
      "num_input_tokens_seen": 45467072,
      "step": 78315
    },
    {
      "epoch": 11.665177241584749,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.013262866593296903,
      "loss": 0.7831,
      "num_input_tokens_seen": 45469984,
      "step": 78320
    },
    {
      "epoch": 11.665921954125707,
      "grad_norm": 0.048828125,
      "learning_rate": 0.013260930073315266,
      "loss": 0.8294,
      "num_input_tokens_seen": 45472800,
      "step": 78325
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 0.04296875,
      "learning_rate": 0.013258993582713452,
      "loss": 0.8087,
      "num_input_tokens_seen": 45475584,
      "step": 78330
    },
    {
      "epoch": 11.667411379207627,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.013257057121524174,
      "loss": 0.7959,
      "num_input_tokens_seen": 45478656,
      "step": 78335
    },
    {
      "epoch": 11.668156091748585,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.013255120689780136,
      "loss": 0.7948,
      "num_input_tokens_seen": 45481664,
      "step": 78340
    },
    {
      "epoch": 11.668900804289544,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01325318428751407,
      "loss": 0.8027,
      "num_input_tokens_seen": 45484864,
      "step": 78345
    },
    {
      "epoch": 11.669645516830503,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.013251247914758674,
      "loss": 0.8324,
      "num_input_tokens_seen": 45487776,
      "step": 78350
    },
    {
      "epoch": 11.670390229371463,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.013249311571546673,
      "loss": 0.7887,
      "num_input_tokens_seen": 45490624,
      "step": 78355
    },
    {
      "epoch": 11.671134941912422,
      "grad_norm": 0.046875,
      "learning_rate": 0.013247375257910764,
      "loss": 0.7869,
      "num_input_tokens_seen": 45493312,
      "step": 78360
    },
    {
      "epoch": 11.671879654453381,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.013245438973883677,
      "loss": 0.7916,
      "num_input_tokens_seen": 45496640,
      "step": 78365
    },
    {
      "epoch": 11.67262436699434,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.01324350271949811,
      "loss": 0.8086,
      "num_input_tokens_seen": 45499648,
      "step": 78370
    },
    {
      "epoch": 11.673369079535298,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.013241566494786784,
      "loss": 0.7908,
      "num_input_tokens_seen": 45502400,
      "step": 78375
    },
    {
      "epoch": 11.674113792076259,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.013239630299782406,
      "loss": 0.8086,
      "num_input_tokens_seen": 45505216,
      "step": 78380
    },
    {
      "epoch": 11.674858504617218,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.01323769413451768,
      "loss": 0.8047,
      "num_input_tokens_seen": 45508160,
      "step": 78385
    },
    {
      "epoch": 11.675603217158177,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.013235757999025327,
      "loss": 0.79,
      "num_input_tokens_seen": 45510912,
      "step": 78390
    },
    {
      "epoch": 11.676347929699137,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.013233821893338039,
      "loss": 0.7972,
      "num_input_tokens_seen": 45513696,
      "step": 78395
    },
    {
      "epoch": 11.677092642240096,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.013231885817488543,
      "loss": 0.7569,
      "num_input_tokens_seen": 45516480,
      "step": 78400
    },
    {
      "epoch": 11.677837354781055,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.013229949771509538,
      "loss": 0.795,
      "num_input_tokens_seen": 45519168,
      "step": 78405
    },
    {
      "epoch": 11.678582067322013,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.01322801375543373,
      "loss": 0.8064,
      "num_input_tokens_seen": 45522080,
      "step": 78410
    },
    {
      "epoch": 11.679326779862972,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.013226077769293833,
      "loss": 0.7904,
      "num_input_tokens_seen": 45524768,
      "step": 78415
    },
    {
      "epoch": 11.680071492403933,
      "grad_norm": 0.041015625,
      "learning_rate": 0.013224141813122545,
      "loss": 0.803,
      "num_input_tokens_seen": 45527456,
      "step": 78420
    },
    {
      "epoch": 11.680816204944891,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01322220588695258,
      "loss": 0.8036,
      "num_input_tokens_seen": 45530368,
      "step": 78425
    },
    {
      "epoch": 11.68156091748585,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.013220269990816633,
      "loss": 0.7714,
      "num_input_tokens_seen": 45533536,
      "step": 78430
    },
    {
      "epoch": 11.682305630026809,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.013218334124747422,
      "loss": 0.7926,
      "num_input_tokens_seen": 45536224,
      "step": 78435
    },
    {
      "epoch": 11.68305034256777,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.013216398288777643,
      "loss": 0.7889,
      "num_input_tokens_seen": 45539168,
      "step": 78440
    },
    {
      "epoch": 11.683795055108728,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.013214462482940004,
      "loss": 0.7908,
      "num_input_tokens_seen": 45542048,
      "step": 78445
    },
    {
      "epoch": 11.684539767649687,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0132125267072672,
      "loss": 0.7928,
      "num_input_tokens_seen": 45545056,
      "step": 78450
    },
    {
      "epoch": 11.685284480190646,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.01321059096179195,
      "loss": 0.7984,
      "num_input_tokens_seen": 45547712,
      "step": 78455
    },
    {
      "epoch": 11.686029192731606,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.013208655246546944,
      "loss": 0.8108,
      "num_input_tokens_seen": 45550432,
      "step": 78460
    },
    {
      "epoch": 11.686773905272565,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.013206719561564882,
      "loss": 0.8152,
      "num_input_tokens_seen": 45553568,
      "step": 78465
    },
    {
      "epoch": 11.687518617813524,
      "grad_norm": 0.05078125,
      "learning_rate": 0.013204783906878476,
      "loss": 0.846,
      "num_input_tokens_seen": 45556288,
      "step": 78470
    },
    {
      "epoch": 11.688263330354483,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.013202848282520418,
      "loss": 0.7921,
      "num_input_tokens_seen": 45559328,
      "step": 78475
    },
    {
      "epoch": 11.689008042895443,
      "grad_norm": 0.041015625,
      "learning_rate": 0.013200912688523416,
      "loss": 0.7912,
      "num_input_tokens_seen": 45561984,
      "step": 78480
    },
    {
      "epoch": 11.689752755436402,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.013198977124920158,
      "loss": 0.7915,
      "num_input_tokens_seen": 45565088,
      "step": 78485
    },
    {
      "epoch": 11.69049746797736,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.013197041591743357,
      "loss": 0.7743,
      "num_input_tokens_seen": 45568096,
      "step": 78490
    },
    {
      "epoch": 11.69124218051832,
      "grad_norm": 0.02734375,
      "learning_rate": 0.013195106089025709,
      "loss": 0.8052,
      "num_input_tokens_seen": 45570880,
      "step": 78495
    },
    {
      "epoch": 11.69198689305928,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.013193170616799898,
      "loss": 0.7988,
      "num_input_tokens_seen": 45573824,
      "step": 78500
    },
    {
      "epoch": 11.692731605600239,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.013191235175098645,
      "loss": 0.7914,
      "num_input_tokens_seen": 45576480,
      "step": 78505
    },
    {
      "epoch": 11.693476318141197,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.013189299763954629,
      "loss": 0.7821,
      "num_input_tokens_seen": 45579680,
      "step": 78510
    },
    {
      "epoch": 11.694221030682156,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.013187364383400556,
      "loss": 0.8113,
      "num_input_tokens_seen": 45582592,
      "step": 78515
    },
    {
      "epoch": 11.694965743223117,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.013185429033469113,
      "loss": 0.8063,
      "num_input_tokens_seen": 45585472,
      "step": 78520
    },
    {
      "epoch": 11.695710455764075,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.013183493714193008,
      "loss": 0.8033,
      "num_input_tokens_seen": 45588512,
      "step": 78525
    },
    {
      "epoch": 11.696455168305034,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.013181558425604926,
      "loss": 0.8239,
      "num_input_tokens_seen": 45591488,
      "step": 78530
    },
    {
      "epoch": 11.697199880845993,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.013179623167737572,
      "loss": 0.8234,
      "num_input_tokens_seen": 45594368,
      "step": 78535
    },
    {
      "epoch": 11.697944593386953,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.013177687940623635,
      "loss": 0.7902,
      "num_input_tokens_seen": 45596992,
      "step": 78540
    },
    {
      "epoch": 11.698689305927912,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.013175752744295803,
      "loss": 0.7906,
      "num_input_tokens_seen": 45599616,
      "step": 78545
    },
    {
      "epoch": 11.699434018468871,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01317381757878678,
      "loss": 0.7914,
      "num_input_tokens_seen": 45602464,
      "step": 78550
    },
    {
      "epoch": 11.70017873100983,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.013171882444129245,
      "loss": 0.7839,
      "num_input_tokens_seen": 45605696,
      "step": 78555
    },
    {
      "epoch": 11.700923443550789,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.013169947340355904,
      "loss": 0.8321,
      "num_input_tokens_seen": 45608480,
      "step": 78560
    },
    {
      "epoch": 11.701668156091749,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.013168012267499437,
      "loss": 0.8425,
      "num_input_tokens_seen": 45611232,
      "step": 78565
    },
    {
      "epoch": 11.702412868632708,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.013166077225592548,
      "loss": 0.8005,
      "num_input_tokens_seen": 45614208,
      "step": 78570
    },
    {
      "epoch": 11.703157581173667,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01316414221466792,
      "loss": 0.7885,
      "num_input_tokens_seen": 45617280,
      "step": 78575
    },
    {
      "epoch": 11.703902293714627,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.01316220723475824,
      "loss": 0.8037,
      "num_input_tokens_seen": 45620384,
      "step": 78580
    },
    {
      "epoch": 11.704647006255586,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.013160272285896205,
      "loss": 0.7935,
      "num_input_tokens_seen": 45623648,
      "step": 78585
    },
    {
      "epoch": 11.705391718796545,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.013158337368114491,
      "loss": 0.836,
      "num_input_tokens_seen": 45626720,
      "step": 78590
    },
    {
      "epoch": 11.706136431337503,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.013156402481445804,
      "loss": 0.8049,
      "num_input_tokens_seen": 45629728,
      "step": 78595
    },
    {
      "epoch": 11.706881143878462,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.013154467625922815,
      "loss": 0.7725,
      "num_input_tokens_seen": 45632256,
      "step": 78600
    },
    {
      "epoch": 11.707625856419423,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.013152532801578229,
      "loss": 0.7932,
      "num_input_tokens_seen": 45635136,
      "step": 78605
    },
    {
      "epoch": 11.708370568960381,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.013150598008444718,
      "loss": 0.7976,
      "num_input_tokens_seen": 45637888,
      "step": 78610
    },
    {
      "epoch": 11.70911528150134,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.013148663246554977,
      "loss": 0.7994,
      "num_input_tokens_seen": 45640928,
      "step": 78615
    },
    {
      "epoch": 11.709859994042299,
      "grad_norm": 0.033203125,
      "learning_rate": 0.013146728515941692,
      "loss": 0.7986,
      "num_input_tokens_seen": 45643904,
      "step": 78620
    },
    {
      "epoch": 11.71060470658326,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.013144793816637535,
      "loss": 0.828,
      "num_input_tokens_seen": 45646976,
      "step": 78625
    },
    {
      "epoch": 11.711349419124218,
      "grad_norm": 0.025390625,
      "learning_rate": 0.013142859148675209,
      "loss": 0.7844,
      "num_input_tokens_seen": 45649760,
      "step": 78630
    },
    {
      "epoch": 11.712094131665177,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.013140924512087383,
      "loss": 0.7906,
      "num_input_tokens_seen": 45652704,
      "step": 78635
    },
    {
      "epoch": 11.712838844206136,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.013138989906906757,
      "loss": 0.7874,
      "num_input_tokens_seen": 45655520,
      "step": 78640
    },
    {
      "epoch": 11.713583556747096,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.013137055333165998,
      "loss": 0.7833,
      "num_input_tokens_seen": 45658432,
      "step": 78645
    },
    {
      "epoch": 11.714328269288055,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.013135120790897802,
      "loss": 0.806,
      "num_input_tokens_seen": 45661216,
      "step": 78650
    },
    {
      "epoch": 11.715072981829014,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.013133186280134835,
      "loss": 0.8185,
      "num_input_tokens_seen": 45663904,
      "step": 78655
    },
    {
      "epoch": 11.715817694369973,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.013131251800909799,
      "loss": 0.7925,
      "num_input_tokens_seen": 45667008,
      "step": 78660
    },
    {
      "epoch": 11.716562406910933,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.013129317353255362,
      "loss": 0.8125,
      "num_input_tokens_seen": 45670080,
      "step": 78665
    },
    {
      "epoch": 11.717307119451892,
      "grad_norm": 0.03125,
      "learning_rate": 0.013127382937204203,
      "loss": 0.7913,
      "num_input_tokens_seen": 45672992,
      "step": 78670
    },
    {
      "epoch": 11.71805183199285,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.01312544855278901,
      "loss": 0.8175,
      "num_input_tokens_seen": 45675936,
      "step": 78675
    },
    {
      "epoch": 11.71879654453381,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.013123514200042454,
      "loss": 0.8035,
      "num_input_tokens_seen": 45678880,
      "step": 78680
    },
    {
      "epoch": 11.71954125707477,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.013121579878997225,
      "loss": 0.7838,
      "num_input_tokens_seen": 45681696,
      "step": 78685
    },
    {
      "epoch": 11.720285969615729,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.013119645589685987,
      "loss": 0.7903,
      "num_input_tokens_seen": 45684256,
      "step": 78690
    },
    {
      "epoch": 11.721030682156687,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.013117711332141431,
      "loss": 0.792,
      "num_input_tokens_seen": 45687232,
      "step": 78695
    },
    {
      "epoch": 11.721775394697646,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.01311577710639623,
      "loss": 0.805,
      "num_input_tokens_seen": 45690496,
      "step": 78700
    },
    {
      "epoch": 11.722520107238607,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.013113842912483055,
      "loss": 0.7878,
      "num_input_tokens_seen": 45693280,
      "step": 78705
    },
    {
      "epoch": 11.723264819779565,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.01311190875043459,
      "loss": 0.7949,
      "num_input_tokens_seen": 45696064,
      "step": 78710
    },
    {
      "epoch": 11.724009532320524,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.013109974620283509,
      "loss": 0.807,
      "num_input_tokens_seen": 45698816,
      "step": 78715
    },
    {
      "epoch": 11.724754244861483,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.013108040522062485,
      "loss": 0.7805,
      "num_input_tokens_seen": 45701696,
      "step": 78720
    },
    {
      "epoch": 11.725498957402444,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.013106106455804189,
      "loss": 0.8074,
      "num_input_tokens_seen": 45704512,
      "step": 78725
    },
    {
      "epoch": 11.726243669943402,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.013104172421541304,
      "loss": 0.8121,
      "num_input_tokens_seen": 45707328,
      "step": 78730
    },
    {
      "epoch": 11.726988382484361,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.013102238419306498,
      "loss": 0.8008,
      "num_input_tokens_seen": 45710080,
      "step": 78735
    },
    {
      "epoch": 11.72773309502532,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.013100304449132443,
      "loss": 0.8126,
      "num_input_tokens_seen": 45712896,
      "step": 78740
    },
    {
      "epoch": 11.728477807566279,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.013098370511051819,
      "loss": 0.7886,
      "num_input_tokens_seen": 45715616,
      "step": 78745
    },
    {
      "epoch": 11.729222520107239,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01309643660509729,
      "loss": 0.786,
      "num_input_tokens_seen": 45718848,
      "step": 78750
    },
    {
      "epoch": 11.729967232648198,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.013094502731301529,
      "loss": 0.797,
      "num_input_tokens_seen": 45721888,
      "step": 78755
    },
    {
      "epoch": 11.730711945189157,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.013092568889697205,
      "loss": 0.8259,
      "num_input_tokens_seen": 45724832,
      "step": 78760
    },
    {
      "epoch": 11.731456657730115,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.013090635080316996,
      "loss": 0.7761,
      "num_input_tokens_seen": 45728000,
      "step": 78765
    },
    {
      "epoch": 11.732201370271076,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.013088701303193565,
      "loss": 0.7747,
      "num_input_tokens_seen": 45730784,
      "step": 78770
    },
    {
      "epoch": 11.732946082812035,
      "grad_norm": 0.0390625,
      "learning_rate": 0.013086767558359583,
      "loss": 0.8021,
      "num_input_tokens_seen": 45733568,
      "step": 78775
    },
    {
      "epoch": 11.733690795352993,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.013084833845847717,
      "loss": 0.8094,
      "num_input_tokens_seen": 45736416,
      "step": 78780
    },
    {
      "epoch": 11.734435507893952,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.013082900165690638,
      "loss": 0.8057,
      "num_input_tokens_seen": 45739104,
      "step": 78785
    },
    {
      "epoch": 11.735180220434913,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.013080966517921013,
      "loss": 0.7924,
      "num_input_tokens_seen": 45741984,
      "step": 78790
    },
    {
      "epoch": 11.735924932975871,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.013079032902571503,
      "loss": 0.7947,
      "num_input_tokens_seen": 45744608,
      "step": 78795
    },
    {
      "epoch": 11.73666964551683,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.013077099319674788,
      "loss": 0.7964,
      "num_input_tokens_seen": 45747328,
      "step": 78800
    },
    {
      "epoch": 11.737414358057789,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.013075165769263519,
      "loss": 0.7841,
      "num_input_tokens_seen": 45750400,
      "step": 78805
    },
    {
      "epoch": 11.73815907059875,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.013073232251370375,
      "loss": 0.7742,
      "num_input_tokens_seen": 45753152,
      "step": 78810
    },
    {
      "epoch": 11.738903783139708,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.013071298766028005,
      "loss": 0.8226,
      "num_input_tokens_seen": 45756320,
      "step": 78815
    },
    {
      "epoch": 11.739648495680667,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.013069365313269087,
      "loss": 0.8079,
      "num_input_tokens_seen": 45759488,
      "step": 78820
    },
    {
      "epoch": 11.740393208221626,
      "grad_norm": 0.05078125,
      "learning_rate": 0.013067431893126284,
      "loss": 0.8016,
      "num_input_tokens_seen": 45762208,
      "step": 78825
    },
    {
      "epoch": 11.741137920762586,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.013065498505632244,
      "loss": 0.7841,
      "num_input_tokens_seen": 45764896,
      "step": 78830
    },
    {
      "epoch": 11.741882633303545,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.013063565150819651,
      "loss": 0.7998,
      "num_input_tokens_seen": 45767712,
      "step": 78835
    },
    {
      "epoch": 11.742627345844504,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.013061631828721151,
      "loss": 0.7917,
      "num_input_tokens_seen": 45770880,
      "step": 78840
    },
    {
      "epoch": 11.743372058385463,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.013059698539369415,
      "loss": 0.7941,
      "num_input_tokens_seen": 45773856,
      "step": 78845
    },
    {
      "epoch": 11.744116770926423,
      "grad_norm": 0.025390625,
      "learning_rate": 0.013057765282797095,
      "loss": 0.822,
      "num_input_tokens_seen": 45776576,
      "step": 78850
    },
    {
      "epoch": 11.744861483467382,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.01305583205903686,
      "loss": 0.7815,
      "num_input_tokens_seen": 45779392,
      "step": 78855
    },
    {
      "epoch": 11.74560619600834,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.01305389886812137,
      "loss": 0.7873,
      "num_input_tokens_seen": 45782368,
      "step": 78860
    },
    {
      "epoch": 11.7463509085493,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.013051965710083272,
      "loss": 0.802,
      "num_input_tokens_seen": 45785056,
      "step": 78865
    },
    {
      "epoch": 11.74709562109026,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.01305003258495524,
      "loss": 0.7848,
      "num_input_tokens_seen": 45787904,
      "step": 78870
    },
    {
      "epoch": 11.747840333631219,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.013048099492769924,
      "loss": 0.8048,
      "num_input_tokens_seen": 45790624,
      "step": 78875
    },
    {
      "epoch": 11.748585046172177,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.013046166433559986,
      "loss": 0.8027,
      "num_input_tokens_seen": 45793824,
      "step": 78880
    },
    {
      "epoch": 11.749329758713136,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.013044233407358071,
      "loss": 0.806,
      "num_input_tokens_seen": 45796736,
      "step": 78885
    },
    {
      "epoch": 11.750074471254095,
      "grad_norm": 0.17578125,
      "learning_rate": 0.013042300414196854,
      "loss": 0.7948,
      "num_input_tokens_seen": 45799712,
      "step": 78890
    },
    {
      "epoch": 11.750819183795056,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.013040367454108975,
      "loss": 0.7804,
      "num_input_tokens_seen": 45802528,
      "step": 78895
    },
    {
      "epoch": 11.751563896336014,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.013038434527127102,
      "loss": 0.8149,
      "num_input_tokens_seen": 45805632,
      "step": 78900
    },
    {
      "epoch": 11.752308608876973,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.013036501633283884,
      "loss": 0.7885,
      "num_input_tokens_seen": 45808256,
      "step": 78905
    },
    {
      "epoch": 11.753053321417934,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.013034568772611973,
      "loss": 0.7808,
      "num_input_tokens_seen": 45811168,
      "step": 78910
    },
    {
      "epoch": 11.753798033958892,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.013032635945144028,
      "loss": 0.7926,
      "num_input_tokens_seen": 45814048,
      "step": 78915
    },
    {
      "epoch": 11.754542746499851,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.013030703150912691,
      "loss": 0.7947,
      "num_input_tokens_seen": 45816960,
      "step": 78920
    },
    {
      "epoch": 11.75528745904081,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.013028770389950631,
      "loss": 0.7917,
      "num_input_tokens_seen": 45819680,
      "step": 78925
    },
    {
      "epoch": 11.756032171581769,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.013026837662290486,
      "loss": 0.8007,
      "num_input_tokens_seen": 45823008,
      "step": 78930
    },
    {
      "epoch": 11.75677688412273,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.013024904967964919,
      "loss": 0.7963,
      "num_input_tokens_seen": 45825952,
      "step": 78935
    },
    {
      "epoch": 11.757521596663688,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.013022972307006574,
      "loss": 0.7866,
      "num_input_tokens_seen": 45828672,
      "step": 78940
    },
    {
      "epoch": 11.758266309204647,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.013021039679448102,
      "loss": 0.8025,
      "num_input_tokens_seen": 45831808,
      "step": 78945
    },
    {
      "epoch": 11.759011021745605,
      "grad_norm": 0.041015625,
      "learning_rate": 0.013019107085322154,
      "loss": 0.8075,
      "num_input_tokens_seen": 45834656,
      "step": 78950
    },
    {
      "epoch": 11.759755734286566,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.013017174524661373,
      "loss": 0.8123,
      "num_input_tokens_seen": 45837376,
      "step": 78955
    },
    {
      "epoch": 11.760500446827525,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.01301524199749842,
      "loss": 0.7924,
      "num_input_tokens_seen": 45840416,
      "step": 78960
    },
    {
      "epoch": 11.761245159368483,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.013013309503865929,
      "loss": 0.7933,
      "num_input_tokens_seen": 45843424,
      "step": 78965
    },
    {
      "epoch": 11.761989871909442,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01301137704379656,
      "loss": 0.7901,
      "num_input_tokens_seen": 45846336,
      "step": 78970
    },
    {
      "epoch": 11.762734584450403,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.013009444617322953,
      "loss": 0.8134,
      "num_input_tokens_seen": 45849024,
      "step": 78975
    },
    {
      "epoch": 11.763479296991362,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.01300751222447776,
      "loss": 0.7986,
      "num_input_tokens_seen": 45851616,
      "step": 78980
    },
    {
      "epoch": 11.76422400953232,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.013005579865293621,
      "loss": 0.8081,
      "num_input_tokens_seen": 45854592,
      "step": 78985
    },
    {
      "epoch": 11.764968722073279,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.013003647539803178,
      "loss": 0.7937,
      "num_input_tokens_seen": 45857696,
      "step": 78990
    },
    {
      "epoch": 11.76571343461424,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.013001715248039086,
      "loss": 0.8076,
      "num_input_tokens_seen": 45860640,
      "step": 78995
    },
    {
      "epoch": 11.766458147155198,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.01299978299003398,
      "loss": 0.8009,
      "num_input_tokens_seen": 45863872,
      "step": 79000
    },
    {
      "epoch": 11.767202859696157,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.012997850765820511,
      "loss": 0.7638,
      "num_input_tokens_seen": 45866624,
      "step": 79005
    },
    {
      "epoch": 11.767947572237116,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.012995918575431316,
      "loss": 0.7603,
      "num_input_tokens_seen": 45869344,
      "step": 79010
    },
    {
      "epoch": 11.768692284778076,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.012993986418899043,
      "loss": 0.8413,
      "num_input_tokens_seen": 45872256,
      "step": 79015
    },
    {
      "epoch": 11.769436997319035,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.012992054296256331,
      "loss": 0.794,
      "num_input_tokens_seen": 45875008,
      "step": 79020
    },
    {
      "epoch": 11.770181709859994,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.012990122207535815,
      "loss": 0.8026,
      "num_input_tokens_seen": 45877760,
      "step": 79025
    },
    {
      "epoch": 11.770926422400953,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.012988190152770147,
      "loss": 0.8085,
      "num_input_tokens_seen": 45880672,
      "step": 79030
    },
    {
      "epoch": 11.771671134941913,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.012986258131991956,
      "loss": 0.8182,
      "num_input_tokens_seen": 45883648,
      "step": 79035
    },
    {
      "epoch": 11.772415847482872,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.012984326145233895,
      "loss": 0.801,
      "num_input_tokens_seen": 45886272,
      "step": 79040
    },
    {
      "epoch": 11.77316056002383,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.012982394192528592,
      "loss": 0.7782,
      "num_input_tokens_seen": 45889184,
      "step": 79045
    },
    {
      "epoch": 11.77390527256479,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.012980462273908687,
      "loss": 0.7861,
      "num_input_tokens_seen": 45892384,
      "step": 79050
    },
    {
      "epoch": 11.77464998510575,
      "grad_norm": 0.044921875,
      "learning_rate": 0.012978530389406816,
      "loss": 0.7906,
      "num_input_tokens_seen": 45895200,
      "step": 79055
    },
    {
      "epoch": 11.775394697646709,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.012976598539055627,
      "loss": 0.7978,
      "num_input_tokens_seen": 45898240,
      "step": 79060
    },
    {
      "epoch": 11.776139410187668,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.012974666722887749,
      "loss": 0.7993,
      "num_input_tokens_seen": 45900992,
      "step": 79065
    },
    {
      "epoch": 11.776884122728626,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.01297273494093581,
      "loss": 0.8013,
      "num_input_tokens_seen": 45904064,
      "step": 79070
    },
    {
      "epoch": 11.777628835269585,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01297080319323246,
      "loss": 0.801,
      "num_input_tokens_seen": 45906784,
      "step": 79075
    },
    {
      "epoch": 11.778373547810546,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.012968871479810327,
      "loss": 0.8035,
      "num_input_tokens_seen": 45909632,
      "step": 79080
    },
    {
      "epoch": 11.779118260351504,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.012966939800702047,
      "loss": 0.7891,
      "num_input_tokens_seen": 45912576,
      "step": 79085
    },
    {
      "epoch": 11.779862972892463,
      "grad_norm": 0.03515625,
      "learning_rate": 0.012965008155940247,
      "loss": 0.8194,
      "num_input_tokens_seen": 45915520,
      "step": 79090
    },
    {
      "epoch": 11.780607685433424,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.012963076545557574,
      "loss": 0.7826,
      "num_input_tokens_seen": 45918752,
      "step": 79095
    },
    {
      "epoch": 11.781352397974382,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01296114496958665,
      "loss": 0.7937,
      "num_input_tokens_seen": 45921600,
      "step": 79100
    },
    {
      "epoch": 11.782097110515341,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.012959213428060107,
      "loss": 0.8043,
      "num_input_tokens_seen": 45924416,
      "step": 79105
    },
    {
      "epoch": 11.7828418230563,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01295728192101058,
      "loss": 0.8131,
      "num_input_tokens_seen": 45927520,
      "step": 79110
    },
    {
      "epoch": 11.783586535597259,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.012955350448470697,
      "loss": 0.7635,
      "num_input_tokens_seen": 45930464,
      "step": 79115
    },
    {
      "epoch": 11.78433124813822,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.012953419010473095,
      "loss": 0.7949,
      "num_input_tokens_seen": 45933248,
      "step": 79120
    },
    {
      "epoch": 11.785075960679178,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.012951487607050392,
      "loss": 0.8095,
      "num_input_tokens_seen": 45936320,
      "step": 79125
    },
    {
      "epoch": 11.785820673220137,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.012949556238235231,
      "loss": 0.7861,
      "num_input_tokens_seen": 45938944,
      "step": 79130
    },
    {
      "epoch": 11.786565385761095,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.012947624904060229,
      "loss": 0.7949,
      "num_input_tokens_seen": 45941824,
      "step": 79135
    },
    {
      "epoch": 11.787310098302056,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.012945693604558021,
      "loss": 0.7883,
      "num_input_tokens_seen": 45944736,
      "step": 79140
    },
    {
      "epoch": 11.788054810843015,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.012943762339761229,
      "loss": 0.7977,
      "num_input_tokens_seen": 45947520,
      "step": 79145
    },
    {
      "epoch": 11.788799523383974,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.012941831109702482,
      "loss": 0.7892,
      "num_input_tokens_seen": 45950208,
      "step": 79150
    },
    {
      "epoch": 11.789544235924932,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.01293989991441441,
      "loss": 0.7992,
      "num_input_tokens_seen": 45953120,
      "step": 79155
    },
    {
      "epoch": 11.790288948465893,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.01293796875392963,
      "loss": 0.7965,
      "num_input_tokens_seen": 45956032,
      "step": 79160
    },
    {
      "epoch": 11.791033661006852,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.012936037628280779,
      "loss": 0.7857,
      "num_input_tokens_seen": 45959136,
      "step": 79165
    },
    {
      "epoch": 11.79177837354781,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.012934106537500469,
      "loss": 0.7965,
      "num_input_tokens_seen": 45962080,
      "step": 79170
    },
    {
      "epoch": 11.792523086088769,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.012932175481621335,
      "loss": 0.8251,
      "num_input_tokens_seen": 45965088,
      "step": 79175
    },
    {
      "epoch": 11.79326779862973,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.012930244460675993,
      "loss": 0.8031,
      "num_input_tokens_seen": 45967936,
      "step": 79180
    },
    {
      "epoch": 11.794012511170688,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01292831347469706,
      "loss": 0.7916,
      "num_input_tokens_seen": 45971072,
      "step": 79185
    },
    {
      "epoch": 11.794757223711647,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.012926382523717174,
      "loss": 0.8167,
      "num_input_tokens_seen": 45974176,
      "step": 79190
    },
    {
      "epoch": 11.795501936252606,
      "grad_norm": 0.03125,
      "learning_rate": 0.01292445160776894,
      "loss": 0.8145,
      "num_input_tokens_seen": 45976832,
      "step": 79195
    },
    {
      "epoch": 11.796246648793566,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.012922520726884991,
      "loss": 0.804,
      "num_input_tokens_seen": 45979616,
      "step": 79200
    },
    {
      "epoch": 11.796991361334525,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.012920589881097943,
      "loss": 0.8049,
      "num_input_tokens_seen": 45982656,
      "step": 79205
    },
    {
      "epoch": 11.797736073875484,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.012918659070440416,
      "loss": 0.8097,
      "num_input_tokens_seen": 45985568,
      "step": 79210
    },
    {
      "epoch": 11.798480786416443,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.012916728294945024,
      "loss": 0.808,
      "num_input_tokens_seen": 45988544,
      "step": 79215
    },
    {
      "epoch": 11.799225498957403,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.012914797554644397,
      "loss": 0.8089,
      "num_input_tokens_seen": 45991232,
      "step": 79220
    },
    {
      "epoch": 11.799970211498362,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.012912866849571145,
      "loss": 0.7965,
      "num_input_tokens_seen": 45994400,
      "step": 79225
    },
    {
      "epoch": 11.80071492403932,
      "grad_norm": 0.019287109375,
      "learning_rate": 0.01291093617975788,
      "loss": 0.8442,
      "num_input_tokens_seen": 45997408,
      "step": 79230
    },
    {
      "epoch": 11.80145963658028,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.012909005545237232,
      "loss": 0.7869,
      "num_input_tokens_seen": 46000480,
      "step": 79235
    },
    {
      "epoch": 11.80220434912124,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.012907074946041807,
      "loss": 0.8018,
      "num_input_tokens_seen": 46003360,
      "step": 79240
    },
    {
      "epoch": 11.802949061662199,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.012905144382204227,
      "loss": 0.8012,
      "num_input_tokens_seen": 46006240,
      "step": 79245
    },
    {
      "epoch": 11.803693774203158,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.012903213853757093,
      "loss": 0.8261,
      "num_input_tokens_seen": 46009088,
      "step": 79250
    },
    {
      "epoch": 11.804438486744116,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.012901283360733042,
      "loss": 0.8005,
      "num_input_tokens_seen": 46012032,
      "step": 79255
    },
    {
      "epoch": 11.805183199285075,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.012899352903164666,
      "loss": 0.7842,
      "num_input_tokens_seen": 46015008,
      "step": 79260
    },
    {
      "epoch": 11.805927911826036,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.012897422481084594,
      "loss": 0.7751,
      "num_input_tokens_seen": 46017632,
      "step": 79265
    },
    {
      "epoch": 11.806672624366994,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.012895492094525432,
      "loss": 0.7903,
      "num_input_tokens_seen": 46020384,
      "step": 79270
    },
    {
      "epoch": 11.807417336907953,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.012893561743519792,
      "loss": 0.8181,
      "num_input_tokens_seen": 46023424,
      "step": 79275
    },
    {
      "epoch": 11.808162049448912,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.012891631428100287,
      "loss": 0.8107,
      "num_input_tokens_seen": 46026208,
      "step": 79280
    },
    {
      "epoch": 11.808906761989872,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.012889701148299518,
      "loss": 0.7928,
      "num_input_tokens_seen": 46028768,
      "step": 79285
    },
    {
      "epoch": 11.809651474530831,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.012887770904150112,
      "loss": 0.7919,
      "num_input_tokens_seen": 46031456,
      "step": 79290
    },
    {
      "epoch": 11.81039618707179,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.012885840695684663,
      "loss": 0.7984,
      "num_input_tokens_seen": 46034304,
      "step": 79295
    },
    {
      "epoch": 11.811140899612749,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.012883910522935794,
      "loss": 0.7952,
      "num_input_tokens_seen": 46037408,
      "step": 79300
    },
    {
      "epoch": 11.81188561215371,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.012881980385936104,
      "loss": 0.7915,
      "num_input_tokens_seen": 46040512,
      "step": 79305
    },
    {
      "epoch": 11.812630324694668,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.0128800502847182,
      "loss": 0.7993,
      "num_input_tokens_seen": 46043616,
      "step": 79310
    },
    {
      "epoch": 11.813375037235627,
      "grad_norm": 0.044921875,
      "learning_rate": 0.012878120219314695,
      "loss": 0.8197,
      "num_input_tokens_seen": 46046432,
      "step": 79315
    },
    {
      "epoch": 11.814119749776586,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.012876190189758186,
      "loss": 0.7905,
      "num_input_tokens_seen": 46049248,
      "step": 79320
    },
    {
      "epoch": 11.814864462317546,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.012874260196081292,
      "loss": 0.8001,
      "num_input_tokens_seen": 46052160,
      "step": 79325
    },
    {
      "epoch": 11.815609174858505,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.012872330238316606,
      "loss": 0.8151,
      "num_input_tokens_seen": 46055264,
      "step": 79330
    },
    {
      "epoch": 11.816353887399464,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.012870400316496743,
      "loss": 0.797,
      "num_input_tokens_seen": 46057952,
      "step": 79335
    },
    {
      "epoch": 11.817098599940422,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.012868470430654299,
      "loss": 0.8049,
      "num_input_tokens_seen": 46061056,
      "step": 79340
    },
    {
      "epoch": 11.817843312481383,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.012866540580821885,
      "loss": 0.8038,
      "num_input_tokens_seen": 46064000,
      "step": 79345
    },
    {
      "epoch": 11.818588025022342,
      "grad_norm": 0.04296875,
      "learning_rate": 0.012864610767032098,
      "loss": 0.8134,
      "num_input_tokens_seen": 46066816,
      "step": 79350
    },
    {
      "epoch": 11.8193327375633,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.012862680989317537,
      "loss": 0.7796,
      "num_input_tokens_seen": 46069728,
      "step": 79355
    },
    {
      "epoch": 11.82007745010426,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.012860751247710814,
      "loss": 0.8025,
      "num_input_tokens_seen": 46072288,
      "step": 79360
    },
    {
      "epoch": 11.82082216264522,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.012858821542244518,
      "loss": 0.7982,
      "num_input_tokens_seen": 46075232,
      "step": 79365
    },
    {
      "epoch": 11.821566875186178,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.01285689187295126,
      "loss": 0.797,
      "num_input_tokens_seen": 46078112,
      "step": 79370
    },
    {
      "epoch": 11.822311587727137,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.012854962239863637,
      "loss": 0.7929,
      "num_input_tokens_seen": 46081376,
      "step": 79375
    },
    {
      "epoch": 11.823056300268096,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.012853032643014248,
      "loss": 0.7738,
      "num_input_tokens_seen": 46084192,
      "step": 79380
    },
    {
      "epoch": 11.823801012809056,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.012851103082435691,
      "loss": 0.7977,
      "num_input_tokens_seen": 46086816,
      "step": 79385
    },
    {
      "epoch": 11.824545725350015,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.012849173558160554,
      "loss": 0.7974,
      "num_input_tokens_seen": 46089504,
      "step": 79390
    },
    {
      "epoch": 11.825290437890974,
      "grad_norm": 0.044921875,
      "learning_rate": 0.012847244070221454,
      "loss": 0.7949,
      "num_input_tokens_seen": 46092512,
      "step": 79395
    },
    {
      "epoch": 11.826035150431933,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.012845314618650969,
      "loss": 0.8048,
      "num_input_tokens_seen": 46095744,
      "step": 79400
    },
    {
      "epoch": 11.826779862972892,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01284338520348171,
      "loss": 0.8258,
      "num_input_tokens_seen": 46098720,
      "step": 79405
    },
    {
      "epoch": 11.827524575513852,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.012841455824746262,
      "loss": 0.7911,
      "num_input_tokens_seen": 46101536,
      "step": 79410
    },
    {
      "epoch": 11.82826928805481,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.01283952648247723,
      "loss": 0.7982,
      "num_input_tokens_seen": 46104608,
      "step": 79415
    },
    {
      "epoch": 11.82901400059577,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.012837597176707193,
      "loss": 0.7966,
      "num_input_tokens_seen": 46107584,
      "step": 79420
    },
    {
      "epoch": 11.82975871313673,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.012835667907468764,
      "loss": 0.7942,
      "num_input_tokens_seen": 46110592,
      "step": 79425
    },
    {
      "epoch": 11.830503425677689,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.012833738674794523,
      "loss": 0.8084,
      "num_input_tokens_seen": 46113600,
      "step": 79430
    },
    {
      "epoch": 11.831248138218648,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.012831809478717062,
      "loss": 0.795,
      "num_input_tokens_seen": 46116512,
      "step": 79435
    },
    {
      "epoch": 11.831992850759606,
      "grad_norm": 0.037109375,
      "learning_rate": 0.012829880319268978,
      "loss": 0.7734,
      "num_input_tokens_seen": 46119328,
      "step": 79440
    },
    {
      "epoch": 11.832737563300565,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.012827951196482858,
      "loss": 0.7854,
      "num_input_tokens_seen": 46122144,
      "step": 79445
    },
    {
      "epoch": 11.833482275841526,
      "grad_norm": 0.046875,
      "learning_rate": 0.0128260221103913,
      "loss": 0.8038,
      "num_input_tokens_seen": 46125216,
      "step": 79450
    },
    {
      "epoch": 11.834226988382484,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.012824093061026881,
      "loss": 0.8016,
      "num_input_tokens_seen": 46128032,
      "step": 79455
    },
    {
      "epoch": 11.834971700923443,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.012822164048422208,
      "loss": 0.7994,
      "num_input_tokens_seen": 46130816,
      "step": 79460
    },
    {
      "epoch": 11.835716413464402,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.012820235072609857,
      "loss": 0.806,
      "num_input_tokens_seen": 46133952,
      "step": 79465
    },
    {
      "epoch": 11.836461126005362,
      "grad_norm": 0.037109375,
      "learning_rate": 0.012818306133622417,
      "loss": 0.7993,
      "num_input_tokens_seen": 46136864,
      "step": 79470
    },
    {
      "epoch": 11.837205838546321,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.012816377231492477,
      "loss": 0.8009,
      "num_input_tokens_seen": 46139808,
      "step": 79475
    },
    {
      "epoch": 11.83795055108728,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.012814448366252625,
      "loss": 0.7913,
      "num_input_tokens_seen": 46142848,
      "step": 79480
    },
    {
      "epoch": 11.838695263628239,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.01281251953793545,
      "loss": 0.788,
      "num_input_tokens_seen": 46145344,
      "step": 79485
    },
    {
      "epoch": 11.8394399761692,
      "grad_norm": 0.048828125,
      "learning_rate": 0.012810590746573527,
      "loss": 0.7961,
      "num_input_tokens_seen": 46148480,
      "step": 79490
    },
    {
      "epoch": 11.840184688710158,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.012808661992199452,
      "loss": 0.7979,
      "num_input_tokens_seen": 46151136,
      "step": 79495
    },
    {
      "epoch": 11.840929401251117,
      "grad_norm": 0.03515625,
      "learning_rate": 0.012806733274845807,
      "loss": 0.8059,
      "num_input_tokens_seen": 46154080,
      "step": 79500
    },
    {
      "epoch": 11.841674113792076,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.012804804594545175,
      "loss": 0.7915,
      "num_input_tokens_seen": 46156928,
      "step": 79505
    },
    {
      "epoch": 11.842418826333036,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.012802875951330138,
      "loss": 0.8232,
      "num_input_tokens_seen": 46159904,
      "step": 79510
    },
    {
      "epoch": 11.843163538873995,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.012800947345233273,
      "loss": 0.7926,
      "num_input_tokens_seen": 46162528,
      "step": 79515
    },
    {
      "epoch": 11.843908251414954,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.012799018776287174,
      "loss": 0.8095,
      "num_input_tokens_seen": 46165408,
      "step": 79520
    },
    {
      "epoch": 11.844652963955912,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.012797090244524406,
      "loss": 0.7952,
      "num_input_tokens_seen": 46168384,
      "step": 79525
    },
    {
      "epoch": 11.845397676496873,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.01279516174997757,
      "loss": 0.808,
      "num_input_tokens_seen": 46171552,
      "step": 79530
    },
    {
      "epoch": 11.846142389037832,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.012793233292679231,
      "loss": 0.7935,
      "num_input_tokens_seen": 46174496,
      "step": 79535
    },
    {
      "epoch": 11.84688710157879,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.012791304872661977,
      "loss": 0.8387,
      "num_input_tokens_seen": 46177472,
      "step": 79540
    },
    {
      "epoch": 11.84763181411975,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.012789376489958381,
      "loss": 0.7871,
      "num_input_tokens_seen": 46180192,
      "step": 79545
    },
    {
      "epoch": 11.84837652666071,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.012787448144601015,
      "loss": 0.8085,
      "num_input_tokens_seen": 46183104,
      "step": 79550
    },
    {
      "epoch": 11.849121239201668,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.012785519836622473,
      "loss": 0.7943,
      "num_input_tokens_seen": 46185984,
      "step": 79555
    },
    {
      "epoch": 11.849865951742627,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.012783591566055316,
      "loss": 0.7849,
      "num_input_tokens_seen": 46189120,
      "step": 79560
    },
    {
      "epoch": 11.850610664283586,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01278166333293213,
      "loss": 0.8008,
      "num_input_tokens_seen": 46191936,
      "step": 79565
    },
    {
      "epoch": 11.851355376824547,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.012779735137285487,
      "loss": 0.7871,
      "num_input_tokens_seen": 46194592,
      "step": 79570
    },
    {
      "epoch": 11.852100089365505,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.012777806979147965,
      "loss": 0.7951,
      "num_input_tokens_seen": 46197376,
      "step": 79575
    },
    {
      "epoch": 11.852844801906464,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.012775878858552128,
      "loss": 0.8014,
      "num_input_tokens_seen": 46200384,
      "step": 79580
    },
    {
      "epoch": 11.853589514447423,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.012773950775530564,
      "loss": 0.7903,
      "num_input_tokens_seen": 46203136,
      "step": 79585
    },
    {
      "epoch": 11.854334226988382,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.012772022730115842,
      "loss": 0.8197,
      "num_input_tokens_seen": 46206016,
      "step": 79590
    },
    {
      "epoch": 11.855078939529342,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.012770094722340522,
      "loss": 0.789,
      "num_input_tokens_seen": 46208736,
      "step": 79595
    },
    {
      "epoch": 11.8558236520703,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.012768166752237192,
      "loss": 0.8009,
      "num_input_tokens_seen": 46211808,
      "step": 79600
    },
    {
      "epoch": 11.85656836461126,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.012766238819838413,
      "loss": 0.7925,
      "num_input_tokens_seen": 46214848,
      "step": 79605
    },
    {
      "epoch": 11.85731307715222,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.012764310925176763,
      "loss": 0.8012,
      "num_input_tokens_seen": 46217664,
      "step": 79610
    },
    {
      "epoch": 11.858057789693179,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0127623830682848,
      "loss": 0.815,
      "num_input_tokens_seen": 46220736,
      "step": 79615
    },
    {
      "epoch": 11.858802502234138,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.01276045524919511,
      "loss": 0.776,
      "num_input_tokens_seen": 46223648,
      "step": 79620
    },
    {
      "epoch": 11.859547214775096,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.01275852746794025,
      "loss": 0.7907,
      "num_input_tokens_seen": 46226560,
      "step": 79625
    },
    {
      "epoch": 11.860291927316055,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.012756599724552784,
      "loss": 0.8005,
      "num_input_tokens_seen": 46229568,
      "step": 79630
    },
    {
      "epoch": 11.861036639857016,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.012754672019065293,
      "loss": 0.8086,
      "num_input_tokens_seen": 46232416,
      "step": 79635
    },
    {
      "epoch": 11.861781352397974,
      "grad_norm": 0.0390625,
      "learning_rate": 0.012752744351510333,
      "loss": 0.8028,
      "num_input_tokens_seen": 46235392,
      "step": 79640
    },
    {
      "epoch": 11.862526064938933,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.012750816721920475,
      "loss": 0.8064,
      "num_input_tokens_seen": 46238400,
      "step": 79645
    },
    {
      "epoch": 11.863270777479892,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.012748889130328277,
      "loss": 0.7904,
      "num_input_tokens_seen": 46241408,
      "step": 79650
    },
    {
      "epoch": 11.864015490020853,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.012746961576766316,
      "loss": 0.7776,
      "num_input_tokens_seen": 46244320,
      "step": 79655
    },
    {
      "epoch": 11.864760202561811,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.012745034061267141,
      "loss": 0.8091,
      "num_input_tokens_seen": 46247008,
      "step": 79660
    },
    {
      "epoch": 11.86550491510277,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.012743106583863333,
      "loss": 0.792,
      "num_input_tokens_seen": 46250112,
      "step": 79665
    },
    {
      "epoch": 11.866249627643729,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.012741179144587443,
      "loss": 0.7864,
      "num_input_tokens_seen": 46253088,
      "step": 79670
    },
    {
      "epoch": 11.86699434018469,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.012739251743472034,
      "loss": 0.8022,
      "num_input_tokens_seen": 46256128,
      "step": 79675
    },
    {
      "epoch": 11.867739052725648,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.012737324380549671,
      "loss": 0.7987,
      "num_input_tokens_seen": 46258880,
      "step": 79680
    },
    {
      "epoch": 11.868483765266607,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.012735397055852908,
      "loss": 0.8083,
      "num_input_tokens_seen": 46262112,
      "step": 79685
    },
    {
      "epoch": 11.869228477807566,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.012733469769414314,
      "loss": 0.8083,
      "num_input_tokens_seen": 46265120,
      "step": 79690
    },
    {
      "epoch": 11.869973190348526,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01273154252126644,
      "loss": 0.8007,
      "num_input_tokens_seen": 46267712,
      "step": 79695
    },
    {
      "epoch": 11.870717902889485,
      "grad_norm": 0.04296875,
      "learning_rate": 0.012729615311441858,
      "loss": 0.7931,
      "num_input_tokens_seen": 46270560,
      "step": 79700
    },
    {
      "epoch": 11.871462615430444,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.012727688139973114,
      "loss": 0.7954,
      "num_input_tokens_seen": 46273632,
      "step": 79705
    },
    {
      "epoch": 11.872207327971402,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01272576100689277,
      "loss": 0.7732,
      "num_input_tokens_seen": 46276416,
      "step": 79710
    },
    {
      "epoch": 11.872952040512363,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.012723833912233383,
      "loss": 0.7955,
      "num_input_tokens_seen": 46279200,
      "step": 79715
    },
    {
      "epoch": 11.873696753053322,
      "grad_norm": 0.03515625,
      "learning_rate": 0.012721906856027502,
      "loss": 0.7909,
      "num_input_tokens_seen": 46282016,
      "step": 79720
    },
    {
      "epoch": 11.87444146559428,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.012719979838307695,
      "loss": 0.7973,
      "num_input_tokens_seen": 46285024,
      "step": 79725
    },
    {
      "epoch": 11.87518617813524,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.012718052859106509,
      "loss": 0.8004,
      "num_input_tokens_seen": 46287872,
      "step": 79730
    },
    {
      "epoch": 11.8759308906762,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.012716125918456504,
      "loss": 0.817,
      "num_input_tokens_seen": 46291040,
      "step": 79735
    },
    {
      "epoch": 11.876675603217159,
      "grad_norm": 0.04296875,
      "learning_rate": 0.012714199016390227,
      "loss": 0.7656,
      "num_input_tokens_seen": 46293920,
      "step": 79740
    },
    {
      "epoch": 11.877420315758117,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.012712272152940237,
      "loss": 0.8165,
      "num_input_tokens_seen": 46296768,
      "step": 79745
    },
    {
      "epoch": 11.878165028299076,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.012710345328139087,
      "loss": 0.7933,
      "num_input_tokens_seen": 46299648,
      "step": 79750
    },
    {
      "epoch": 11.878909740840037,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.012708418542019316,
      "loss": 0.8047,
      "num_input_tokens_seen": 46302592,
      "step": 79755
    },
    {
      "epoch": 11.879654453380995,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.012706491794613492,
      "loss": 0.8099,
      "num_input_tokens_seen": 46305728,
      "step": 79760
    },
    {
      "epoch": 11.880399165921954,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.012704565085954156,
      "loss": 0.7982,
      "num_input_tokens_seen": 46308544,
      "step": 79765
    },
    {
      "epoch": 11.881143878462913,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.012702638416073858,
      "loss": 0.7873,
      "num_input_tokens_seen": 46311520,
      "step": 79770
    },
    {
      "epoch": 11.881888591003872,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.012700711785005151,
      "loss": 0.7722,
      "num_input_tokens_seen": 46314336,
      "step": 79775
    },
    {
      "epoch": 11.882633303544832,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.012698785192780582,
      "loss": 0.8037,
      "num_input_tokens_seen": 46317344,
      "step": 79780
    },
    {
      "epoch": 11.883378016085791,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.012696858639432701,
      "loss": 0.788,
      "num_input_tokens_seen": 46320224,
      "step": 79785
    },
    {
      "epoch": 11.88412272862675,
      "grad_norm": 0.044921875,
      "learning_rate": 0.012694932124994045,
      "loss": 0.7788,
      "num_input_tokens_seen": 46323104,
      "step": 79790
    },
    {
      "epoch": 11.884867441167708,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.012693005649497175,
      "loss": 0.7825,
      "num_input_tokens_seen": 46326048,
      "step": 79795
    },
    {
      "epoch": 11.885612153708669,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.012691079212974627,
      "loss": 0.7919,
      "num_input_tokens_seen": 46329376,
      "step": 79800
    },
    {
      "epoch": 11.886356866249628,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.012689152815458947,
      "loss": 0.7897,
      "num_input_tokens_seen": 46332064,
      "step": 79805
    },
    {
      "epoch": 11.887101578790586,
      "grad_norm": 0.068359375,
      "learning_rate": 0.012687226456982684,
      "loss": 0.7821,
      "num_input_tokens_seen": 46334944,
      "step": 79810
    },
    {
      "epoch": 11.887846291331545,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.012685300137578383,
      "loss": 0.7882,
      "num_input_tokens_seen": 46338016,
      "step": 79815
    },
    {
      "epoch": 11.888591003872506,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.012683373857278574,
      "loss": 0.8223,
      "num_input_tokens_seen": 46340960,
      "step": 79820
    },
    {
      "epoch": 11.889335716413465,
      "grad_norm": 0.04296875,
      "learning_rate": 0.012681447616115817,
      "loss": 0.8087,
      "num_input_tokens_seen": 46343968,
      "step": 79825
    },
    {
      "epoch": 11.890080428954423,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.012679521414122647,
      "loss": 0.7927,
      "num_input_tokens_seen": 46347168,
      "step": 79830
    },
    {
      "epoch": 11.890825141495382,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0126775952513316,
      "loss": 0.8045,
      "num_input_tokens_seen": 46349952,
      "step": 79835
    },
    {
      "epoch": 11.891569854036343,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.012675669127775223,
      "loss": 0.8,
      "num_input_tokens_seen": 46352768,
      "step": 79840
    },
    {
      "epoch": 11.892314566577301,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.01267374304348605,
      "loss": 0.8156,
      "num_input_tokens_seen": 46355776,
      "step": 79845
    },
    {
      "epoch": 11.89305927911826,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.012671816998496628,
      "loss": 0.8,
      "num_input_tokens_seen": 46358624,
      "step": 79850
    },
    {
      "epoch": 11.893803991659219,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.012669890992839487,
      "loss": 0.8049,
      "num_input_tokens_seen": 46361504,
      "step": 79855
    },
    {
      "epoch": 11.89454870420018,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.012667965026547174,
      "loss": 0.7974,
      "num_input_tokens_seen": 46364384,
      "step": 79860
    },
    {
      "epoch": 11.895293416741138,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.012666039099652221,
      "loss": 0.7814,
      "num_input_tokens_seen": 46367168,
      "step": 79865
    },
    {
      "epoch": 11.896038129282097,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.012664113212187167,
      "loss": 0.7922,
      "num_input_tokens_seen": 46370048,
      "step": 79870
    },
    {
      "epoch": 11.896782841823056,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.012662187364184545,
      "loss": 0.7825,
      "num_input_tokens_seen": 46372608,
      "step": 79875
    },
    {
      "epoch": 11.897527554364016,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.012660261555676887,
      "loss": 0.7786,
      "num_input_tokens_seen": 46375584,
      "step": 79880
    },
    {
      "epoch": 11.898272266904975,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.012658335786696739,
      "loss": 0.7984,
      "num_input_tokens_seen": 46378368,
      "step": 79885
    },
    {
      "epoch": 11.899016979445934,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.012656410057276619,
      "loss": 0.7868,
      "num_input_tokens_seen": 46381248,
      "step": 79890
    },
    {
      "epoch": 11.899761691986892,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.012654484367449077,
      "loss": 0.7977,
      "num_input_tokens_seen": 46384160,
      "step": 79895
    },
    {
      "epoch": 11.900506404527853,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.012652558717246635,
      "loss": 0.7936,
      "num_input_tokens_seen": 46386880,
      "step": 79900
    },
    {
      "epoch": 11.901251117068812,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.01265063310670183,
      "loss": 0.7999,
      "num_input_tokens_seen": 46389408,
      "step": 79905
    },
    {
      "epoch": 11.90199582960977,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.01264870753584719,
      "loss": 0.8768,
      "num_input_tokens_seen": 46392576,
      "step": 79910
    },
    {
      "epoch": 11.90274054215073,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.012646782004715243,
      "loss": 0.7713,
      "num_input_tokens_seen": 46395456,
      "step": 79915
    },
    {
      "epoch": 11.90348525469169,
      "grad_norm": 0.03125,
      "learning_rate": 0.012644856513338526,
      "loss": 0.797,
      "num_input_tokens_seen": 46398304,
      "step": 79920
    },
    {
      "epoch": 11.904229967232649,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.012642931061749558,
      "loss": 0.8165,
      "num_input_tokens_seen": 46401120,
      "step": 79925
    },
    {
      "epoch": 11.904974679773607,
      "grad_norm": 0.064453125,
      "learning_rate": 0.012641005649980882,
      "loss": 0.7785,
      "num_input_tokens_seen": 46404128,
      "step": 79930
    },
    {
      "epoch": 11.905719392314566,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.012639080278065014,
      "loss": 0.8107,
      "num_input_tokens_seen": 46406880,
      "step": 79935
    },
    {
      "epoch": 11.906464104855527,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.01263715494603449,
      "loss": 0.8009,
      "num_input_tokens_seen": 46409952,
      "step": 79940
    },
    {
      "epoch": 11.907208817396485,
      "grad_norm": 0.046875,
      "learning_rate": 0.012635229653921822,
      "loss": 0.7999,
      "num_input_tokens_seen": 46413088,
      "step": 79945
    },
    {
      "epoch": 11.907953529937444,
      "grad_norm": 0.044921875,
      "learning_rate": 0.012633304401759554,
      "loss": 0.7885,
      "num_input_tokens_seen": 46415968,
      "step": 79950
    },
    {
      "epoch": 11.908698242478403,
      "grad_norm": 0.03515625,
      "learning_rate": 0.012631379189580202,
      "loss": 0.8014,
      "num_input_tokens_seen": 46418656,
      "step": 79955
    },
    {
      "epoch": 11.909442955019362,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.012629454017416287,
      "loss": 0.8162,
      "num_input_tokens_seen": 46421600,
      "step": 79960
    },
    {
      "epoch": 11.910187667560322,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.01262752888530034,
      "loss": 0.799,
      "num_input_tokens_seen": 46424512,
      "step": 79965
    },
    {
      "epoch": 11.910932380101281,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.01262560379326488,
      "loss": 0.7944,
      "num_input_tokens_seen": 46427488,
      "step": 79970
    },
    {
      "epoch": 11.91167709264224,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.012623678741342435,
      "loss": 0.7925,
      "num_input_tokens_seen": 46430112,
      "step": 79975
    },
    {
      "epoch": 11.912421805183198,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.012621753729565515,
      "loss": 0.8018,
      "num_input_tokens_seen": 46433024,
      "step": 79980
    },
    {
      "epoch": 11.913166517724159,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.012619828757966656,
      "loss": 0.8461,
      "num_input_tokens_seen": 46435904,
      "step": 79985
    },
    {
      "epoch": 11.913911230265118,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.01261790382657837,
      "loss": 0.8068,
      "num_input_tokens_seen": 46438688,
      "step": 79990
    },
    {
      "epoch": 11.914655942806077,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.01261597893543317,
      "loss": 0.7875,
      "num_input_tokens_seen": 46441312,
      "step": 79995
    },
    {
      "epoch": 11.915400655347035,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.01261405408456359,
      "loss": 0.8,
      "num_input_tokens_seen": 46444256,
      "step": 80000
    },
    {
      "epoch": 11.916145367887996,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.012612129274002138,
      "loss": 0.7962,
      "num_input_tokens_seen": 46447200,
      "step": 80005
    },
    {
      "epoch": 11.916890080428955,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.01261020450378134,
      "loss": 0.7843,
      "num_input_tokens_seen": 46450048,
      "step": 80010
    },
    {
      "epoch": 11.917634792969913,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.0126082797739337,
      "loss": 0.804,
      "num_input_tokens_seen": 46453088,
      "step": 80015
    },
    {
      "epoch": 11.918379505510872,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.01260635508449175,
      "loss": 0.794,
      "num_input_tokens_seen": 46456256,
      "step": 80020
    },
    {
      "epoch": 11.919124218051833,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.01260443043548799,
      "loss": 0.7882,
      "num_input_tokens_seen": 46459040,
      "step": 80025
    },
    {
      "epoch": 11.919868930592791,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01260250582695495,
      "loss": 0.804,
      "num_input_tokens_seen": 46461952,
      "step": 80030
    },
    {
      "epoch": 11.92061364313375,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.012600581258925138,
      "loss": 0.8092,
      "num_input_tokens_seen": 46464896,
      "step": 80035
    },
    {
      "epoch": 11.921358355674709,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.012598656731431065,
      "loss": 0.8066,
      "num_input_tokens_seen": 46467680,
      "step": 80040
    },
    {
      "epoch": 11.92210306821567,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.012596732244505247,
      "loss": 0.7946,
      "num_input_tokens_seen": 46470432,
      "step": 80045
    },
    {
      "epoch": 11.922847780756628,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.012594807798180188,
      "loss": 0.7997,
      "num_input_tokens_seen": 46473408,
      "step": 80050
    },
    {
      "epoch": 11.923592493297587,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.012592883392488414,
      "loss": 0.7839,
      "num_input_tokens_seen": 46476096,
      "step": 80055
    },
    {
      "epoch": 11.924337205838546,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.012590959027462423,
      "loss": 0.7807,
      "num_input_tokens_seen": 46479168,
      "step": 80060
    },
    {
      "epoch": 11.925081918379506,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.012589034703134739,
      "loss": 0.7915,
      "num_input_tokens_seen": 46482240,
      "step": 80065
    },
    {
      "epoch": 11.925826630920465,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.012587110419537858,
      "loss": 0.7999,
      "num_input_tokens_seen": 46485088,
      "step": 80070
    },
    {
      "epoch": 11.926571343461424,
      "grad_norm": 0.03125,
      "learning_rate": 0.012585186176704295,
      "loss": 0.8126,
      "num_input_tokens_seen": 46487840,
      "step": 80075
    },
    {
      "epoch": 11.927316056002383,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.01258326197466656,
      "loss": 0.8028,
      "num_input_tokens_seen": 46490976,
      "step": 80080
    },
    {
      "epoch": 11.928060768543343,
      "grad_norm": 0.02734375,
      "learning_rate": 0.01258133781345715,
      "loss": 0.8044,
      "num_input_tokens_seen": 46493536,
      "step": 80085
    },
    {
      "epoch": 11.928805481084302,
      "grad_norm": 0.03515625,
      "learning_rate": 0.012579413693108587,
      "loss": 0.7938,
      "num_input_tokens_seen": 46496416,
      "step": 80090
    },
    {
      "epoch": 11.92955019362526,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.012577489613653365,
      "loss": 0.7931,
      "num_input_tokens_seen": 46499264,
      "step": 80095
    },
    {
      "epoch": 11.93029490616622,
      "grad_norm": 0.0390625,
      "learning_rate": 0.012575565575123996,
      "loss": 0.8073,
      "num_input_tokens_seen": 46501920,
      "step": 80100
    },
    {
      "epoch": 11.931039618707178,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.012573641577552981,
      "loss": 0.7924,
      "num_input_tokens_seen": 46504704,
      "step": 80105
    },
    {
      "epoch": 11.931784331248139,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.01257171762097283,
      "loss": 0.7996,
      "num_input_tokens_seen": 46507552,
      "step": 80110
    },
    {
      "epoch": 11.932529043789097,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.012569793705416039,
      "loss": 0.7977,
      "num_input_tokens_seen": 46510080,
      "step": 80115
    },
    {
      "epoch": 11.933273756330056,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01256786983091511,
      "loss": 0.8271,
      "num_input_tokens_seen": 46512736,
      "step": 80120
    },
    {
      "epoch": 11.934018468871017,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.012565945997502553,
      "loss": 0.8051,
      "num_input_tokens_seen": 46515584,
      "step": 80125
    },
    {
      "epoch": 11.934763181411975,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.012564022205210862,
      "loss": 0.7921,
      "num_input_tokens_seen": 46518688,
      "step": 80130
    },
    {
      "epoch": 11.935507893952934,
      "grad_norm": 0.076171875,
      "learning_rate": 0.012562098454072539,
      "loss": 0.7904,
      "num_input_tokens_seen": 46521568,
      "step": 80135
    },
    {
      "epoch": 11.936252606493893,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.012560174744120085,
      "loss": 0.8122,
      "num_input_tokens_seen": 46524480,
      "step": 80140
    },
    {
      "epoch": 11.936997319034852,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.012558251075386004,
      "loss": 0.7941,
      "num_input_tokens_seen": 46527648,
      "step": 80145
    },
    {
      "epoch": 11.937742031575812,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.012556327447902789,
      "loss": 0.7916,
      "num_input_tokens_seen": 46530784,
      "step": 80150
    },
    {
      "epoch": 11.938486744116771,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01255440386170293,
      "loss": 0.807,
      "num_input_tokens_seen": 46533664,
      "step": 80155
    },
    {
      "epoch": 11.93923145665773,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.012552480316818939,
      "loss": 0.8383,
      "num_input_tokens_seen": 46536512,
      "step": 80160
    },
    {
      "epoch": 11.939976169198689,
      "grad_norm": 0.072265625,
      "learning_rate": 0.012550556813283303,
      "loss": 0.8045,
      "num_input_tokens_seen": 46539584,
      "step": 80165
    },
    {
      "epoch": 11.940720881739649,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.012548633351128522,
      "loss": 0.8106,
      "num_input_tokens_seen": 46542624,
      "step": 80170
    },
    {
      "epoch": 11.941465594280608,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.012546709930387085,
      "loss": 0.7847,
      "num_input_tokens_seen": 46545632,
      "step": 80175
    },
    {
      "epoch": 11.942210306821567,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.012544786551091496,
      "loss": 0.8022,
      "num_input_tokens_seen": 46548256,
      "step": 80180
    },
    {
      "epoch": 11.942955019362525,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.012542863213274235,
      "loss": 0.7875,
      "num_input_tokens_seen": 46551040,
      "step": 80185
    },
    {
      "epoch": 11.943699731903486,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.012540939916967812,
      "loss": 0.7777,
      "num_input_tokens_seen": 46553728,
      "step": 80190
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.012539016662204706,
      "loss": 0.7794,
      "num_input_tokens_seen": 46556800,
      "step": 80195
    },
    {
      "epoch": 11.945189156985403,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.012537093449017411,
      "loss": 0.7907,
      "num_input_tokens_seen": 46559840,
      "step": 80200
    },
    {
      "epoch": 11.945933869526362,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.012535170277438423,
      "loss": 0.7949,
      "num_input_tokens_seen": 46562336,
      "step": 80205
    },
    {
      "epoch": 11.946678582067323,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.01253324714750022,
      "loss": 0.8076,
      "num_input_tokens_seen": 46565056,
      "step": 80210
    },
    {
      "epoch": 11.947423294608281,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.012531324059235306,
      "loss": 0.8042,
      "num_input_tokens_seen": 46568032,
      "step": 80215
    },
    {
      "epoch": 11.94816800714924,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.012529401012676158,
      "loss": 0.8067,
      "num_input_tokens_seen": 46570656,
      "step": 80220
    },
    {
      "epoch": 11.948912719690199,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.012527478007855274,
      "loss": 0.7887,
      "num_input_tokens_seen": 46573568,
      "step": 80225
    },
    {
      "epoch": 11.94965743223116,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.012525555044805136,
      "loss": 0.7802,
      "num_input_tokens_seen": 46576384,
      "step": 80230
    },
    {
      "epoch": 11.950402144772118,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.012523632123558228,
      "loss": 0.8109,
      "num_input_tokens_seen": 46579424,
      "step": 80235
    },
    {
      "epoch": 11.951146857313077,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.012521709244147042,
      "loss": 0.8064,
      "num_input_tokens_seen": 46582048,
      "step": 80240
    },
    {
      "epoch": 11.951891569854036,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.01251978640660405,
      "loss": 0.7888,
      "num_input_tokens_seen": 46584768,
      "step": 80245
    },
    {
      "epoch": 11.952636282394996,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.012517863610961756,
      "loss": 0.7834,
      "num_input_tokens_seen": 46587872,
      "step": 80250
    },
    {
      "epoch": 11.953380994935955,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.012515940857252628,
      "loss": 0.7995,
      "num_input_tokens_seen": 46590880,
      "step": 80255
    },
    {
      "epoch": 11.954125707476914,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.012514018145509157,
      "loss": 0.7865,
      "num_input_tokens_seen": 46593600,
      "step": 80260
    },
    {
      "epoch": 11.954870420017873,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.012512095475763822,
      "loss": 0.7928,
      "num_input_tokens_seen": 46596320,
      "step": 80265
    },
    {
      "epoch": 11.955615132558833,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.012510172848049108,
      "loss": 0.8241,
      "num_input_tokens_seen": 46599008,
      "step": 80270
    },
    {
      "epoch": 11.956359845099792,
      "grad_norm": 0.03515625,
      "learning_rate": 0.012508250262397493,
      "loss": 0.8179,
      "num_input_tokens_seen": 46602112,
      "step": 80275
    },
    {
      "epoch": 11.95710455764075,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.012506327718841451,
      "loss": 0.7929,
      "num_input_tokens_seen": 46605216,
      "step": 80280
    },
    {
      "epoch": 11.95784927018171,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.012504405217413476,
      "loss": 0.7879,
      "num_input_tokens_seen": 46608352,
      "step": 80285
    },
    {
      "epoch": 11.958593982722668,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.012502482758146029,
      "loss": 0.8463,
      "num_input_tokens_seen": 46611136,
      "step": 80290
    },
    {
      "epoch": 11.959338695263629,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.012500560341071604,
      "loss": 0.7891,
      "num_input_tokens_seen": 46613824,
      "step": 80295
    },
    {
      "epoch": 11.960083407804587,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.012498637966222671,
      "loss": 0.8172,
      "num_input_tokens_seen": 46616896,
      "step": 80300
    },
    {
      "epoch": 11.960828120345546,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.01249671563363171,
      "loss": 0.8069,
      "num_input_tokens_seen": 46619680,
      "step": 80305
    },
    {
      "epoch": 11.961572832886507,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.012494793343331192,
      "loss": 0.7944,
      "num_input_tokens_seen": 46622336,
      "step": 80310
    },
    {
      "epoch": 11.962317545427466,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.012492871095353589,
      "loss": 0.8204,
      "num_input_tokens_seen": 46625184,
      "step": 80315
    },
    {
      "epoch": 11.963062257968424,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.01249094888973139,
      "loss": 0.7923,
      "num_input_tokens_seen": 46627840,
      "step": 80320
    },
    {
      "epoch": 11.963806970509383,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.012489026726497049,
      "loss": 0.8005,
      "num_input_tokens_seen": 46630816,
      "step": 80325
    },
    {
      "epoch": 11.964551683050342,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.012487104605683055,
      "loss": 0.807,
      "num_input_tokens_seen": 46634208,
      "step": 80330
    },
    {
      "epoch": 11.965296395591302,
      "grad_norm": 0.037109375,
      "learning_rate": 0.012485182527321875,
      "loss": 0.7765,
      "num_input_tokens_seen": 46637024,
      "step": 80335
    },
    {
      "epoch": 11.966041108132261,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.012483260491445982,
      "loss": 0.7965,
      "num_input_tokens_seen": 46639872,
      "step": 80340
    },
    {
      "epoch": 11.96678582067322,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.012481338498087838,
      "loss": 0.826,
      "num_input_tokens_seen": 46642624,
      "step": 80345
    },
    {
      "epoch": 11.967530533214179,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.012479416547279928,
      "loss": 0.8101,
      "num_input_tokens_seen": 46645216,
      "step": 80350
    },
    {
      "epoch": 11.96827524575514,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.012477494639054712,
      "loss": 0.782,
      "num_input_tokens_seen": 46648096,
      "step": 80355
    },
    {
      "epoch": 11.969019958296098,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.012475572773444656,
      "loss": 0.8022,
      "num_input_tokens_seen": 46650976,
      "step": 80360
    },
    {
      "epoch": 11.969764670837057,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.012473650950482239,
      "loss": 0.8056,
      "num_input_tokens_seen": 46654208,
      "step": 80365
    },
    {
      "epoch": 11.970509383378015,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.012471729170199916,
      "loss": 0.7891,
      "num_input_tokens_seen": 46656768,
      "step": 80370
    },
    {
      "epoch": 11.971254095918976,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.012469807432630164,
      "loss": 0.8103,
      "num_input_tokens_seen": 46659424,
      "step": 80375
    },
    {
      "epoch": 11.971998808459935,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.012467885737805436,
      "loss": 0.7874,
      "num_input_tokens_seen": 46662336,
      "step": 80380
    },
    {
      "epoch": 11.972743521000893,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01246596408575821,
      "loss": 0.7993,
      "num_input_tokens_seen": 46665152,
      "step": 80385
    },
    {
      "epoch": 11.973488233541852,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.01246404247652095,
      "loss": 0.7777,
      "num_input_tokens_seen": 46668032,
      "step": 80390
    },
    {
      "epoch": 11.974232946082813,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.012462120910126103,
      "loss": 0.7666,
      "num_input_tokens_seen": 46670944,
      "step": 80395
    },
    {
      "epoch": 11.974977658623772,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.012460199386606155,
      "loss": 0.8124,
      "num_input_tokens_seen": 46673664,
      "step": 80400
    },
    {
      "epoch": 11.97572237116473,
      "grad_norm": 0.03125,
      "learning_rate": 0.012458277905993551,
      "loss": 0.8118,
      "num_input_tokens_seen": 46676640,
      "step": 80405
    },
    {
      "epoch": 11.976467083705689,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.012456356468320762,
      "loss": 0.8034,
      "num_input_tokens_seen": 46679648,
      "step": 80410
    },
    {
      "epoch": 11.97721179624665,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.012454435073620242,
      "loss": 0.8161,
      "num_input_tokens_seen": 46682528,
      "step": 80415
    },
    {
      "epoch": 11.977956508787608,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.012452513721924458,
      "loss": 0.7931,
      "num_input_tokens_seen": 46685920,
      "step": 80420
    },
    {
      "epoch": 11.978701221328567,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01245059241326586,
      "loss": 0.8041,
      "num_input_tokens_seen": 46688672,
      "step": 80425
    },
    {
      "epoch": 11.979445933869526,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.012448671147676916,
      "loss": 0.8098,
      "num_input_tokens_seen": 46691776,
      "step": 80430
    },
    {
      "epoch": 11.980190646410486,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.01244674992519008,
      "loss": 0.7976,
      "num_input_tokens_seen": 46694656,
      "step": 80435
    },
    {
      "epoch": 11.980935358951445,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.012444828745837808,
      "loss": 0.8043,
      "num_input_tokens_seen": 46697312,
      "step": 80440
    },
    {
      "epoch": 11.981680071492404,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01244290760965256,
      "loss": 0.8,
      "num_input_tokens_seen": 46700256,
      "step": 80445
    },
    {
      "epoch": 11.982424784033363,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.012440986516666783,
      "loss": 0.7977,
      "num_input_tokens_seen": 46703136,
      "step": 80450
    },
    {
      "epoch": 11.983169496574323,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.012439065466912945,
      "loss": 0.7814,
      "num_input_tokens_seen": 46706048,
      "step": 80455
    },
    {
      "epoch": 11.983914209115282,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.012437144460423486,
      "loss": 0.7988,
      "num_input_tokens_seen": 46708832,
      "step": 80460
    },
    {
      "epoch": 11.98465892165624,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.012435223497230872,
      "loss": 0.8201,
      "num_input_tokens_seen": 46711648,
      "step": 80465
    },
    {
      "epoch": 11.9854036341972,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.012433302577367546,
      "loss": 0.8027,
      "num_input_tokens_seen": 46714752,
      "step": 80470
    },
    {
      "epoch": 11.986148346738158,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.012431381700865971,
      "loss": 0.7847,
      "num_input_tokens_seen": 46717888,
      "step": 80475
    },
    {
      "epoch": 11.986893059279119,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.012429460867758589,
      "loss": 0.7998,
      "num_input_tokens_seen": 46720736,
      "step": 80480
    },
    {
      "epoch": 11.987637771820078,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.012427540078077845,
      "loss": 0.8045,
      "num_input_tokens_seen": 46723520,
      "step": 80485
    },
    {
      "epoch": 11.988382484361036,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.012425619331856206,
      "loss": 0.8068,
      "num_input_tokens_seen": 46727040,
      "step": 80490
    },
    {
      "epoch": 11.989127196901995,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.012423698629126111,
      "loss": 0.8035,
      "num_input_tokens_seen": 46729824,
      "step": 80495
    },
    {
      "epoch": 11.989871909442956,
      "grad_norm": 0.03515625,
      "learning_rate": 0.012421777969920009,
      "loss": 0.7926,
      "num_input_tokens_seen": 46732608,
      "step": 80500
    },
    {
      "epoch": 11.990616621983914,
      "grad_norm": 0.0390625,
      "learning_rate": 0.01241985735427034,
      "loss": 0.7844,
      "num_input_tokens_seen": 46735456,
      "step": 80505
    },
    {
      "epoch": 11.991361334524873,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.012417936782209569,
      "loss": 0.7812,
      "num_input_tokens_seen": 46738656,
      "step": 80510
    },
    {
      "epoch": 11.992106047065832,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.012416016253770128,
      "loss": 0.7982,
      "num_input_tokens_seen": 46741888,
      "step": 80515
    },
    {
      "epoch": 11.992850759606792,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.012414095768984465,
      "loss": 0.8002,
      "num_input_tokens_seen": 46744704,
      "step": 80520
    },
    {
      "epoch": 11.993595472147751,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.012412175327885027,
      "loss": 0.8086,
      "num_input_tokens_seen": 46747264,
      "step": 80525
    },
    {
      "epoch": 11.99434018468871,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.012410254930504254,
      "loss": 0.8046,
      "num_input_tokens_seen": 46750240,
      "step": 80530
    },
    {
      "epoch": 11.995084897229669,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.012408334576874597,
      "loss": 0.7924,
      "num_input_tokens_seen": 46753344,
      "step": 80535
    },
    {
      "epoch": 11.99582960977063,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.012406414267028487,
      "loss": 0.8147,
      "num_input_tokens_seen": 46756128,
      "step": 80540
    },
    {
      "epoch": 11.996574322311588,
      "grad_norm": 0.041015625,
      "learning_rate": 0.012404494000998377,
      "loss": 0.7798,
      "num_input_tokens_seen": 46759168,
      "step": 80545
    },
    {
      "epoch": 11.997319034852547,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.012402573778816699,
      "loss": 0.7931,
      "num_input_tokens_seen": 46762144,
      "step": 80550
    },
    {
      "epoch": 11.998063747393505,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0124006536005159,
      "loss": 0.8008,
      "num_input_tokens_seen": 46765024,
      "step": 80555
    },
    {
      "epoch": 11.998808459934466,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.01239873346612842,
      "loss": 0.812,
      "num_input_tokens_seen": 46768192,
      "step": 80560
    },
    {
      "epoch": 11.999553172475425,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.012396813375686688,
      "loss": 0.7982,
      "num_input_tokens_seen": 46771104,
      "step": 80565
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.7999743819236755,
      "eval_runtime": 70.6809,
      "eval_samples_per_second": 42.218,
      "eval_steps_per_second": 10.554,
      "num_input_tokens_seen": 46772480,
      "step": 80568
    },
    {
      "epoch": 12.000297885016384,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.012394893329223153,
      "loss": 0.7892,
      "num_input_tokens_seen": 46773440,
      "step": 80570
    },
    {
      "epoch": 12.001042597557342,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.012392973326770241,
      "loss": 0.7992,
      "num_input_tokens_seen": 46776384,
      "step": 80575
    },
    {
      "epoch": 12.001787310098303,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.012391053368360402,
      "loss": 0.7996,
      "num_input_tokens_seen": 46779200,
      "step": 80580
    },
    {
      "epoch": 12.002532022639262,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.012389133454026058,
      "loss": 0.7984,
      "num_input_tokens_seen": 46782176,
      "step": 80585
    },
    {
      "epoch": 12.00327673518022,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.012387213583799656,
      "loss": 0.7891,
      "num_input_tokens_seen": 46784992,
      "step": 80590
    },
    {
      "epoch": 12.004021447721179,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.012385293757713625,
      "loss": 0.7764,
      "num_input_tokens_seen": 46788032,
      "step": 80595
    },
    {
      "epoch": 12.00476616026214,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.012383373975800395,
      "loss": 0.7867,
      "num_input_tokens_seen": 46790880,
      "step": 80600
    },
    {
      "epoch": 12.005510872803098,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.012381454238092409,
      "loss": 0.8056,
      "num_input_tokens_seen": 46794464,
      "step": 80605
    },
    {
      "epoch": 12.006255585344057,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.012379534544622081,
      "loss": 0.8003,
      "num_input_tokens_seen": 46797536,
      "step": 80610
    },
    {
      "epoch": 12.007000297885016,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.012377614895421862,
      "loss": 0.804,
      "num_input_tokens_seen": 46800896,
      "step": 80615
    },
    {
      "epoch": 12.007745010425976,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01237569529052417,
      "loss": 0.7843,
      "num_input_tokens_seen": 46803648,
      "step": 80620
    },
    {
      "epoch": 12.008489722966935,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.012373775729961442,
      "loss": 0.7956,
      "num_input_tokens_seen": 46806400,
      "step": 80625
    },
    {
      "epoch": 12.009234435507894,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0123718562137661,
      "loss": 0.8143,
      "num_input_tokens_seen": 46809088,
      "step": 80630
    },
    {
      "epoch": 12.009979148048853,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.012369936741970584,
      "loss": 0.7928,
      "num_input_tokens_seen": 46812096,
      "step": 80635
    },
    {
      "epoch": 12.010723860589811,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.01236801731460731,
      "loss": 0.809,
      "num_input_tokens_seen": 46814720,
      "step": 80640
    },
    {
      "epoch": 12.011468573130772,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.012366097931708703,
      "loss": 0.7923,
      "num_input_tokens_seen": 46817600,
      "step": 80645
    },
    {
      "epoch": 12.01221328567173,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0123641785933072,
      "loss": 0.7923,
      "num_input_tokens_seen": 46820288,
      "step": 80650
    },
    {
      "epoch": 12.01295799821269,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.012362259299435215,
      "loss": 0.8033,
      "num_input_tokens_seen": 46823104,
      "step": 80655
    },
    {
      "epoch": 12.013702710753648,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.012360340050125186,
      "loss": 0.8061,
      "num_input_tokens_seen": 46825888,
      "step": 80660
    },
    {
      "epoch": 12.014447423294609,
      "grad_norm": 0.03515625,
      "learning_rate": 0.012358420845409524,
      "loss": 0.8353,
      "num_input_tokens_seen": 46828864,
      "step": 80665
    },
    {
      "epoch": 12.015192135835568,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.01235650168532066,
      "loss": 0.8106,
      "num_input_tokens_seen": 46831712,
      "step": 80670
    },
    {
      "epoch": 12.015936848376526,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.012354582569891014,
      "loss": 0.7923,
      "num_input_tokens_seen": 46834720,
      "step": 80675
    },
    {
      "epoch": 12.016681560917485,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.012352663499153002,
      "loss": 0.8025,
      "num_input_tokens_seen": 46837728,
      "step": 80680
    },
    {
      "epoch": 12.017426273458446,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.012350744473139054,
      "loss": 0.8081,
      "num_input_tokens_seen": 46840640,
      "step": 80685
    },
    {
      "epoch": 12.018170985999404,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01234882549188158,
      "loss": 0.7879,
      "num_input_tokens_seen": 46844128,
      "step": 80690
    },
    {
      "epoch": 12.018915698540363,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01234690655541301,
      "loss": 0.7992,
      "num_input_tokens_seen": 46846912,
      "step": 80695
    },
    {
      "epoch": 12.019660411081322,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.012344987663765754,
      "loss": 0.7803,
      "num_input_tokens_seen": 46849632,
      "step": 80700
    },
    {
      "epoch": 12.020405123622282,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.012343068816972239,
      "loss": 0.7927,
      "num_input_tokens_seen": 46852416,
      "step": 80705
    },
    {
      "epoch": 12.021149836163241,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.012341150015064866,
      "loss": 0.798,
      "num_input_tokens_seen": 46855520,
      "step": 80710
    },
    {
      "epoch": 12.0218945487042,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.012339231258076068,
      "loss": 0.798,
      "num_input_tokens_seen": 46858240,
      "step": 80715
    },
    {
      "epoch": 12.022639261245159,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.012337312546038256,
      "loss": 0.7923,
      "num_input_tokens_seen": 46861152,
      "step": 80720
    },
    {
      "epoch": 12.02338397378612,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.012335393878983835,
      "loss": 0.7941,
      "num_input_tokens_seen": 46863904,
      "step": 80725
    },
    {
      "epoch": 12.024128686327078,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.012333475256945232,
      "loss": 0.8045,
      "num_input_tokens_seen": 46867200,
      "step": 80730
    },
    {
      "epoch": 12.024873398868037,
      "grad_norm": 0.03125,
      "learning_rate": 0.012331556679954852,
      "loss": 0.7878,
      "num_input_tokens_seen": 46870240,
      "step": 80735
    },
    {
      "epoch": 12.025618111408996,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.012329638148045113,
      "loss": 0.7995,
      "num_input_tokens_seen": 46873184,
      "step": 80740
    },
    {
      "epoch": 12.026362823949956,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.012327719661248415,
      "loss": 0.7917,
      "num_input_tokens_seen": 46876544,
      "step": 80745
    },
    {
      "epoch": 12.027107536490915,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.012325801219597187,
      "loss": 0.7906,
      "num_input_tokens_seen": 46879520,
      "step": 80750
    },
    {
      "epoch": 12.027852249031874,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.012323882823123826,
      "loss": 0.7753,
      "num_input_tokens_seen": 46882688,
      "step": 80755
    },
    {
      "epoch": 12.028596961572832,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.012321964471860741,
      "loss": 0.7853,
      "num_input_tokens_seen": 46886720,
      "step": 80760
    },
    {
      "epoch": 12.029341674113793,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.01232004616584035,
      "loss": 0.8075,
      "num_input_tokens_seen": 46889760,
      "step": 80765
    },
    {
      "epoch": 12.030086386654752,
      "grad_norm": 0.0390625,
      "learning_rate": 0.012318127905095051,
      "loss": 0.7935,
      "num_input_tokens_seen": 46892448,
      "step": 80770
    },
    {
      "epoch": 12.03083109919571,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.01231620968965726,
      "loss": 0.7878,
      "num_input_tokens_seen": 46895168,
      "step": 80775
    },
    {
      "epoch": 12.03157581173667,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.012314291519559371,
      "loss": 0.7682,
      "num_input_tokens_seen": 46898336,
      "step": 80780
    },
    {
      "epoch": 12.03232052427763,
      "grad_norm": 0.0859375,
      "learning_rate": 0.012312373394833804,
      "loss": 0.8127,
      "num_input_tokens_seen": 46901184,
      "step": 80785
    },
    {
      "epoch": 12.033065236818588,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.012310455315512956,
      "loss": 0.8055,
      "num_input_tokens_seen": 46904448,
      "step": 80790
    },
    {
      "epoch": 12.033809949359547,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.012308537281629228,
      "loss": 0.7748,
      "num_input_tokens_seen": 46907360,
      "step": 80795
    },
    {
      "epoch": 12.034554661900506,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.012306619293215033,
      "loss": 0.7653,
      "num_input_tokens_seen": 46910240,
      "step": 80800
    },
    {
      "epoch": 12.035299374441466,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.012304701350302764,
      "loss": 0.7958,
      "num_input_tokens_seen": 46913024,
      "step": 80805
    },
    {
      "epoch": 12.036044086982425,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.012302783452924829,
      "loss": 0.8367,
      "num_input_tokens_seen": 46915840,
      "step": 80810
    },
    {
      "epoch": 12.036788799523384,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01230086560111362,
      "loss": 0.7844,
      "num_input_tokens_seen": 46918848,
      "step": 80815
    },
    {
      "epoch": 12.037533512064343,
      "grad_norm": 0.06640625,
      "learning_rate": 0.01229894779490155,
      "loss": 0.813,
      "num_input_tokens_seen": 46921888,
      "step": 80820
    },
    {
      "epoch": 12.038278224605302,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.012297030034321008,
      "loss": 0.7872,
      "num_input_tokens_seen": 46924896,
      "step": 80825
    },
    {
      "epoch": 12.039022937146262,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0122951123194044,
      "loss": 0.7787,
      "num_input_tokens_seen": 46927872,
      "step": 80830
    },
    {
      "epoch": 12.03976764968722,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.012293194650184117,
      "loss": 0.7738,
      "num_input_tokens_seen": 46930848,
      "step": 80835
    },
    {
      "epoch": 12.04051236222818,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01229127702669256,
      "loss": 0.7865,
      "num_input_tokens_seen": 46933984,
      "step": 80840
    },
    {
      "epoch": 12.041257074769138,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.012289359448962126,
      "loss": 0.7958,
      "num_input_tokens_seen": 46936736,
      "step": 80845
    },
    {
      "epoch": 12.042001787310099,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.012287441917025204,
      "loss": 0.7926,
      "num_input_tokens_seen": 46939552,
      "step": 80850
    },
    {
      "epoch": 12.042746499851058,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.012285524430914201,
      "loss": 0.8338,
      "num_input_tokens_seen": 46942304,
      "step": 80855
    },
    {
      "epoch": 12.043491212392016,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0122836069906615,
      "loss": 0.7727,
      "num_input_tokens_seen": 46945088,
      "step": 80860
    },
    {
      "epoch": 12.044235924932975,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0122816895962995,
      "loss": 0.8057,
      "num_input_tokens_seen": 46947680,
      "step": 80865
    },
    {
      "epoch": 12.044980637473936,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.012279772247860586,
      "loss": 0.7812,
      "num_input_tokens_seen": 46950336,
      "step": 80870
    },
    {
      "epoch": 12.045725350014894,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.012277854945377163,
      "loss": 0.7789,
      "num_input_tokens_seen": 46953280,
      "step": 80875
    },
    {
      "epoch": 12.046470062555853,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.012275937688881612,
      "loss": 0.782,
      "num_input_tokens_seen": 46956064,
      "step": 80880
    },
    {
      "epoch": 12.047214775096812,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.01227402047840632,
      "loss": 0.8589,
      "num_input_tokens_seen": 46958848,
      "step": 80885
    },
    {
      "epoch": 12.047959487637772,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.012272103313983687,
      "loss": 0.7995,
      "num_input_tokens_seen": 46961760,
      "step": 80890
    },
    {
      "epoch": 12.048704200178731,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.012270186195646094,
      "loss": 0.7872,
      "num_input_tokens_seen": 46964672,
      "step": 80895
    },
    {
      "epoch": 12.04944891271969,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.012268269123425931,
      "loss": 0.7912,
      "num_input_tokens_seen": 46967328,
      "step": 80900
    },
    {
      "epoch": 12.050193625260649,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.012266352097355583,
      "loss": 0.7877,
      "num_input_tokens_seen": 46970112,
      "step": 80905
    },
    {
      "epoch": 12.05093833780161,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.012264435117467441,
      "loss": 0.8095,
      "num_input_tokens_seen": 46972800,
      "step": 80910
    },
    {
      "epoch": 12.051683050342568,
      "grad_norm": 0.033203125,
      "learning_rate": 0.012262518183793888,
      "loss": 0.8007,
      "num_input_tokens_seen": 46976032,
      "step": 80915
    },
    {
      "epoch": 12.052427762883527,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.012260601296367302,
      "loss": 0.7887,
      "num_input_tokens_seen": 46979232,
      "step": 80920
    },
    {
      "epoch": 12.053172475424486,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.012258684455220078,
      "loss": 0.8053,
      "num_input_tokens_seen": 46981984,
      "step": 80925
    },
    {
      "epoch": 12.053917187965446,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.01225676766038459,
      "loss": 0.7925,
      "num_input_tokens_seen": 46985056,
      "step": 80930
    },
    {
      "epoch": 12.054661900506405,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.012254850911893232,
      "loss": 0.7761,
      "num_input_tokens_seen": 46987968,
      "step": 80935
    },
    {
      "epoch": 12.055406613047364,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.012252934209778368,
      "loss": 0.8186,
      "num_input_tokens_seen": 46990752,
      "step": 80940
    },
    {
      "epoch": 12.056151325588322,
      "grad_norm": 0.04296875,
      "learning_rate": 0.012251017554072397,
      "loss": 0.7957,
      "num_input_tokens_seen": 46993888,
      "step": 80945
    },
    {
      "epoch": 12.056896038129283,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.012249100944807682,
      "loss": 0.781,
      "num_input_tokens_seen": 46996736,
      "step": 80950
    },
    {
      "epoch": 12.057640750670242,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.012247184382016617,
      "loss": 0.7928,
      "num_input_tokens_seen": 46999584,
      "step": 80955
    },
    {
      "epoch": 12.0583854632112,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.012245267865731576,
      "loss": 0.7961,
      "num_input_tokens_seen": 47002208,
      "step": 80960
    },
    {
      "epoch": 12.05913017575216,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.01224335139598493,
      "loss": 0.8244,
      "num_input_tokens_seen": 47005088,
      "step": 80965
    },
    {
      "epoch": 12.05987488829312,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.012241434972809065,
      "loss": 0.7643,
      "num_input_tokens_seen": 47007872,
      "step": 80970
    },
    {
      "epoch": 12.060619600834078,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.012239518596236345,
      "loss": 0.8882,
      "num_input_tokens_seen": 47010656,
      "step": 80975
    },
    {
      "epoch": 12.061364313375037,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.012237602266299158,
      "loss": 0.8059,
      "num_input_tokens_seen": 47013536,
      "step": 80980
    },
    {
      "epoch": 12.062109025915996,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.012235685983029867,
      "loss": 0.7781,
      "num_input_tokens_seen": 47016576,
      "step": 80985
    },
    {
      "epoch": 12.062853738456955,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.01223376974646086,
      "loss": 0.8037,
      "num_input_tokens_seen": 47019424,
      "step": 80990
    },
    {
      "epoch": 12.063598450997915,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.012231853556624503,
      "loss": 0.7979,
      "num_input_tokens_seen": 47022528,
      "step": 80995
    },
    {
      "epoch": 12.064343163538874,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.01222993741355316,
      "loss": 0.7869,
      "num_input_tokens_seen": 47025440,
      "step": 81000
    },
    {
      "epoch": 12.065087876079833,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.012228021317279214,
      "loss": 0.7909,
      "num_input_tokens_seen": 47028288,
      "step": 81005
    },
    {
      "epoch": 12.065832588620792,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.012226105267835023,
      "loss": 0.7894,
      "num_input_tokens_seen": 47031360,
      "step": 81010
    },
    {
      "epoch": 12.066577301161752,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.01222418926525297,
      "loss": 0.787,
      "num_input_tokens_seen": 47034208,
      "step": 81015
    },
    {
      "epoch": 12.06732201370271,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.012222273309565414,
      "loss": 0.7907,
      "num_input_tokens_seen": 47036864,
      "step": 81020
    },
    {
      "epoch": 12.06806672624367,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.012220357400804733,
      "loss": 0.8191,
      "num_input_tokens_seen": 47039456,
      "step": 81025
    },
    {
      "epoch": 12.068811438784628,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.012218441539003285,
      "loss": 0.7963,
      "num_input_tokens_seen": 47042528,
      "step": 81030
    },
    {
      "epoch": 12.069556151325589,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.012216525724193444,
      "loss": 0.8113,
      "num_input_tokens_seen": 47045440,
      "step": 81035
    },
    {
      "epoch": 12.070300863866548,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.01221460995640757,
      "loss": 0.7681,
      "num_input_tokens_seen": 47048384,
      "step": 81040
    },
    {
      "epoch": 12.071045576407506,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.012212694235678025,
      "loss": 0.7888,
      "num_input_tokens_seen": 47051328,
      "step": 81045
    },
    {
      "epoch": 12.071790288948465,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.012210778562037184,
      "loss": 0.8107,
      "num_input_tokens_seen": 47054304,
      "step": 81050
    },
    {
      "epoch": 12.072535001489426,
      "grad_norm": 0.0625,
      "learning_rate": 0.012208862935517399,
      "loss": 0.8303,
      "num_input_tokens_seen": 47056960,
      "step": 81055
    },
    {
      "epoch": 12.073279714030384,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.012206947356151044,
      "loss": 0.792,
      "num_input_tokens_seen": 47060224,
      "step": 81060
    },
    {
      "epoch": 12.074024426571343,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01220503182397047,
      "loss": 0.7723,
      "num_input_tokens_seen": 47063360,
      "step": 81065
    },
    {
      "epoch": 12.074769139112302,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.012203116339008047,
      "loss": 0.7888,
      "num_input_tokens_seen": 47066080,
      "step": 81070
    },
    {
      "epoch": 12.075513851653263,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.012201200901296123,
      "loss": 0.781,
      "num_input_tokens_seen": 47068864,
      "step": 81075
    },
    {
      "epoch": 12.076258564194221,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.012199285510867073,
      "loss": 0.7854,
      "num_input_tokens_seen": 47071680,
      "step": 81080
    },
    {
      "epoch": 12.07700327673518,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.012197370167753248,
      "loss": 0.8144,
      "num_input_tokens_seen": 47074688,
      "step": 81085
    },
    {
      "epoch": 12.077747989276139,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.012195454871987,
      "loss": 0.7836,
      "num_input_tokens_seen": 47077760,
      "step": 81090
    },
    {
      "epoch": 12.0784927018171,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.012193539623600697,
      "loss": 0.7949,
      "num_input_tokens_seen": 47080864,
      "step": 81095
    },
    {
      "epoch": 12.079237414358058,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.012191624422626684,
      "loss": 0.7934,
      "num_input_tokens_seen": 47083744,
      "step": 81100
    },
    {
      "epoch": 12.079982126899017,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.01218970926909733,
      "loss": 0.8304,
      "num_input_tokens_seen": 47086720,
      "step": 81105
    },
    {
      "epoch": 12.080726839439976,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01218779416304497,
      "loss": 0.7997,
      "num_input_tokens_seen": 47089536,
      "step": 81110
    },
    {
      "epoch": 12.081471551980936,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.01218587910450198,
      "loss": 0.7894,
      "num_input_tokens_seen": 47092288,
      "step": 81115
    },
    {
      "epoch": 12.082216264521895,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0121839640935007,
      "loss": 0.7828,
      "num_input_tokens_seen": 47094976,
      "step": 81120
    },
    {
      "epoch": 12.082960977062854,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.01218204913007348,
      "loss": 0.8196,
      "num_input_tokens_seen": 47098016,
      "step": 81125
    },
    {
      "epoch": 12.083705689603812,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01218013421425268,
      "loss": 0.7859,
      "num_input_tokens_seen": 47100832,
      "step": 81130
    },
    {
      "epoch": 12.084450402144773,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.012178219346070645,
      "loss": 0.8299,
      "num_input_tokens_seen": 47103744,
      "step": 81135
    },
    {
      "epoch": 12.085195114685732,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.01217630452555973,
      "loss": 0.7934,
      "num_input_tokens_seen": 47106560,
      "step": 81140
    },
    {
      "epoch": 12.08593982722669,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.012174389752752273,
      "loss": 0.794,
      "num_input_tokens_seen": 47109568,
      "step": 81145
    },
    {
      "epoch": 12.08668453976765,
      "grad_norm": 0.033203125,
      "learning_rate": 0.012172475027680636,
      "loss": 0.7947,
      "num_input_tokens_seen": 47112288,
      "step": 81150
    },
    {
      "epoch": 12.08742925230861,
      "grad_norm": 0.0390625,
      "learning_rate": 0.012170560350377153,
      "loss": 0.7746,
      "num_input_tokens_seen": 47115424,
      "step": 81155
    },
    {
      "epoch": 12.088173964849569,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.012168645720874186,
      "loss": 0.8133,
      "num_input_tokens_seen": 47118336,
      "step": 81160
    },
    {
      "epoch": 12.088918677390527,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.012166731139204065,
      "loss": 0.8001,
      "num_input_tokens_seen": 47121248,
      "step": 81165
    },
    {
      "epoch": 12.089663389931486,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.012164816605399142,
      "loss": 0.7999,
      "num_input_tokens_seen": 47124128,
      "step": 81170
    },
    {
      "epoch": 12.090408102472445,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.012162902119491765,
      "loss": 0.7903,
      "num_input_tokens_seen": 47127008,
      "step": 81175
    },
    {
      "epoch": 12.091152815013405,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.012160987681514268,
      "loss": 0.7892,
      "num_input_tokens_seen": 47129504,
      "step": 81180
    },
    {
      "epoch": 12.091897527554364,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.012159073291499003,
      "loss": 0.7689,
      "num_input_tokens_seen": 47132352,
      "step": 81185
    },
    {
      "epoch": 12.092642240095323,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.012157158949478306,
      "loss": 0.7832,
      "num_input_tokens_seen": 47135200,
      "step": 81190
    },
    {
      "epoch": 12.093386952636282,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.012155244655484522,
      "loss": 0.7827,
      "num_input_tokens_seen": 47137888,
      "step": 81195
    },
    {
      "epoch": 12.094131665177242,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.012153330409549986,
      "loss": 0.8128,
      "num_input_tokens_seen": 47140960,
      "step": 81200
    },
    {
      "epoch": 12.094876377718201,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.012151416211707036,
      "loss": 0.8018,
      "num_input_tokens_seen": 47143904,
      "step": 81205
    },
    {
      "epoch": 12.09562109025916,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.012149502061988017,
      "loss": 0.7991,
      "num_input_tokens_seen": 47147008,
      "step": 81210
    },
    {
      "epoch": 12.096365802800118,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.012147587960425262,
      "loss": 0.8109,
      "num_input_tokens_seen": 47149504,
      "step": 81215
    },
    {
      "epoch": 12.097110515341079,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.01214567390705111,
      "loss": 0.8141,
      "num_input_tokens_seen": 47152544,
      "step": 81220
    },
    {
      "epoch": 12.097855227882038,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.012143759901897897,
      "loss": 0.8121,
      "num_input_tokens_seen": 47155680,
      "step": 81225
    },
    {
      "epoch": 12.098599940422996,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01214184594499796,
      "loss": 0.772,
      "num_input_tokens_seen": 47158784,
      "step": 81230
    },
    {
      "epoch": 12.099344652963955,
      "grad_norm": 0.060546875,
      "learning_rate": 0.012139932036383623,
      "loss": 0.7696,
      "num_input_tokens_seen": 47161536,
      "step": 81235
    },
    {
      "epoch": 12.100089365504916,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.012138018176087237,
      "loss": 0.7793,
      "num_input_tokens_seen": 47164448,
      "step": 81240
    },
    {
      "epoch": 12.100834078045875,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.012136104364141123,
      "loss": 0.7787,
      "num_input_tokens_seen": 47167424,
      "step": 81245
    },
    {
      "epoch": 12.101578790586833,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.012134190600577608,
      "loss": 0.7797,
      "num_input_tokens_seen": 47170336,
      "step": 81250
    },
    {
      "epoch": 12.102323503127792,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.012132276885429037,
      "loss": 0.8253,
      "num_input_tokens_seen": 47173152,
      "step": 81255
    },
    {
      "epoch": 12.103068215668753,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.01213036321872773,
      "loss": 0.7788,
      "num_input_tokens_seen": 47176192,
      "step": 81260
    },
    {
      "epoch": 12.103812928209711,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.012128449600506025,
      "loss": 0.7765,
      "num_input_tokens_seen": 47179104,
      "step": 81265
    },
    {
      "epoch": 12.10455764075067,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.012126536030796236,
      "loss": 0.7731,
      "num_input_tokens_seen": 47181792,
      "step": 81270
    },
    {
      "epoch": 12.105302353291629,
      "grad_norm": 0.064453125,
      "learning_rate": 0.01212462250963071,
      "loss": 0.7731,
      "num_input_tokens_seen": 47184640,
      "step": 81275
    },
    {
      "epoch": 12.10604706583259,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.012122709037041763,
      "loss": 0.8313,
      "num_input_tokens_seen": 47187936,
      "step": 81280
    },
    {
      "epoch": 12.106791778373548,
      "grad_norm": 0.044921875,
      "learning_rate": 0.012120795613061714,
      "loss": 0.8088,
      "num_input_tokens_seen": 47191200,
      "step": 81285
    },
    {
      "epoch": 12.107536490914507,
      "grad_norm": 0.033203125,
      "learning_rate": 0.012118882237722905,
      "loss": 0.7894,
      "num_input_tokens_seen": 47194112,
      "step": 81290
    },
    {
      "epoch": 12.108281203455466,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.01211696891105765,
      "loss": 0.8531,
      "num_input_tokens_seen": 47197248,
      "step": 81295
    },
    {
      "epoch": 12.109025915996426,
      "grad_norm": 0.05859375,
      "learning_rate": 0.012115055633098278,
      "loss": 0.8113,
      "num_input_tokens_seen": 47199968,
      "step": 81300
    },
    {
      "epoch": 12.109770628537385,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.012113142403877102,
      "loss": 0.8156,
      "num_input_tokens_seen": 47203136,
      "step": 81305
    },
    {
      "epoch": 12.110515341078344,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.012111229223426456,
      "loss": 0.8027,
      "num_input_tokens_seen": 47206176,
      "step": 81310
    },
    {
      "epoch": 12.111260053619302,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.01210931609177865,
      "loss": 0.7996,
      "num_input_tokens_seen": 47209088,
      "step": 81315
    },
    {
      "epoch": 12.112004766160263,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.012107403008966022,
      "loss": 0.7975,
      "num_input_tokens_seen": 47211936,
      "step": 81320
    },
    {
      "epoch": 12.112749478701222,
      "grad_norm": 0.046875,
      "learning_rate": 0.012105489975020875,
      "loss": 0.8054,
      "num_input_tokens_seen": 47215072,
      "step": 81325
    },
    {
      "epoch": 12.11349419124218,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.01210357698997553,
      "loss": 0.8015,
      "num_input_tokens_seen": 47217920,
      "step": 81330
    },
    {
      "epoch": 12.11423890378314,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.012101664053862314,
      "loss": 0.7892,
      "num_input_tokens_seen": 47220672,
      "step": 81335
    },
    {
      "epoch": 12.114983616324098,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.01209975116671353,
      "loss": 0.7935,
      "num_input_tokens_seen": 47223584,
      "step": 81340
    },
    {
      "epoch": 12.115728328865059,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.01209783832856151,
      "loss": 0.7914,
      "num_input_tokens_seen": 47226528,
      "step": 81345
    },
    {
      "epoch": 12.116473041406017,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.012095925539438553,
      "loss": 0.8445,
      "num_input_tokens_seen": 47229472,
      "step": 81350
    },
    {
      "epoch": 12.117217753946976,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.01209401279937699,
      "loss": 0.7967,
      "num_input_tokens_seen": 47232544,
      "step": 81355
    },
    {
      "epoch": 12.117962466487935,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.012092100108409127,
      "loss": 0.8093,
      "num_input_tokens_seen": 47235488,
      "step": 81360
    },
    {
      "epoch": 12.118707179028895,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.012090187466567275,
      "loss": 0.8229,
      "num_input_tokens_seen": 47238400,
      "step": 81365
    },
    {
      "epoch": 12.119451891569854,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.012088274873883751,
      "loss": 0.7993,
      "num_input_tokens_seen": 47241152,
      "step": 81370
    },
    {
      "epoch": 12.120196604110813,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.012086362330390859,
      "loss": 0.8002,
      "num_input_tokens_seen": 47244000,
      "step": 81375
    },
    {
      "epoch": 12.120941316651772,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.012084449836120916,
      "loss": 0.8128,
      "num_input_tokens_seen": 47246720,
      "step": 81380
    },
    {
      "epoch": 12.121686029192732,
      "grad_norm": 0.041015625,
      "learning_rate": 0.012082537391106228,
      "loss": 0.7921,
      "num_input_tokens_seen": 47249472,
      "step": 81385
    },
    {
      "epoch": 12.122430741733691,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.012080624995379109,
      "loss": 0.7993,
      "num_input_tokens_seen": 47252768,
      "step": 81390
    },
    {
      "epoch": 12.12317545427465,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.01207871264897186,
      "loss": 0.808,
      "num_input_tokens_seen": 47255872,
      "step": 81395
    },
    {
      "epoch": 12.123920166815608,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.012076800351916798,
      "loss": 0.8,
      "num_input_tokens_seen": 47258816,
      "step": 81400
    },
    {
      "epoch": 12.124664879356569,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.012074888104246223,
      "loss": 0.8057,
      "num_input_tokens_seen": 47262048,
      "step": 81405
    },
    {
      "epoch": 12.125409591897528,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.012072975905992432,
      "loss": 0.7955,
      "num_input_tokens_seen": 47264896,
      "step": 81410
    },
    {
      "epoch": 12.126154304438487,
      "grad_norm": 0.0546875,
      "learning_rate": 0.012071063757187745,
      "loss": 0.7836,
      "num_input_tokens_seen": 47267776,
      "step": 81415
    },
    {
      "epoch": 12.126899016979445,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.012069151657864452,
      "loss": 0.8157,
      "num_input_tokens_seen": 47270912,
      "step": 81420
    },
    {
      "epoch": 12.127643729520406,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.01206723960805487,
      "loss": 0.806,
      "num_input_tokens_seen": 47273888,
      "step": 81425
    },
    {
      "epoch": 12.128388442061365,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.012065327607791292,
      "loss": 0.7833,
      "num_input_tokens_seen": 47276672,
      "step": 81430
    },
    {
      "epoch": 12.129133154602323,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.012063415657106025,
      "loss": 0.7826,
      "num_input_tokens_seen": 47279424,
      "step": 81435
    },
    {
      "epoch": 12.129877867143282,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.012061503756031362,
      "loss": 0.79,
      "num_input_tokens_seen": 47282432,
      "step": 81440
    },
    {
      "epoch": 12.130622579684243,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.012059591904599605,
      "loss": 0.7952,
      "num_input_tokens_seen": 47285472,
      "step": 81445
    },
    {
      "epoch": 12.131367292225201,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.012057680102843056,
      "loss": 0.8012,
      "num_input_tokens_seen": 47288224,
      "step": 81450
    },
    {
      "epoch": 12.13211200476616,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.012055768350794008,
      "loss": 0.7645,
      "num_input_tokens_seen": 47290976,
      "step": 81455
    },
    {
      "epoch": 12.132856717307119,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.012053856648484766,
      "loss": 0.7992,
      "num_input_tokens_seen": 47294048,
      "step": 81460
    },
    {
      "epoch": 12.13360142984808,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.01205194499594762,
      "loss": 0.7834,
      "num_input_tokens_seen": 47296800,
      "step": 81465
    },
    {
      "epoch": 12.134346142389038,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.01205003339321487,
      "loss": 0.7825,
      "num_input_tokens_seen": 47299648,
      "step": 81470
    },
    {
      "epoch": 12.135090854929997,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0120481218403188,
      "loss": 0.8042,
      "num_input_tokens_seen": 47302528,
      "step": 81475
    },
    {
      "epoch": 12.135835567470956,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.012046210337291719,
      "loss": 0.7883,
      "num_input_tokens_seen": 47305280,
      "step": 81480
    },
    {
      "epoch": 12.136580280011916,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.012044298884165913,
      "loss": 0.8055,
      "num_input_tokens_seen": 47308032,
      "step": 81485
    },
    {
      "epoch": 12.137324992552875,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.012042387480973667,
      "loss": 0.7934,
      "num_input_tokens_seen": 47310944,
      "step": 81490
    },
    {
      "epoch": 12.138069705093834,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.012040476127747281,
      "loss": 0.7894,
      "num_input_tokens_seen": 47314336,
      "step": 81495
    },
    {
      "epoch": 12.138814417634793,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.012038564824519041,
      "loss": 0.7725,
      "num_input_tokens_seen": 47317312,
      "step": 81500
    },
    {
      "epoch": 12.139559130175751,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.012036653571321245,
      "loss": 0.7966,
      "num_input_tokens_seen": 47320480,
      "step": 81505
    },
    {
      "epoch": 12.140303842716712,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.012034742368186165,
      "loss": 0.8289,
      "num_input_tokens_seen": 47323648,
      "step": 81510
    },
    {
      "epoch": 12.14104855525767,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.012032831215146107,
      "loss": 0.7883,
      "num_input_tokens_seen": 47326688,
      "step": 81515
    },
    {
      "epoch": 12.14179326779863,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.01203092011223335,
      "loss": 0.7952,
      "num_input_tokens_seen": 47329568,
      "step": 81520
    },
    {
      "epoch": 12.142537980339588,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.012029009059480176,
      "loss": 0.801,
      "num_input_tokens_seen": 47332352,
      "step": 81525
    },
    {
      "epoch": 12.143282692880549,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01202709805691888,
      "loss": 0.8077,
      "num_input_tokens_seen": 47335200,
      "step": 81530
    },
    {
      "epoch": 12.144027405421507,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.012025187104581732,
      "loss": 0.7749,
      "num_input_tokens_seen": 47337984,
      "step": 81535
    },
    {
      "epoch": 12.144772117962466,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.012023276202501031,
      "loss": 0.8012,
      "num_input_tokens_seen": 47340864,
      "step": 81540
    },
    {
      "epoch": 12.145516830503425,
      "grad_norm": 0.044921875,
      "learning_rate": 0.012021365350709048,
      "loss": 0.7872,
      "num_input_tokens_seen": 47343584,
      "step": 81545
    },
    {
      "epoch": 12.146261543044385,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.012019454549238074,
      "loss": 0.8037,
      "num_input_tokens_seen": 47346368,
      "step": 81550
    },
    {
      "epoch": 12.147006255585344,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.012017543798120386,
      "loss": 0.8045,
      "num_input_tokens_seen": 47349120,
      "step": 81555
    },
    {
      "epoch": 12.147750968126303,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.012015633097388265,
      "loss": 0.8126,
      "num_input_tokens_seen": 47352032,
      "step": 81560
    },
    {
      "epoch": 12.148495680667262,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.012013722447073992,
      "loss": 0.7871,
      "num_input_tokens_seen": 47354688,
      "step": 81565
    },
    {
      "epoch": 12.149240393208222,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.012011811847209834,
      "loss": 0.8152,
      "num_input_tokens_seen": 47357568,
      "step": 81570
    },
    {
      "epoch": 12.149985105749181,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.012009901297828087,
      "loss": 0.7913,
      "num_input_tokens_seen": 47360960,
      "step": 81575
    },
    {
      "epoch": 12.15072981829014,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.01200799079896101,
      "loss": 0.8117,
      "num_input_tokens_seen": 47363712,
      "step": 81580
    },
    {
      "epoch": 12.151474530831099,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.012006080350640894,
      "loss": 0.7828,
      "num_input_tokens_seen": 47366432,
      "step": 81585
    },
    {
      "epoch": 12.152219243372059,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.012004169952900006,
      "loss": 0.8153,
      "num_input_tokens_seen": 47369088,
      "step": 81590
    },
    {
      "epoch": 12.152963955913018,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.012002259605770624,
      "loss": 0.7923,
      "num_input_tokens_seen": 47371936,
      "step": 81595
    },
    {
      "epoch": 12.153708668453977,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.012000349309285018,
      "loss": 0.784,
      "num_input_tokens_seen": 47374752,
      "step": 81600
    },
    {
      "epoch": 12.154453380994935,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.011998439063475457,
      "loss": 0.7952,
      "num_input_tokens_seen": 47377504,
      "step": 81605
    },
    {
      "epoch": 12.155198093535896,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.011996528868374224,
      "loss": 0.7985,
      "num_input_tokens_seen": 47380384,
      "step": 81610
    },
    {
      "epoch": 12.155942806076855,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.011994618724013577,
      "loss": 0.7923,
      "num_input_tokens_seen": 47383456,
      "step": 81615
    },
    {
      "epoch": 12.156687518617813,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.011992708630425796,
      "loss": 0.7784,
      "num_input_tokens_seen": 47386464,
      "step": 81620
    },
    {
      "epoch": 12.157432231158772,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.011990798587643145,
      "loss": 0.7831,
      "num_input_tokens_seen": 47389504,
      "step": 81625
    },
    {
      "epoch": 12.158176943699733,
      "grad_norm": 0.03125,
      "learning_rate": 0.011988888595697899,
      "loss": 0.7766,
      "num_input_tokens_seen": 47392192,
      "step": 81630
    },
    {
      "epoch": 12.158921656240691,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.01198697865462231,
      "loss": 0.777,
      "num_input_tokens_seen": 47395168,
      "step": 81635
    },
    {
      "epoch": 12.15966636878165,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.01198506876444866,
      "loss": 0.7566,
      "num_input_tokens_seen": 47398208,
      "step": 81640
    },
    {
      "epoch": 12.160411081322609,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.011983158925209212,
      "loss": 0.781,
      "num_input_tokens_seen": 47401216,
      "step": 81645
    },
    {
      "epoch": 12.16115579386357,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.011981249136936219,
      "loss": 0.778,
      "num_input_tokens_seen": 47404320,
      "step": 81650
    },
    {
      "epoch": 12.161900506404528,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.01197933939966196,
      "loss": 0.79,
      "num_input_tokens_seen": 47407296,
      "step": 81655
    },
    {
      "epoch": 12.162645218945487,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.011977429713418692,
      "loss": 0.8286,
      "num_input_tokens_seen": 47410048,
      "step": 81660
    },
    {
      "epoch": 12.163389931486446,
      "grad_norm": 0.046875,
      "learning_rate": 0.011975520078238676,
      "loss": 0.834,
      "num_input_tokens_seen": 47412896,
      "step": 81665
    },
    {
      "epoch": 12.164134644027406,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.01197361049415417,
      "loss": 0.7772,
      "num_input_tokens_seen": 47415840,
      "step": 81670
    },
    {
      "epoch": 12.164879356568365,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.011971700961197445,
      "loss": 0.7893,
      "num_input_tokens_seen": 47418528,
      "step": 81675
    },
    {
      "epoch": 12.165624069109324,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.011969791479400747,
      "loss": 0.7985,
      "num_input_tokens_seen": 47421376,
      "step": 81680
    },
    {
      "epoch": 12.166368781650283,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.01196788204879635,
      "loss": 0.8065,
      "num_input_tokens_seen": 47424768,
      "step": 81685
    },
    {
      "epoch": 12.167113494191241,
      "grad_norm": 0.052734375,
      "learning_rate": 0.011965972669416505,
      "loss": 0.8034,
      "num_input_tokens_seen": 47427744,
      "step": 81690
    },
    {
      "epoch": 12.167858206732202,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.011964063341293461,
      "loss": 0.8647,
      "num_input_tokens_seen": 47430464,
      "step": 81695
    },
    {
      "epoch": 12.16860291927316,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.011962154064459487,
      "loss": 0.7838,
      "num_input_tokens_seen": 47433184,
      "step": 81700
    },
    {
      "epoch": 12.16934763181412,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.011960244838946829,
      "loss": 0.8094,
      "num_input_tokens_seen": 47435968,
      "step": 81705
    },
    {
      "epoch": 12.170092344355078,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.011958335664787746,
      "loss": 0.842,
      "num_input_tokens_seen": 47438880,
      "step": 81710
    },
    {
      "epoch": 12.170837056896039,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.011956426542014488,
      "loss": 0.8036,
      "num_input_tokens_seen": 47441824,
      "step": 81715
    },
    {
      "epoch": 12.171581769436997,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.011954517470659315,
      "loss": 0.7945,
      "num_input_tokens_seen": 47444672,
      "step": 81720
    },
    {
      "epoch": 12.172326481977956,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.011952608450754474,
      "loss": 0.8048,
      "num_input_tokens_seen": 47447360,
      "step": 81725
    },
    {
      "epoch": 12.173071194518915,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.011950699482332213,
      "loss": 0.7878,
      "num_input_tokens_seen": 47450176,
      "step": 81730
    },
    {
      "epoch": 12.173815907059875,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.011948790565424788,
      "loss": 0.8386,
      "num_input_tokens_seen": 47453088,
      "step": 81735
    },
    {
      "epoch": 12.174560619600834,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.011946881700064438,
      "loss": 0.8075,
      "num_input_tokens_seen": 47455936,
      "step": 81740
    },
    {
      "epoch": 12.175305332141793,
      "grad_norm": 0.037109375,
      "learning_rate": 0.011944972886283428,
      "loss": 0.7768,
      "num_input_tokens_seen": 47458880,
      "step": 81745
    },
    {
      "epoch": 12.176050044682752,
      "grad_norm": 0.037109375,
      "learning_rate": 0.011943064124113988,
      "loss": 0.7853,
      "num_input_tokens_seen": 47462112,
      "step": 81750
    },
    {
      "epoch": 12.176794757223712,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.011941155413588379,
      "loss": 0.803,
      "num_input_tokens_seen": 47464832,
      "step": 81755
    },
    {
      "epoch": 12.177539469764671,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.011939246754738838,
      "loss": 0.8019,
      "num_input_tokens_seen": 47467808,
      "step": 81760
    },
    {
      "epoch": 12.17828418230563,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.011937338147597613,
      "loss": 0.7961,
      "num_input_tokens_seen": 47470752,
      "step": 81765
    },
    {
      "epoch": 12.179028894846589,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.01193542959219695,
      "loss": 0.7987,
      "num_input_tokens_seen": 47473696,
      "step": 81770
    },
    {
      "epoch": 12.179773607387549,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.011933521088569082,
      "loss": 0.7979,
      "num_input_tokens_seen": 47476864,
      "step": 81775
    },
    {
      "epoch": 12.180518319928508,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.011931612636746265,
      "loss": 0.8211,
      "num_input_tokens_seen": 47479744,
      "step": 81780
    },
    {
      "epoch": 12.181263032469467,
      "grad_norm": 0.037109375,
      "learning_rate": 0.011929704236760727,
      "loss": 0.806,
      "num_input_tokens_seen": 47482464,
      "step": 81785
    },
    {
      "epoch": 12.182007745010425,
      "grad_norm": 0.044921875,
      "learning_rate": 0.011927795888644722,
      "loss": 0.8041,
      "num_input_tokens_seen": 47485408,
      "step": 81790
    },
    {
      "epoch": 12.182752457551386,
      "grad_norm": 0.064453125,
      "learning_rate": 0.011925887592430481,
      "loss": 0.7706,
      "num_input_tokens_seen": 47488192,
      "step": 81795
    },
    {
      "epoch": 12.183497170092345,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.011923979348150246,
      "loss": 0.8079,
      "num_input_tokens_seen": 47491040,
      "step": 81800
    },
    {
      "epoch": 12.184241882633303,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.011922071155836255,
      "loss": 0.7954,
      "num_input_tokens_seen": 47493856,
      "step": 81805
    },
    {
      "epoch": 12.184986595174262,
      "grad_norm": 0.0174560546875,
      "learning_rate": 0.011920163015520737,
      "loss": 0.7937,
      "num_input_tokens_seen": 47496608,
      "step": 81810
    },
    {
      "epoch": 12.185731307715223,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.01191825492723594,
      "loss": 0.7916,
      "num_input_tokens_seen": 47499616,
      "step": 81815
    },
    {
      "epoch": 12.186476020256181,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.011916346891014092,
      "loss": 0.803,
      "num_input_tokens_seen": 47502528,
      "step": 81820
    },
    {
      "epoch": 12.18722073279714,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.01191443890688743,
      "loss": 0.7913,
      "num_input_tokens_seen": 47505376,
      "step": 81825
    },
    {
      "epoch": 12.187965445338099,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.011912530974888185,
      "loss": 0.7952,
      "num_input_tokens_seen": 47508448,
      "step": 81830
    },
    {
      "epoch": 12.18871015787906,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.011910623095048594,
      "loss": 0.8057,
      "num_input_tokens_seen": 47511296,
      "step": 81835
    },
    {
      "epoch": 12.189454870420018,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.01190871526740088,
      "loss": 0.8099,
      "num_input_tokens_seen": 47514176,
      "step": 81840
    },
    {
      "epoch": 12.190199582960977,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.011906807491977285,
      "loss": 0.7919,
      "num_input_tokens_seen": 47517216,
      "step": 81845
    },
    {
      "epoch": 12.190944295501936,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.011904899768810037,
      "loss": 0.8121,
      "num_input_tokens_seen": 47520256,
      "step": 81850
    },
    {
      "epoch": 12.191689008042895,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.011902992097931355,
      "loss": 0.8362,
      "num_input_tokens_seen": 47523008,
      "step": 81855
    },
    {
      "epoch": 12.192433720583855,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.01190108447937348,
      "loss": 0.7913,
      "num_input_tokens_seen": 47526048,
      "step": 81860
    },
    {
      "epoch": 12.193178433124814,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.011899176913168625,
      "loss": 0.8015,
      "num_input_tokens_seen": 47528928,
      "step": 81865
    },
    {
      "epoch": 12.193923145665773,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.01189726939934903,
      "loss": 0.7895,
      "num_input_tokens_seen": 47531904,
      "step": 81870
    },
    {
      "epoch": 12.194667858206731,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.01189536193794691,
      "loss": 0.8043,
      "num_input_tokens_seen": 47534848,
      "step": 81875
    },
    {
      "epoch": 12.195412570747692,
      "grad_norm": 0.029296875,
      "learning_rate": 0.0118934545289945,
      "loss": 0.8015,
      "num_input_tokens_seen": 47537824,
      "step": 81880
    },
    {
      "epoch": 12.19615728328865,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.011891547172524022,
      "loss": 0.8006,
      "num_input_tokens_seen": 47540768,
      "step": 81885
    },
    {
      "epoch": 12.19690199582961,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.011889639868567687,
      "loss": 0.7803,
      "num_input_tokens_seen": 47543968,
      "step": 81890
    },
    {
      "epoch": 12.197646708370568,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.011887732617157733,
      "loss": 0.7747,
      "num_input_tokens_seen": 47546592,
      "step": 81895
    },
    {
      "epoch": 12.198391420911529,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.011885825418326362,
      "loss": 0.7988,
      "num_input_tokens_seen": 47549728,
      "step": 81900
    },
    {
      "epoch": 12.199136133452487,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.011883918272105814,
      "loss": 0.7961,
      "num_input_tokens_seen": 47554112,
      "step": 81905
    },
    {
      "epoch": 12.199880845993446,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.011882011178528294,
      "loss": 0.7871,
      "num_input_tokens_seen": 47556992,
      "step": 81910
    },
    {
      "epoch": 12.200625558534405,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.011880104137626033,
      "loss": 0.8035,
      "num_input_tokens_seen": 47559808,
      "step": 81915
    },
    {
      "epoch": 12.201370271075366,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.011878197149431237,
      "loss": 0.7957,
      "num_input_tokens_seen": 47562976,
      "step": 81920
    },
    {
      "epoch": 12.202114983616324,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.011876290213976131,
      "loss": 0.7871,
      "num_input_tokens_seen": 47566016,
      "step": 81925
    },
    {
      "epoch": 12.202859696157283,
      "grad_norm": 0.046875,
      "learning_rate": 0.01187438333129293,
      "loss": 0.7855,
      "num_input_tokens_seen": 47569088,
      "step": 81930
    },
    {
      "epoch": 12.203604408698242,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.011872476501413834,
      "loss": 0.7999,
      "num_input_tokens_seen": 47572000,
      "step": 81935
    },
    {
      "epoch": 12.204349121239202,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.011870569724371078,
      "loss": 0.8219,
      "num_input_tokens_seen": 47574752,
      "step": 81940
    },
    {
      "epoch": 12.205093833780161,
      "grad_norm": 0.044921875,
      "learning_rate": 0.01186866300019686,
      "loss": 0.7995,
      "num_input_tokens_seen": 47577920,
      "step": 81945
    },
    {
      "epoch": 12.20583854632112,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.011866756328923403,
      "loss": 0.7963,
      "num_input_tokens_seen": 47580992,
      "step": 81950
    },
    {
      "epoch": 12.206583258862079,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.011864849710582912,
      "loss": 0.7985,
      "num_input_tokens_seen": 47583808,
      "step": 81955
    },
    {
      "epoch": 12.20732797140304,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.011862943145207602,
      "loss": 0.8075,
      "num_input_tokens_seen": 47586784,
      "step": 81960
    },
    {
      "epoch": 12.208072683943998,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.011861036632829678,
      "loss": 0.7953,
      "num_input_tokens_seen": 47589728,
      "step": 81965
    },
    {
      "epoch": 12.208817396484957,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.011859130173481343,
      "loss": 0.8033,
      "num_input_tokens_seen": 47592672,
      "step": 81970
    },
    {
      "epoch": 12.209562109025915,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.011857223767194818,
      "loss": 0.7957,
      "num_input_tokens_seen": 47595808,
      "step": 81975
    },
    {
      "epoch": 12.210306821566876,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.011855317414002298,
      "loss": 0.8024,
      "num_input_tokens_seen": 47598816,
      "step": 81980
    },
    {
      "epoch": 12.211051534107835,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.011853411113935999,
      "loss": 0.8123,
      "num_input_tokens_seen": 47601952,
      "step": 81985
    },
    {
      "epoch": 12.211796246648793,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.011851504867028118,
      "loss": 0.8436,
      "num_input_tokens_seen": 47604736,
      "step": 81990
    },
    {
      "epoch": 12.212540959189752,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.011849598673310867,
      "loss": 0.7872,
      "num_input_tokens_seen": 47607680,
      "step": 81995
    },
    {
      "epoch": 12.213285671730713,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.011847692532816438,
      "loss": 0.8035,
      "num_input_tokens_seen": 47610560,
      "step": 82000
    },
    {
      "epoch": 12.214030384271672,
      "grad_norm": 0.056640625,
      "learning_rate": 0.011845786445577045,
      "loss": 0.7923,
      "num_input_tokens_seen": 47613376,
      "step": 82005
    },
    {
      "epoch": 12.21477509681263,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.011843880411624885,
      "loss": 0.7994,
      "num_input_tokens_seen": 47616128,
      "step": 82010
    },
    {
      "epoch": 12.215519809353589,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.011841974430992152,
      "loss": 0.7819,
      "num_input_tokens_seen": 47619136,
      "step": 82015
    },
    {
      "epoch": 12.216264521894548,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.011840068503711056,
      "loss": 0.7957,
      "num_input_tokens_seen": 47622080,
      "step": 82020
    },
    {
      "epoch": 12.217009234435508,
      "grad_norm": 0.041015625,
      "learning_rate": 0.011838162629813789,
      "loss": 0.7942,
      "num_input_tokens_seen": 47624832,
      "step": 82025
    },
    {
      "epoch": 12.217753946976467,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.011836256809332554,
      "loss": 0.7995,
      "num_input_tokens_seen": 47627712,
      "step": 82030
    },
    {
      "epoch": 12.218498659517426,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.011834351042299537,
      "loss": 0.7913,
      "num_input_tokens_seen": 47630688,
      "step": 82035
    },
    {
      "epoch": 12.219243372058385,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.01183244532874695,
      "loss": 0.8099,
      "num_input_tokens_seen": 47633824,
      "step": 82040
    },
    {
      "epoch": 12.219988084599345,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.01183053966870698,
      "loss": 0.7823,
      "num_input_tokens_seen": 47636576,
      "step": 82045
    },
    {
      "epoch": 12.220732797140304,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.011828634062211813,
      "loss": 0.7864,
      "num_input_tokens_seen": 47639232,
      "step": 82050
    },
    {
      "epoch": 12.221477509681263,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.011826728509293656,
      "loss": 0.7938,
      "num_input_tokens_seen": 47642240,
      "step": 82055
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.011824823009984695,
      "loss": 0.7783,
      "num_input_tokens_seen": 47644960,
      "step": 82060
    },
    {
      "epoch": 12.222966934763182,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.011822917564317123,
      "loss": 0.7957,
      "num_input_tokens_seen": 47647904,
      "step": 82065
    },
    {
      "epoch": 12.22371164730414,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.011821012172323127,
      "loss": 0.8157,
      "num_input_tokens_seen": 47650752,
      "step": 82070
    },
    {
      "epoch": 12.2244563598451,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.011819106834034903,
      "loss": 0.7931,
      "num_input_tokens_seen": 47653952,
      "step": 82075
    },
    {
      "epoch": 12.225201072386058,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.011817201549484629,
      "loss": 0.7809,
      "num_input_tokens_seen": 47656832,
      "step": 82080
    },
    {
      "epoch": 12.225945784927019,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.011815296318704507,
      "loss": 0.8034,
      "num_input_tokens_seen": 47659584,
      "step": 82085
    },
    {
      "epoch": 12.226690497467978,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.011813391141726717,
      "loss": 0.7853,
      "num_input_tokens_seen": 47662400,
      "step": 82090
    },
    {
      "epoch": 12.227435210008936,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.011811486018583442,
      "loss": 0.8713,
      "num_input_tokens_seen": 47665600,
      "step": 82095
    },
    {
      "epoch": 12.228179922549895,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.011809580949306873,
      "loss": 0.8088,
      "num_input_tokens_seen": 47668768,
      "step": 82100
    },
    {
      "epoch": 12.228924635090856,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.011807675933929184,
      "loss": 0.8156,
      "num_input_tokens_seen": 47671616,
      "step": 82105
    },
    {
      "epoch": 12.229669347631814,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.011805770972482574,
      "loss": 0.8139,
      "num_input_tokens_seen": 47674432,
      "step": 82110
    },
    {
      "epoch": 12.230414060172773,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.011803866064999209,
      "loss": 0.784,
      "num_input_tokens_seen": 47677152,
      "step": 82115
    },
    {
      "epoch": 12.231158772713732,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.011801961211511286,
      "loss": 0.7921,
      "num_input_tokens_seen": 47680160,
      "step": 82120
    },
    {
      "epoch": 12.231903485254692,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.011800056412050976,
      "loss": 0.7976,
      "num_input_tokens_seen": 47682848,
      "step": 82125
    },
    {
      "epoch": 12.232648197795651,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.01179815166665046,
      "loss": 0.7912,
      "num_input_tokens_seen": 47685728,
      "step": 82130
    },
    {
      "epoch": 12.23339291033661,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.011796246975341918,
      "loss": 0.7967,
      "num_input_tokens_seen": 47688384,
      "step": 82135
    },
    {
      "epoch": 12.234137622877569,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.011794342338157524,
      "loss": 0.8025,
      "num_input_tokens_seen": 47691136,
      "step": 82140
    },
    {
      "epoch": 12.23488233541853,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.011792437755129462,
      "loss": 0.7908,
      "num_input_tokens_seen": 47694144,
      "step": 82145
    },
    {
      "epoch": 12.235627047959488,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0117905332262899,
      "loss": 0.791,
      "num_input_tokens_seen": 47697408,
      "step": 82150
    },
    {
      "epoch": 12.236371760500447,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01178862875167102,
      "loss": 0.8305,
      "num_input_tokens_seen": 47700384,
      "step": 82155
    },
    {
      "epoch": 12.237116473041405,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.011786724331304992,
      "loss": 0.8005,
      "num_input_tokens_seen": 47704032,
      "step": 82160
    },
    {
      "epoch": 12.237861185582366,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.011784819965223993,
      "loss": 0.811,
      "num_input_tokens_seen": 47706880,
      "step": 82165
    },
    {
      "epoch": 12.238605898123325,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.011782915653460195,
      "loss": 0.8036,
      "num_input_tokens_seen": 47709696,
      "step": 82170
    },
    {
      "epoch": 12.239350610664284,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01178101139604576,
      "loss": 0.7969,
      "num_input_tokens_seen": 47712448,
      "step": 82175
    },
    {
      "epoch": 12.240095323205242,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.011779107193012872,
      "loss": 0.8001,
      "num_input_tokens_seen": 47715328,
      "step": 82180
    },
    {
      "epoch": 12.240840035746203,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.011777203044393692,
      "loss": 0.796,
      "num_input_tokens_seen": 47718272,
      "step": 82185
    },
    {
      "epoch": 12.241584748287162,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.011775298950220392,
      "loss": 0.7725,
      "num_input_tokens_seen": 47721248,
      "step": 82190
    },
    {
      "epoch": 12.24232946082812,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.011773394910525134,
      "loss": 0.7716,
      "num_input_tokens_seen": 47724000,
      "step": 82195
    },
    {
      "epoch": 12.243074173369079,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.011771490925340095,
      "loss": 0.7865,
      "num_input_tokens_seen": 47726848,
      "step": 82200
    },
    {
      "epoch": 12.243818885910038,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.011769586994697438,
      "loss": 0.7848,
      "num_input_tokens_seen": 47730048,
      "step": 82205
    },
    {
      "epoch": 12.244563598450998,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.011767683118629317,
      "loss": 0.8251,
      "num_input_tokens_seen": 47732608,
      "step": 82210
    },
    {
      "epoch": 12.245308310991957,
      "grad_norm": 0.02734375,
      "learning_rate": 0.01176577929716791,
      "loss": 0.794,
      "num_input_tokens_seen": 47735840,
      "step": 82215
    },
    {
      "epoch": 12.246053023532916,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.011763875530345375,
      "loss": 0.7926,
      "num_input_tokens_seen": 47738816,
      "step": 82220
    },
    {
      "epoch": 12.246797736073875,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.011761971818193875,
      "loss": 0.7944,
      "num_input_tokens_seen": 47741952,
      "step": 82225
    },
    {
      "epoch": 12.247542448614835,
      "grad_norm": 0.05078125,
      "learning_rate": 0.011760068160745563,
      "loss": 0.7979,
      "num_input_tokens_seen": 47744928,
      "step": 82230
    },
    {
      "epoch": 12.248287161155794,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.011758164558032615,
      "loss": 0.7865,
      "num_input_tokens_seen": 47747616,
      "step": 82235
    },
    {
      "epoch": 12.249031873696753,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.011756261010087173,
      "loss": 0.8043,
      "num_input_tokens_seen": 47750656,
      "step": 82240
    },
    {
      "epoch": 12.249776586237711,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.011754357516941412,
      "loss": 0.7956,
      "num_input_tokens_seen": 47753504,
      "step": 82245
    },
    {
      "epoch": 12.250521298778672,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.011752454078627482,
      "loss": 0.7864,
      "num_input_tokens_seen": 47756256,
      "step": 82250
    },
    {
      "epoch": 12.25126601131963,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.011750550695177536,
      "loss": 0.7934,
      "num_input_tokens_seen": 47759200,
      "step": 82255
    },
    {
      "epoch": 12.25201072386059,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.011748647366623738,
      "loss": 0.7836,
      "num_input_tokens_seen": 47762240,
      "step": 82260
    },
    {
      "epoch": 12.252755436401548,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.01174674409299823,
      "loss": 0.806,
      "num_input_tokens_seen": 47765408,
      "step": 82265
    },
    {
      "epoch": 12.253500148942509,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.011744840874333182,
      "loss": 0.8063,
      "num_input_tokens_seen": 47768224,
      "step": 82270
    },
    {
      "epoch": 12.254244861483468,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.011742937710660732,
      "loss": 0.8048,
      "num_input_tokens_seen": 47771296,
      "step": 82275
    },
    {
      "epoch": 12.254989574024426,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.011741034602013044,
      "loss": 0.7881,
      "num_input_tokens_seen": 47774432,
      "step": 82280
    },
    {
      "epoch": 12.255734286565385,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.011739131548422262,
      "loss": 0.8076,
      "num_input_tokens_seen": 47777216,
      "step": 82285
    },
    {
      "epoch": 12.256478999106346,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.011737228549920544,
      "loss": 0.7774,
      "num_input_tokens_seen": 47779872,
      "step": 82290
    },
    {
      "epoch": 12.257223711647304,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01173532560654003,
      "loss": 0.8041,
      "num_input_tokens_seen": 47782720,
      "step": 82295
    },
    {
      "epoch": 12.257968424188263,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.011733422718312866,
      "loss": 0.7998,
      "num_input_tokens_seen": 47785280,
      "step": 82300
    },
    {
      "epoch": 12.258713136729222,
      "grad_norm": 0.044921875,
      "learning_rate": 0.011731519885271213,
      "loss": 0.7754,
      "num_input_tokens_seen": 47788096,
      "step": 82305
    },
    {
      "epoch": 12.259457849270182,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.011729617107447203,
      "loss": 0.8103,
      "num_input_tokens_seen": 47791232,
      "step": 82310
    },
    {
      "epoch": 12.260202561811141,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.011727714384872995,
      "loss": 0.7888,
      "num_input_tokens_seen": 47794336,
      "step": 82315
    },
    {
      "epoch": 12.2609472743521,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.011725811717580721,
      "loss": 0.7945,
      "num_input_tokens_seen": 47797280,
      "step": 82320
    },
    {
      "epoch": 12.261691986893059,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.011723909105602535,
      "loss": 0.8307,
      "num_input_tokens_seen": 47800224,
      "step": 82325
    },
    {
      "epoch": 12.26243669943402,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.011722006548970577,
      "loss": 0.8292,
      "num_input_tokens_seen": 47802880,
      "step": 82330
    },
    {
      "epoch": 12.263181411974978,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.011720104047716976,
      "loss": 0.7956,
      "num_input_tokens_seen": 47805664,
      "step": 82335
    },
    {
      "epoch": 12.263926124515937,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.011718201601873892,
      "loss": 0.8464,
      "num_input_tokens_seen": 47808384,
      "step": 82340
    },
    {
      "epoch": 12.264670837056896,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.011716299211473453,
      "loss": 0.7893,
      "num_input_tokens_seen": 47811296,
      "step": 82345
    },
    {
      "epoch": 12.265415549597856,
      "grad_norm": 0.037109375,
      "learning_rate": 0.011714396876547805,
      "loss": 0.8009,
      "num_input_tokens_seen": 47814048,
      "step": 82350
    },
    {
      "epoch": 12.266160262138815,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.011712494597129077,
      "loss": 0.7931,
      "num_input_tokens_seen": 47816800,
      "step": 82355
    },
    {
      "epoch": 12.266904974679774,
      "grad_norm": 0.021484375,
      "learning_rate": 0.011710592373249418,
      "loss": 0.8091,
      "num_input_tokens_seen": 47819456,
      "step": 82360
    },
    {
      "epoch": 12.267649687220732,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.011708690204940952,
      "loss": 0.7942,
      "num_input_tokens_seen": 47822176,
      "step": 82365
    },
    {
      "epoch": 12.268394399761693,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.011706788092235822,
      "loss": 0.8009,
      "num_input_tokens_seen": 47825088,
      "step": 82370
    },
    {
      "epoch": 12.269139112302652,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.011704886035166165,
      "loss": 0.8019,
      "num_input_tokens_seen": 47827840,
      "step": 82375
    },
    {
      "epoch": 12.26988382484361,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.011702984033764099,
      "loss": 0.7976,
      "num_input_tokens_seen": 47830912,
      "step": 82380
    },
    {
      "epoch": 12.27062853738457,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.011701082088061776,
      "loss": 0.8104,
      "num_input_tokens_seen": 47833984,
      "step": 82385
    },
    {
      "epoch": 12.271373249925528,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.011699180198091316,
      "loss": 0.8017,
      "num_input_tokens_seen": 47836736,
      "step": 82390
    },
    {
      "epoch": 12.272117962466488,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.011697278363884851,
      "loss": 0.8075,
      "num_input_tokens_seen": 47839552,
      "step": 82395
    },
    {
      "epoch": 12.272862675007447,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.011695376585474508,
      "loss": 0.8078,
      "num_input_tokens_seen": 47842304,
      "step": 82400
    },
    {
      "epoch": 12.273607387548406,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.011693474862892423,
      "loss": 0.7991,
      "num_input_tokens_seen": 47845184,
      "step": 82405
    },
    {
      "epoch": 12.274352100089365,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.011691573196170722,
      "loss": 0.8117,
      "num_input_tokens_seen": 47848160,
      "step": 82410
    },
    {
      "epoch": 12.275096812630325,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01168967158534152,
      "loss": 0.8032,
      "num_input_tokens_seen": 47851360,
      "step": 82415
    },
    {
      "epoch": 12.275841525171284,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.011687770030436962,
      "loss": 0.7859,
      "num_input_tokens_seen": 47854176,
      "step": 82420
    },
    {
      "epoch": 12.276586237712243,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.011685868531489157,
      "loss": 0.8015,
      "num_input_tokens_seen": 47857088,
      "step": 82425
    },
    {
      "epoch": 12.277330950253202,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.011683967088530239,
      "loss": 0.7982,
      "num_input_tokens_seen": 47859712,
      "step": 82430
    },
    {
      "epoch": 12.278075662794162,
      "grad_norm": 0.052734375,
      "learning_rate": 0.011682065701592323,
      "loss": 0.813,
      "num_input_tokens_seen": 47862880,
      "step": 82435
    },
    {
      "epoch": 12.27882037533512,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.011680164370707537,
      "loss": 0.8114,
      "num_input_tokens_seen": 47866080,
      "step": 82440
    },
    {
      "epoch": 12.27956508787608,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.011678263095907995,
      "loss": 0.7939,
      "num_input_tokens_seen": 47868832,
      "step": 82445
    },
    {
      "epoch": 12.280309800417038,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.011676361877225828,
      "loss": 0.786,
      "num_input_tokens_seen": 47871648,
      "step": 82450
    },
    {
      "epoch": 12.281054512957999,
      "grad_norm": 0.14453125,
      "learning_rate": 0.011674460714693148,
      "loss": 0.7578,
      "num_input_tokens_seen": 47874592,
      "step": 82455
    },
    {
      "epoch": 12.281799225498958,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.011672559608342074,
      "loss": 0.7962,
      "num_input_tokens_seen": 47877952,
      "step": 82460
    },
    {
      "epoch": 12.282543938039916,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.011670658558204726,
      "loss": 0.8281,
      "num_input_tokens_seen": 47881152,
      "step": 82465
    },
    {
      "epoch": 12.283288650580875,
      "grad_norm": 0.0390625,
      "learning_rate": 0.01166875756431321,
      "loss": 0.7907,
      "num_input_tokens_seen": 47883904,
      "step": 82470
    },
    {
      "epoch": 12.284033363121836,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.011666856626699657,
      "loss": 0.7978,
      "num_input_tokens_seen": 47886880,
      "step": 82475
    },
    {
      "epoch": 12.284778075662794,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.01166495574539617,
      "loss": 0.7983,
      "num_input_tokens_seen": 47890048,
      "step": 82480
    },
    {
      "epoch": 12.285522788203753,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.011663054920434866,
      "loss": 0.788,
      "num_input_tokens_seen": 47892960,
      "step": 82485
    },
    {
      "epoch": 12.286267500744712,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01166115415184786,
      "loss": 0.7914,
      "num_input_tokens_seen": 47895872,
      "step": 82490
    },
    {
      "epoch": 12.287012213285673,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.011659253439667259,
      "loss": 0.7843,
      "num_input_tokens_seen": 47898912,
      "step": 82495
    },
    {
      "epoch": 12.287756925826631,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.011657352783925178,
      "loss": 0.8001,
      "num_input_tokens_seen": 47901760,
      "step": 82500
    },
    {
      "epoch": 12.28850163836759,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01165545218465372,
      "loss": 0.7806,
      "num_input_tokens_seen": 47904640,
      "step": 82505
    },
    {
      "epoch": 12.289246350908549,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.011653551641885002,
      "loss": 0.8148,
      "num_input_tokens_seen": 47907680,
      "step": 82510
    },
    {
      "epoch": 12.28999106344951,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.011651651155651126,
      "loss": 0.7982,
      "num_input_tokens_seen": 47910848,
      "step": 82515
    },
    {
      "epoch": 12.290735775990468,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.011649750725984204,
      "loss": 0.8017,
      "num_input_tokens_seen": 47913792,
      "step": 82520
    },
    {
      "epoch": 12.291480488531427,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.011647850352916331,
      "loss": 0.8209,
      "num_input_tokens_seen": 47916672,
      "step": 82525
    },
    {
      "epoch": 12.292225201072386,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.011645950036479627,
      "loss": 0.7972,
      "num_input_tokens_seen": 47919712,
      "step": 82530
    },
    {
      "epoch": 12.292969913613344,
      "grad_norm": 0.04296875,
      "learning_rate": 0.011644049776706186,
      "loss": 0.7991,
      "num_input_tokens_seen": 47922592,
      "step": 82535
    },
    {
      "epoch": 12.293714626154305,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.01164214957362811,
      "loss": 0.7788,
      "num_input_tokens_seen": 47925376,
      "step": 82540
    },
    {
      "epoch": 12.294459338695264,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.011640249427277509,
      "loss": 0.7841,
      "num_input_tokens_seen": 47928256,
      "step": 82545
    },
    {
      "epoch": 12.295204051236222,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.011638349337686475,
      "loss": 0.7821,
      "num_input_tokens_seen": 47931072,
      "step": 82550
    },
    {
      "epoch": 12.295948763777181,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.011636449304887116,
      "loss": 0.7925,
      "num_input_tokens_seen": 47933760,
      "step": 82555
    },
    {
      "epoch": 12.296693476318142,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.01163454932891152,
      "loss": 0.8245,
      "num_input_tokens_seen": 47936832,
      "step": 82560
    },
    {
      "epoch": 12.2974381888591,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.011632649409791803,
      "loss": 0.7965,
      "num_input_tokens_seen": 47939872,
      "step": 82565
    },
    {
      "epoch": 12.29818290140006,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.011630749547560049,
      "loss": 0.7872,
      "num_input_tokens_seen": 47943072,
      "step": 82570
    },
    {
      "epoch": 12.298927613941018,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.01162884974224835,
      "loss": 0.7952,
      "num_input_tokens_seen": 47945888,
      "step": 82575
    },
    {
      "epoch": 12.299672326481979,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.011626949993888816,
      "loss": 0.7932,
      "num_input_tokens_seen": 47948864,
      "step": 82580
    },
    {
      "epoch": 12.300417039022937,
      "grad_norm": 0.05078125,
      "learning_rate": 0.01162505030251353,
      "loss": 0.7965,
      "num_input_tokens_seen": 47951648,
      "step": 82585
    },
    {
      "epoch": 12.301161751563896,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.011623150668154593,
      "loss": 0.8059,
      "num_input_tokens_seen": 47954560,
      "step": 82590
    },
    {
      "epoch": 12.301906464104855,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.011621251090844085,
      "loss": 0.8033,
      "num_input_tokens_seen": 47957536,
      "step": 82595
    },
    {
      "epoch": 12.302651176645815,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.011619351570614113,
      "loss": 0.7944,
      "num_input_tokens_seen": 47960256,
      "step": 82600
    },
    {
      "epoch": 12.303395889186774,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.011617452107496757,
      "loss": 0.7919,
      "num_input_tokens_seen": 47963008,
      "step": 82605
    },
    {
      "epoch": 12.304140601727733,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.011615552701524113,
      "loss": 0.8049,
      "num_input_tokens_seen": 47965824,
      "step": 82610
    },
    {
      "epoch": 12.304885314268692,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.011613653352728265,
      "loss": 0.7945,
      "num_input_tokens_seen": 47968576,
      "step": 82615
    },
    {
      "epoch": 12.305630026809652,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0116117540611413,
      "loss": 0.7816,
      "num_input_tokens_seen": 47971712,
      "step": 82620
    },
    {
      "epoch": 12.30637473935061,
      "grad_norm": 0.041015625,
      "learning_rate": 0.011609854826795312,
      "loss": 0.7992,
      "num_input_tokens_seen": 47974464,
      "step": 82625
    },
    {
      "epoch": 12.30711945189157,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.011607955649722373,
      "loss": 0.8108,
      "num_input_tokens_seen": 47977408,
      "step": 82630
    },
    {
      "epoch": 12.307864164432528,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.01160605652995458,
      "loss": 0.7889,
      "num_input_tokens_seen": 47980320,
      "step": 82635
    },
    {
      "epoch": 12.308608876973489,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.011604157467524011,
      "loss": 0.837,
      "num_input_tokens_seen": 47983328,
      "step": 82640
    },
    {
      "epoch": 12.309353589514448,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.011602258462462753,
      "loss": 0.8161,
      "num_input_tokens_seen": 47986240,
      "step": 82645
    },
    {
      "epoch": 12.310098302055406,
      "grad_norm": 0.041015625,
      "learning_rate": 0.011600359514802885,
      "loss": 0.7841,
      "num_input_tokens_seen": 47989312,
      "step": 82650
    },
    {
      "epoch": 12.310843014596365,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.011598460624576486,
      "loss": 0.7785,
      "num_input_tokens_seen": 47992224,
      "step": 82655
    },
    {
      "epoch": 12.311587727137326,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.011596561791815639,
      "loss": 0.7759,
      "num_input_tokens_seen": 47995136,
      "step": 82660
    },
    {
      "epoch": 12.312332439678285,
      "grad_norm": 0.033203125,
      "learning_rate": 0.011594663016552417,
      "loss": 0.7751,
      "num_input_tokens_seen": 47998016,
      "step": 82665
    },
    {
      "epoch": 12.313077152219243,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.011592764298818907,
      "loss": 0.8124,
      "num_input_tokens_seen": 48000864,
      "step": 82670
    },
    {
      "epoch": 12.313821864760202,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.011590865638647177,
      "loss": 0.7742,
      "num_input_tokens_seen": 48004000,
      "step": 82675
    },
    {
      "epoch": 12.314566577301163,
      "grad_norm": 0.041015625,
      "learning_rate": 0.011588967036069311,
      "loss": 0.7922,
      "num_input_tokens_seen": 48006848,
      "step": 82680
    },
    {
      "epoch": 12.315311289842121,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.011587068491117378,
      "loss": 0.7998,
      "num_input_tokens_seen": 48009728,
      "step": 82685
    },
    {
      "epoch": 12.31605600238308,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.011585170003823458,
      "loss": 0.7761,
      "num_input_tokens_seen": 48012480,
      "step": 82690
    },
    {
      "epoch": 12.316800714924039,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.011583271574219619,
      "loss": 0.7896,
      "num_input_tokens_seen": 48015296,
      "step": 82695
    },
    {
      "epoch": 12.317545427465,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.011581373202337928,
      "loss": 0.8,
      "num_input_tokens_seen": 48018048,
      "step": 82700
    },
    {
      "epoch": 12.318290140005958,
      "grad_norm": 0.04296875,
      "learning_rate": 0.011579474888210468,
      "loss": 0.7867,
      "num_input_tokens_seen": 48020928,
      "step": 82705
    },
    {
      "epoch": 12.319034852546917,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.011577576631869299,
      "loss": 0.7945,
      "num_input_tokens_seen": 48023616,
      "step": 82710
    },
    {
      "epoch": 12.319779565087876,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.011575678433346498,
      "loss": 0.8221,
      "num_input_tokens_seen": 48026656,
      "step": 82715
    },
    {
      "epoch": 12.320524277628834,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01157378029267413,
      "loss": 0.7814,
      "num_input_tokens_seen": 48029440,
      "step": 82720
    },
    {
      "epoch": 12.321268990169795,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.011571882209884263,
      "loss": 0.7877,
      "num_input_tokens_seen": 48032512,
      "step": 82725
    },
    {
      "epoch": 12.322013702710754,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.011569984185008961,
      "loss": 0.7735,
      "num_input_tokens_seen": 48035392,
      "step": 82730
    },
    {
      "epoch": 12.322758415251712,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.011568086218080283,
      "loss": 0.8028,
      "num_input_tokens_seen": 48038336,
      "step": 82735
    },
    {
      "epoch": 12.323503127792671,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.011566188309130308,
      "loss": 0.7925,
      "num_input_tokens_seen": 48041600,
      "step": 82740
    },
    {
      "epoch": 12.324247840333632,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.011564290458191085,
      "loss": 0.8123,
      "num_input_tokens_seen": 48044448,
      "step": 82745
    },
    {
      "epoch": 12.32499255287459,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.011562392665294692,
      "loss": 0.805,
      "num_input_tokens_seen": 48047168,
      "step": 82750
    },
    {
      "epoch": 12.32573726541555,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.011560494930473173,
      "loss": 0.8033,
      "num_input_tokens_seen": 48050016,
      "step": 82755
    },
    {
      "epoch": 12.326481977956508,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.011558597253758601,
      "loss": 0.7877,
      "num_input_tokens_seen": 48052992,
      "step": 82760
    },
    {
      "epoch": 12.327226690497469,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.011556699635183028,
      "loss": 0.8173,
      "num_input_tokens_seen": 48055776,
      "step": 82765
    },
    {
      "epoch": 12.327971403038427,
      "grad_norm": 0.0234375,
      "learning_rate": 0.011554802074778516,
      "loss": 0.7917,
      "num_input_tokens_seen": 48058720,
      "step": 82770
    },
    {
      "epoch": 12.328716115579386,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.011552904572577126,
      "loss": 0.8143,
      "num_input_tokens_seen": 48061600,
      "step": 82775
    },
    {
      "epoch": 12.329460828120345,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.011551007128610902,
      "loss": 0.7915,
      "num_input_tokens_seen": 48064352,
      "step": 82780
    },
    {
      "epoch": 12.330205540661305,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.011549109742911911,
      "loss": 0.8085,
      "num_input_tokens_seen": 48067040,
      "step": 82785
    },
    {
      "epoch": 12.330950253202264,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.011547212415512204,
      "loss": 0.7854,
      "num_input_tokens_seen": 48070208,
      "step": 82790
    },
    {
      "epoch": 12.331694965743223,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.011545315146443835,
      "loss": 0.8039,
      "num_input_tokens_seen": 48073280,
      "step": 82795
    },
    {
      "epoch": 12.332439678284182,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.01154341793573885,
      "loss": 0.7888,
      "num_input_tokens_seen": 48076096,
      "step": 82800
    },
    {
      "epoch": 12.333184390825142,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.011541520783429312,
      "loss": 0.8384,
      "num_input_tokens_seen": 48078944,
      "step": 82805
    },
    {
      "epoch": 12.333929103366101,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.011539623689547264,
      "loss": 0.7921,
      "num_input_tokens_seen": 48081952,
      "step": 82810
    },
    {
      "epoch": 12.33467381590706,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.01153772665412475,
      "loss": 0.8015,
      "num_input_tokens_seen": 48084768,
      "step": 82815
    },
    {
      "epoch": 12.335418528448018,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.011535829677193837,
      "loss": 0.805,
      "num_input_tokens_seen": 48087584,
      "step": 82820
    },
    {
      "epoch": 12.336163240988979,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.011533932758786551,
      "loss": 0.7665,
      "num_input_tokens_seen": 48090656,
      "step": 82825
    },
    {
      "epoch": 12.336907953529938,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.011532035898934955,
      "loss": 0.821,
      "num_input_tokens_seen": 48093696,
      "step": 82830
    },
    {
      "epoch": 12.337652666070897,
      "grad_norm": 0.03125,
      "learning_rate": 0.01153013909767108,
      "loss": 0.7883,
      "num_input_tokens_seen": 48096448,
      "step": 82835
    },
    {
      "epoch": 12.338397378611855,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.011528242355026986,
      "loss": 0.7918,
      "num_input_tokens_seen": 48099712,
      "step": 82840
    },
    {
      "epoch": 12.339142091152816,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.011526345671034706,
      "loss": 0.7926,
      "num_input_tokens_seen": 48102592,
      "step": 82845
    },
    {
      "epoch": 12.339886803693775,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.011524449045726286,
      "loss": 0.8276,
      "num_input_tokens_seen": 48105632,
      "step": 82850
    },
    {
      "epoch": 12.340631516234733,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.011522552479133766,
      "loss": 0.806,
      "num_input_tokens_seen": 48108768,
      "step": 82855
    },
    {
      "epoch": 12.341376228775692,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.011520655971289187,
      "loss": 0.806,
      "num_input_tokens_seen": 48111872,
      "step": 82860
    },
    {
      "epoch": 12.342120941316653,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.011518759522224593,
      "loss": 0.7716,
      "num_input_tokens_seen": 48114976,
      "step": 82865
    },
    {
      "epoch": 12.342865653857611,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.011516863131972012,
      "loss": 0.7949,
      "num_input_tokens_seen": 48117728,
      "step": 82870
    },
    {
      "epoch": 12.34361036639857,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.011514966800563494,
      "loss": 0.7849,
      "num_input_tokens_seen": 48120576,
      "step": 82875
    },
    {
      "epoch": 12.344355078939529,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.011513070528031065,
      "loss": 0.7987,
      "num_input_tokens_seen": 48123456,
      "step": 82880
    },
    {
      "epoch": 12.34509979148049,
      "grad_norm": 0.03125,
      "learning_rate": 0.011511174314406774,
      "loss": 0.7973,
      "num_input_tokens_seen": 48126400,
      "step": 82885
    },
    {
      "epoch": 12.345844504021448,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.011509278159722636,
      "loss": 0.7791,
      "num_input_tokens_seen": 48129280,
      "step": 82890
    },
    {
      "epoch": 12.346589216562407,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.011507382064010703,
      "loss": 0.8131,
      "num_input_tokens_seen": 48131968,
      "step": 82895
    },
    {
      "epoch": 12.347333929103366,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.011505486027303002,
      "loss": 0.7921,
      "num_input_tokens_seen": 48134784,
      "step": 82900
    },
    {
      "epoch": 12.348078641644324,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.011503590049631555,
      "loss": 0.792,
      "num_input_tokens_seen": 48137600,
      "step": 82905
    },
    {
      "epoch": 12.348823354185285,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.011501694131028407,
      "loss": 0.7716,
      "num_input_tokens_seen": 48140288,
      "step": 82910
    },
    {
      "epoch": 12.349568066726244,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.011499798271525577,
      "loss": 0.786,
      "num_input_tokens_seen": 48143424,
      "step": 82915
    },
    {
      "epoch": 12.350312779267203,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.011497902471155102,
      "loss": 0.7747,
      "num_input_tokens_seen": 48146656,
      "step": 82920
    },
    {
      "epoch": 12.351057491808161,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.011496006729949,
      "loss": 0.796,
      "num_input_tokens_seen": 48149472,
      "step": 82925
    },
    {
      "epoch": 12.351802204349122,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.011494111047939308,
      "loss": 0.8052,
      "num_input_tokens_seen": 48152576,
      "step": 82930
    },
    {
      "epoch": 12.35254691689008,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.011492215425158046,
      "loss": 0.7904,
      "num_input_tokens_seen": 48155648,
      "step": 82935
    },
    {
      "epoch": 12.35329162943104,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.011490319861637234,
      "loss": 0.7622,
      "num_input_tokens_seen": 48158752,
      "step": 82940
    },
    {
      "epoch": 12.354036341971998,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.011488424357408909,
      "loss": 0.7824,
      "num_input_tokens_seen": 48161568,
      "step": 82945
    },
    {
      "epoch": 12.354781054512959,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.01148652891250508,
      "loss": 0.8091,
      "num_input_tokens_seen": 48164352,
      "step": 82950
    },
    {
      "epoch": 12.355525767053917,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.011484633526957777,
      "loss": 0.7952,
      "num_input_tokens_seen": 48167328,
      "step": 82955
    },
    {
      "epoch": 12.356270479594876,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.011482738200799012,
      "loss": 0.8001,
      "num_input_tokens_seen": 48170432,
      "step": 82960
    },
    {
      "epoch": 12.357015192135835,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.011480842934060816,
      "loss": 0.7847,
      "num_input_tokens_seen": 48173408,
      "step": 82965
    },
    {
      "epoch": 12.357759904676795,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.011478947726775197,
      "loss": 0.7758,
      "num_input_tokens_seen": 48176512,
      "step": 82970
    },
    {
      "epoch": 12.358504617217754,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.011477052578974183,
      "loss": 0.7995,
      "num_input_tokens_seen": 48179552,
      "step": 82975
    },
    {
      "epoch": 12.359249329758713,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.011475157490689785,
      "loss": 0.8161,
      "num_input_tokens_seen": 48182368,
      "step": 82980
    },
    {
      "epoch": 12.359994042299672,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.011473262461954015,
      "loss": 0.7948,
      "num_input_tokens_seen": 48185504,
      "step": 82985
    },
    {
      "epoch": 12.360738754840632,
      "grad_norm": 0.03125,
      "learning_rate": 0.011471367492798897,
      "loss": 0.7999,
      "num_input_tokens_seen": 48188256,
      "step": 82990
    },
    {
      "epoch": 12.361483467381591,
      "grad_norm": 0.0390625,
      "learning_rate": 0.01146947258325643,
      "loss": 0.7899,
      "num_input_tokens_seen": 48190912,
      "step": 82995
    },
    {
      "epoch": 12.36222817992255,
      "grad_norm": 0.0390625,
      "learning_rate": 0.011467577733358643,
      "loss": 0.7903,
      "num_input_tokens_seen": 48193696,
      "step": 83000
    },
    {
      "epoch": 12.362972892463509,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.011465682943137533,
      "loss": 0.8121,
      "num_input_tokens_seen": 48196544,
      "step": 83005
    },
    {
      "epoch": 12.363717605004469,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.011463788212625125,
      "loss": 0.8076,
      "num_input_tokens_seen": 48199264,
      "step": 83010
    },
    {
      "epoch": 12.364462317545428,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01146189354185342,
      "loss": 0.8225,
      "num_input_tokens_seen": 48202016,
      "step": 83015
    },
    {
      "epoch": 12.365207030086387,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.011459998930854424,
      "loss": 0.7927,
      "num_input_tokens_seen": 48205024,
      "step": 83020
    },
    {
      "epoch": 12.365951742627345,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.011458104379660152,
      "loss": 0.7832,
      "num_input_tokens_seen": 48208192,
      "step": 83025
    },
    {
      "epoch": 12.366696455168306,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0114562098883026,
      "loss": 0.7816,
      "num_input_tokens_seen": 48210720,
      "step": 83030
    },
    {
      "epoch": 12.367441167709265,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.011454315456813788,
      "loss": 0.8092,
      "num_input_tokens_seen": 48213504,
      "step": 83035
    },
    {
      "epoch": 12.368185880250223,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.011452421085225706,
      "loss": 0.8025,
      "num_input_tokens_seen": 48216320,
      "step": 83040
    },
    {
      "epoch": 12.368930592791182,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.011450526773570368,
      "loss": 0.7877,
      "num_input_tokens_seen": 48219360,
      "step": 83045
    },
    {
      "epoch": 12.36967530533214,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01144863252187977,
      "loss": 0.8184,
      "num_input_tokens_seen": 48222240,
      "step": 83050
    },
    {
      "epoch": 12.370420017873101,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.011446738330185922,
      "loss": 0.7925,
      "num_input_tokens_seen": 48225216,
      "step": 83055
    },
    {
      "epoch": 12.37116473041406,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.011444844198520817,
      "loss": 0.8314,
      "num_input_tokens_seen": 48227840,
      "step": 83060
    },
    {
      "epoch": 12.371909442955019,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.011442950126916448,
      "loss": 0.7939,
      "num_input_tokens_seen": 48230656,
      "step": 83065
    },
    {
      "epoch": 12.37265415549598,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.011441056115404831,
      "loss": 0.7792,
      "num_input_tokens_seen": 48233408,
      "step": 83070
    },
    {
      "epoch": 12.373398868036938,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.011439162164017946,
      "loss": 0.7897,
      "num_input_tokens_seen": 48236416,
      "step": 83075
    },
    {
      "epoch": 12.374143580577897,
      "grad_norm": 0.044921875,
      "learning_rate": 0.011437268272787803,
      "loss": 0.7892,
      "num_input_tokens_seen": 48239200,
      "step": 83080
    },
    {
      "epoch": 12.374888293118856,
      "grad_norm": 0.048828125,
      "learning_rate": 0.011435374441746388,
      "loss": 0.7814,
      "num_input_tokens_seen": 48241952,
      "step": 83085
    },
    {
      "epoch": 12.375633005659815,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.011433480670925703,
      "loss": 0.7919,
      "num_input_tokens_seen": 48244864,
      "step": 83090
    },
    {
      "epoch": 12.376377718200775,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.011431586960357738,
      "loss": 0.7837,
      "num_input_tokens_seen": 48247680,
      "step": 83095
    },
    {
      "epoch": 12.377122430741734,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.011429693310074476,
      "loss": 0.8014,
      "num_input_tokens_seen": 48250912,
      "step": 83100
    },
    {
      "epoch": 12.377867143282693,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.011427799720107925,
      "loss": 0.7769,
      "num_input_tokens_seen": 48253696,
      "step": 83105
    },
    {
      "epoch": 12.378611855823651,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.01142590619049006,
      "loss": 0.8167,
      "num_input_tokens_seen": 48256320,
      "step": 83110
    },
    {
      "epoch": 12.379356568364612,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.011424012721252883,
      "loss": 0.8202,
      "num_input_tokens_seen": 48259360,
      "step": 83115
    },
    {
      "epoch": 12.38010128090557,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.011422119312428375,
      "loss": 0.8441,
      "num_input_tokens_seen": 48262336,
      "step": 83120
    },
    {
      "epoch": 12.38084599344653,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.011420225964048526,
      "loss": 0.7966,
      "num_input_tokens_seen": 48265248,
      "step": 83125
    },
    {
      "epoch": 12.381590705987488,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.011418332676145317,
      "loss": 0.8106,
      "num_input_tokens_seen": 48268128,
      "step": 83130
    },
    {
      "epoch": 12.382335418528449,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.011416439448750744,
      "loss": 0.7876,
      "num_input_tokens_seen": 48271008,
      "step": 83135
    },
    {
      "epoch": 12.383080131069407,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.011414546281896781,
      "loss": 0.7875,
      "num_input_tokens_seen": 48274368,
      "step": 83140
    },
    {
      "epoch": 12.383824843610366,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01141265317561541,
      "loss": 0.7905,
      "num_input_tokens_seen": 48277088,
      "step": 83145
    },
    {
      "epoch": 12.384569556151325,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.011410760129938623,
      "loss": 0.8035,
      "num_input_tokens_seen": 48280000,
      "step": 83150
    },
    {
      "epoch": 12.385314268692285,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.011408867144898393,
      "loss": 0.8016,
      "num_input_tokens_seen": 48282752,
      "step": 83155
    },
    {
      "epoch": 12.386058981233244,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.011406974220526707,
      "loss": 0.8019,
      "num_input_tokens_seen": 48285728,
      "step": 83160
    },
    {
      "epoch": 12.386803693774203,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.011405081356855534,
      "loss": 0.7989,
      "num_input_tokens_seen": 48289408,
      "step": 83165
    },
    {
      "epoch": 12.387548406315162,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.011403188553916862,
      "loss": 0.8077,
      "num_input_tokens_seen": 48292576,
      "step": 83170
    },
    {
      "epoch": 12.388293118856122,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.011401295811742663,
      "loss": 0.8115,
      "num_input_tokens_seen": 48295872,
      "step": 83175
    },
    {
      "epoch": 12.389037831397081,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.011399403130364912,
      "loss": 0.779,
      "num_input_tokens_seen": 48298336,
      "step": 83180
    },
    {
      "epoch": 12.38978254393804,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.011397510509815587,
      "loss": 0.7988,
      "num_input_tokens_seen": 48300960,
      "step": 83185
    },
    {
      "epoch": 12.390527256478999,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.01139561795012666,
      "loss": 0.7874,
      "num_input_tokens_seen": 48303616,
      "step": 83190
    },
    {
      "epoch": 12.391271969019959,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.011393725451330106,
      "loss": 0.8122,
      "num_input_tokens_seen": 48306656,
      "step": 83195
    },
    {
      "epoch": 12.392016681560918,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.011391833013457888,
      "loss": 0.8118,
      "num_input_tokens_seen": 48309696,
      "step": 83200
    },
    {
      "epoch": 12.392761394101877,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.011389940636541994,
      "loss": 0.8117,
      "num_input_tokens_seen": 48312608,
      "step": 83205
    },
    {
      "epoch": 12.393506106642835,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.011388048320614377,
      "loss": 0.7941,
      "num_input_tokens_seen": 48315456,
      "step": 83210
    },
    {
      "epoch": 12.394250819183796,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.01138615606570702,
      "loss": 0.8059,
      "num_input_tokens_seen": 48318720,
      "step": 83215
    },
    {
      "epoch": 12.394995531724755,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.01138426387185188,
      "loss": 0.8014,
      "num_input_tokens_seen": 48321792,
      "step": 83220
    },
    {
      "epoch": 12.395740244265713,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01138237173908092,
      "loss": 0.7861,
      "num_input_tokens_seen": 48324544,
      "step": 83225
    },
    {
      "epoch": 12.396484956806672,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.01138047966742612,
      "loss": 0.8009,
      "num_input_tokens_seen": 48327520,
      "step": 83230
    },
    {
      "epoch": 12.397229669347631,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.011378587656919431,
      "loss": 0.8191,
      "num_input_tokens_seen": 48330080,
      "step": 83235
    },
    {
      "epoch": 12.397974381888591,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01137669570759283,
      "loss": 0.7823,
      "num_input_tokens_seen": 48333184,
      "step": 83240
    },
    {
      "epoch": 12.39871909442955,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01137480381947827,
      "loss": 0.8195,
      "num_input_tokens_seen": 48336160,
      "step": 83245
    },
    {
      "epoch": 12.399463806970509,
      "grad_norm": 0.03125,
      "learning_rate": 0.011372911992607718,
      "loss": 0.7979,
      "num_input_tokens_seen": 48339616,
      "step": 83250
    },
    {
      "epoch": 12.400208519511468,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.011371020227013131,
      "loss": 0.7992,
      "num_input_tokens_seen": 48342592,
      "step": 83255
    },
    {
      "epoch": 12.400953232052428,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.011369128522726462,
      "loss": 0.8078,
      "num_input_tokens_seen": 48345440,
      "step": 83260
    },
    {
      "epoch": 12.401697944593387,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.01136723687977968,
      "loss": 0.7932,
      "num_input_tokens_seen": 48348448,
      "step": 83265
    },
    {
      "epoch": 12.402442657134346,
      "grad_norm": 0.05078125,
      "learning_rate": 0.011365345298204736,
      "loss": 0.8078,
      "num_input_tokens_seen": 48351072,
      "step": 83270
    },
    {
      "epoch": 12.403187369675305,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.011363453778033597,
      "loss": 0.7717,
      "num_input_tokens_seen": 48353760,
      "step": 83275
    },
    {
      "epoch": 12.403932082216265,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.011361562319298205,
      "loss": 0.7908,
      "num_input_tokens_seen": 48356768,
      "step": 83280
    },
    {
      "epoch": 12.404676794757224,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.011359670922030522,
      "loss": 0.7964,
      "num_input_tokens_seen": 48359744,
      "step": 83285
    },
    {
      "epoch": 12.405421507298183,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.011357779586262492,
      "loss": 0.8004,
      "num_input_tokens_seen": 48362720,
      "step": 83290
    },
    {
      "epoch": 12.406166219839141,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.011355888312026082,
      "loss": 0.7896,
      "num_input_tokens_seen": 48366016,
      "step": 83295
    },
    {
      "epoch": 12.406910932380102,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.011353997099353235,
      "loss": 0.7976,
      "num_input_tokens_seen": 48369056,
      "step": 83300
    },
    {
      "epoch": 12.40765564492106,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.011352105948275897,
      "loss": 0.7938,
      "num_input_tokens_seen": 48371968,
      "step": 83305
    },
    {
      "epoch": 12.40840035746202,
      "grad_norm": 0.01953125,
      "learning_rate": 0.011350214858826025,
      "loss": 0.8035,
      "num_input_tokens_seen": 48375008,
      "step": 83310
    },
    {
      "epoch": 12.409145070002978,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.01134832383103556,
      "loss": 0.8043,
      "num_input_tokens_seen": 48377792,
      "step": 83315
    },
    {
      "epoch": 12.409889782543939,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.011346432864936458,
      "loss": 0.7557,
      "num_input_tokens_seen": 48380832,
      "step": 83320
    },
    {
      "epoch": 12.410634495084897,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.011344541960560652,
      "loss": 0.7968,
      "num_input_tokens_seen": 48383616,
      "step": 83325
    },
    {
      "epoch": 12.411379207625856,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0113426511179401,
      "loss": 0.8273,
      "num_input_tokens_seen": 48386112,
      "step": 83330
    },
    {
      "epoch": 12.412123920166815,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.011340760337106742,
      "loss": 0.8027,
      "num_input_tokens_seen": 48388896,
      "step": 83335
    },
    {
      "epoch": 12.412868632707776,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.011338869618092511,
      "loss": 0.8269,
      "num_input_tokens_seen": 48391616,
      "step": 83340
    },
    {
      "epoch": 12.413613345248734,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.011336978960929363,
      "loss": 0.7871,
      "num_input_tokens_seen": 48394400,
      "step": 83345
    },
    {
      "epoch": 12.414358057789693,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.01133508836564923,
      "loss": 0.8312,
      "num_input_tokens_seen": 48397152,
      "step": 83350
    },
    {
      "epoch": 12.415102770330652,
      "grad_norm": 0.03515625,
      "learning_rate": 0.011333197832284059,
      "loss": 0.7966,
      "num_input_tokens_seen": 48400000,
      "step": 83355
    },
    {
      "epoch": 12.415847482871612,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.011331307360865775,
      "loss": 0.7929,
      "num_input_tokens_seen": 48402688,
      "step": 83360
    },
    {
      "epoch": 12.416592195412571,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.011329416951426332,
      "loss": 0.7711,
      "num_input_tokens_seen": 48405536,
      "step": 83365
    },
    {
      "epoch": 12.41733690795353,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.011327526603997651,
      "loss": 0.8134,
      "num_input_tokens_seen": 48408448,
      "step": 83370
    },
    {
      "epoch": 12.418081620494489,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.011325636318611684,
      "loss": 0.8082,
      "num_input_tokens_seen": 48411264,
      "step": 83375
    },
    {
      "epoch": 12.41882633303545,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.011323746095300357,
      "loss": 0.7804,
      "num_input_tokens_seen": 48414304,
      "step": 83380
    },
    {
      "epoch": 12.419571045576408,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.011321855934095598,
      "loss": 0.7966,
      "num_input_tokens_seen": 48417376,
      "step": 83385
    },
    {
      "epoch": 12.420315758117367,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.011319965835029349,
      "loss": 0.8005,
      "num_input_tokens_seen": 48420288,
      "step": 83390
    },
    {
      "epoch": 12.421060470658325,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.01131807579813353,
      "loss": 0.7871,
      "num_input_tokens_seen": 48423296,
      "step": 83395
    },
    {
      "epoch": 12.421805183199286,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.011316185823440086,
      "loss": 0.8154,
      "num_input_tokens_seen": 48426560,
      "step": 83400
    },
    {
      "epoch": 12.422549895740245,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.011314295910980932,
      "loss": 0.8127,
      "num_input_tokens_seen": 48429504,
      "step": 83405
    },
    {
      "epoch": 12.423294608281203,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.011312406060788009,
      "loss": 0.7972,
      "num_input_tokens_seen": 48432288,
      "step": 83410
    },
    {
      "epoch": 12.424039320822162,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.011310516272893232,
      "loss": 0.8119,
      "num_input_tokens_seen": 48435072,
      "step": 83415
    },
    {
      "epoch": 12.424784033363121,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.011308626547328539,
      "loss": 0.7812,
      "num_input_tokens_seen": 48437728,
      "step": 83420
    },
    {
      "epoch": 12.425528745904082,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.011306736884125848,
      "loss": 0.7776,
      "num_input_tokens_seen": 48440672,
      "step": 83425
    },
    {
      "epoch": 12.42627345844504,
      "grad_norm": 0.357421875,
      "learning_rate": 0.011304847283317078,
      "loss": 0.8093,
      "num_input_tokens_seen": 48443552,
      "step": 83430
    },
    {
      "epoch": 12.427018170985999,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01130295774493416,
      "loss": 0.8095,
      "num_input_tokens_seen": 48446368,
      "step": 83435
    },
    {
      "epoch": 12.427762883526958,
      "grad_norm": 0.04296875,
      "learning_rate": 0.011301068269009012,
      "loss": 0.7781,
      "num_input_tokens_seen": 48449248,
      "step": 83440
    },
    {
      "epoch": 12.428507596067918,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.01129917885557356,
      "loss": 0.807,
      "num_input_tokens_seen": 48452320,
      "step": 83445
    },
    {
      "epoch": 12.429252308608877,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.011297289504659715,
      "loss": 0.8204,
      "num_input_tokens_seen": 48455008,
      "step": 83450
    },
    {
      "epoch": 12.429997021149836,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.011295400216299408,
      "loss": 0.8123,
      "num_input_tokens_seen": 48458080,
      "step": 83455
    },
    {
      "epoch": 12.430741733690795,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.011293510990524547,
      "loss": 0.834,
      "num_input_tokens_seen": 48460928,
      "step": 83460
    },
    {
      "epoch": 12.431486446231755,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.011291621827367045,
      "loss": 0.8056,
      "num_input_tokens_seen": 48464448,
      "step": 83465
    },
    {
      "epoch": 12.432231158772714,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.011289732726858829,
      "loss": 0.827,
      "num_input_tokens_seen": 48467264,
      "step": 83470
    },
    {
      "epoch": 12.432975871313673,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0112878436890318,
      "loss": 0.8008,
      "num_input_tokens_seen": 48470080,
      "step": 83475
    },
    {
      "epoch": 12.433720583854631,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.011285954713917886,
      "loss": 0.8097,
      "num_input_tokens_seen": 48473152,
      "step": 83480
    },
    {
      "epoch": 12.434465296395592,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.01128406580154899,
      "loss": 0.8204,
      "num_input_tokens_seen": 48475840,
      "step": 83485
    },
    {
      "epoch": 12.43521000893655,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.011282176951957028,
      "loss": 0.8264,
      "num_input_tokens_seen": 48479104,
      "step": 83490
    },
    {
      "epoch": 12.43595472147751,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.011280288165173902,
      "loss": 0.797,
      "num_input_tokens_seen": 48482144,
      "step": 83495
    },
    {
      "epoch": 12.436699434018468,
      "grad_norm": 0.04296875,
      "learning_rate": 0.011278399441231534,
      "loss": 0.8003,
      "num_input_tokens_seen": 48484992,
      "step": 83500
    },
    {
      "epoch": 12.437444146559429,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.011276510780161824,
      "loss": 0.8127,
      "num_input_tokens_seen": 48488032,
      "step": 83505
    },
    {
      "epoch": 12.438188859100388,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.011274622181996676,
      "loss": 0.8055,
      "num_input_tokens_seen": 48491040,
      "step": 83510
    },
    {
      "epoch": 12.438933571641346,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.011272733646768002,
      "loss": 0.8116,
      "num_input_tokens_seen": 48493792,
      "step": 83515
    },
    {
      "epoch": 12.439678284182305,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.011270845174507703,
      "loss": 0.8058,
      "num_input_tokens_seen": 48496576,
      "step": 83520
    },
    {
      "epoch": 12.440422996723266,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.01126895676524769,
      "loss": 0.7854,
      "num_input_tokens_seen": 48499296,
      "step": 83525
    },
    {
      "epoch": 12.441167709264224,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.01126706841901985,
      "loss": 0.8,
      "num_input_tokens_seen": 48501984,
      "step": 83530
    },
    {
      "epoch": 12.441912421805183,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.011265180135856104,
      "loss": 0.813,
      "num_input_tokens_seen": 48504800,
      "step": 83535
    },
    {
      "epoch": 12.442657134346142,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.011263291915788344,
      "loss": 0.8148,
      "num_input_tokens_seen": 48507520,
      "step": 83540
    },
    {
      "epoch": 12.443401846887102,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.011261403758848466,
      "loss": 0.8186,
      "num_input_tokens_seen": 48510560,
      "step": 83545
    },
    {
      "epoch": 12.444146559428061,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.011259515665068375,
      "loss": 0.8074,
      "num_input_tokens_seen": 48513440,
      "step": 83550
    },
    {
      "epoch": 12.44489127196902,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.011257627634479959,
      "loss": 0.8097,
      "num_input_tokens_seen": 48516352,
      "step": 83555
    },
    {
      "epoch": 12.445635984509979,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.011255739667115126,
      "loss": 0.8023,
      "num_input_tokens_seen": 48519584,
      "step": 83560
    },
    {
      "epoch": 12.44638069705094,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.01125385176300576,
      "loss": 0.7865,
      "num_input_tokens_seen": 48522240,
      "step": 83565
    },
    {
      "epoch": 12.447125409591898,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.011251963922183767,
      "loss": 0.7819,
      "num_input_tokens_seen": 48525248,
      "step": 83570
    },
    {
      "epoch": 12.447870122132857,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.011250076144681032,
      "loss": 0.782,
      "num_input_tokens_seen": 48528032,
      "step": 83575
    },
    {
      "epoch": 12.448614834673815,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.011248188430529452,
      "loss": 0.8004,
      "num_input_tokens_seen": 48530848,
      "step": 83580
    },
    {
      "epoch": 12.449359547214776,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.011246300779760915,
      "loss": 0.7779,
      "num_input_tokens_seen": 48533632,
      "step": 83585
    },
    {
      "epoch": 12.450104259755735,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.011244413192407304,
      "loss": 0.7865,
      "num_input_tokens_seen": 48536608,
      "step": 83590
    },
    {
      "epoch": 12.450848972296694,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.011242525668500522,
      "loss": 0.805,
      "num_input_tokens_seen": 48539584,
      "step": 83595
    },
    {
      "epoch": 12.451593684837652,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.011240638208072444,
      "loss": 0.8099,
      "num_input_tokens_seen": 48542688,
      "step": 83600
    },
    {
      "epoch": 12.452338397378611,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.011238750811154968,
      "loss": 0.7942,
      "num_input_tokens_seen": 48545760,
      "step": 83605
    },
    {
      "epoch": 12.453083109919572,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.011236863477779972,
      "loss": 0.8037,
      "num_input_tokens_seen": 48548608,
      "step": 83610
    },
    {
      "epoch": 12.45382782246053,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.011234976207979348,
      "loss": 0.7838,
      "num_input_tokens_seen": 48551424,
      "step": 83615
    },
    {
      "epoch": 12.454572535001489,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.011233089001784971,
      "loss": 0.81,
      "num_input_tokens_seen": 48554176,
      "step": 83620
    },
    {
      "epoch": 12.455317247542448,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.011231201859228723,
      "loss": 0.8062,
      "num_input_tokens_seen": 48556672,
      "step": 83625
    },
    {
      "epoch": 12.456061960083408,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.011229314780342494,
      "loss": 0.7971,
      "num_input_tokens_seen": 48559520,
      "step": 83630
    },
    {
      "epoch": 12.456806672624367,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.011227427765158154,
      "loss": 0.7909,
      "num_input_tokens_seen": 48562592,
      "step": 83635
    },
    {
      "epoch": 12.457551385165326,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.011225540813707592,
      "loss": 0.8072,
      "num_input_tokens_seen": 48565376,
      "step": 83640
    },
    {
      "epoch": 12.458296097706285,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.01122365392602268,
      "loss": 0.8107,
      "num_input_tokens_seen": 48568192,
      "step": 83645
    },
    {
      "epoch": 12.459040810247245,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0112217671021353,
      "loss": 0.8026,
      "num_input_tokens_seen": 48571072,
      "step": 83650
    },
    {
      "epoch": 12.459785522788204,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.01121988034207732,
      "loss": 0.8245,
      "num_input_tokens_seen": 48573888,
      "step": 83655
    },
    {
      "epoch": 12.460530235329163,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.011217993645880622,
      "loss": 0.8096,
      "num_input_tokens_seen": 48576768,
      "step": 83660
    },
    {
      "epoch": 12.461274947870121,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.011216107013577082,
      "loss": 0.7831,
      "num_input_tokens_seen": 48579712,
      "step": 83665
    },
    {
      "epoch": 12.462019660411082,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.011214220445198559,
      "loss": 0.7814,
      "num_input_tokens_seen": 48583072,
      "step": 83670
    },
    {
      "epoch": 12.46276437295204,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.011212333940776942,
      "loss": 0.8117,
      "num_input_tokens_seen": 48585728,
      "step": 83675
    },
    {
      "epoch": 12.463509085493,
      "grad_norm": 0.03515625,
      "learning_rate": 0.011210447500344087,
      "loss": 0.7876,
      "num_input_tokens_seen": 48588256,
      "step": 83680
    },
    {
      "epoch": 12.464253798033958,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.011208561123931877,
      "loss": 0.8254,
      "num_input_tokens_seen": 48591168,
      "step": 83685
    },
    {
      "epoch": 12.464998510574919,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.011206674811572165,
      "loss": 0.8157,
      "num_input_tokens_seen": 48594112,
      "step": 83690
    },
    {
      "epoch": 12.465743223115878,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.011204788563296835,
      "loss": 0.7974,
      "num_input_tokens_seen": 48597024,
      "step": 83695
    },
    {
      "epoch": 12.466487935656836,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.011202902379137743,
      "loss": 0.8037,
      "num_input_tokens_seen": 48600288,
      "step": 83700
    },
    {
      "epoch": 12.467232648197795,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.01120101625912675,
      "loss": 0.8099,
      "num_input_tokens_seen": 48603648,
      "step": 83705
    },
    {
      "epoch": 12.467977360738756,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.011199130203295734,
      "loss": 0.8182,
      "num_input_tokens_seen": 48606528,
      "step": 83710
    },
    {
      "epoch": 12.468722073279714,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.011197244211676545,
      "loss": 0.8093,
      "num_input_tokens_seen": 48609376,
      "step": 83715
    },
    {
      "epoch": 12.469466785820673,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.011195358284301054,
      "loss": 0.8005,
      "num_input_tokens_seen": 48612320,
      "step": 83720
    },
    {
      "epoch": 12.470211498361632,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.011193472421201112,
      "loss": 0.8081,
      "num_input_tokens_seen": 48615072,
      "step": 83725
    },
    {
      "epoch": 12.470956210902592,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.01119158662240859,
      "loss": 0.8155,
      "num_input_tokens_seen": 48618304,
      "step": 83730
    },
    {
      "epoch": 12.471700923443551,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.011189700887955336,
      "loss": 0.7817,
      "num_input_tokens_seen": 48621152,
      "step": 83735
    },
    {
      "epoch": 12.47244563598451,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.011187815217873218,
      "loss": 0.7907,
      "num_input_tokens_seen": 48624224,
      "step": 83740
    },
    {
      "epoch": 12.473190348525469,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.01118592961219409,
      "loss": 0.8045,
      "num_input_tokens_seen": 48627296,
      "step": 83745
    },
    {
      "epoch": 12.473935061066427,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0111840440709498,
      "loss": 0.8148,
      "num_input_tokens_seen": 48630336,
      "step": 83750
    },
    {
      "epoch": 12.474679773607388,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.011182158594172209,
      "loss": 0.7945,
      "num_input_tokens_seen": 48633120,
      "step": 83755
    },
    {
      "epoch": 12.475424486148347,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.011180273181893163,
      "loss": 0.7965,
      "num_input_tokens_seen": 48635872,
      "step": 83760
    },
    {
      "epoch": 12.476169198689306,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.011178387834144526,
      "loss": 0.7971,
      "num_input_tokens_seen": 48638720,
      "step": 83765
    },
    {
      "epoch": 12.476913911230264,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.011176502550958136,
      "loss": 0.7828,
      "num_input_tokens_seen": 48641600,
      "step": 83770
    },
    {
      "epoch": 12.477658623771225,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.011174617332365855,
      "loss": 0.8224,
      "num_input_tokens_seen": 48644480,
      "step": 83775
    },
    {
      "epoch": 12.478403336312184,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.011172732178399526,
      "loss": 0.8118,
      "num_input_tokens_seen": 48647328,
      "step": 83780
    },
    {
      "epoch": 12.479148048853142,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.011170847089090993,
      "loss": 0.8057,
      "num_input_tokens_seen": 48650368,
      "step": 83785
    },
    {
      "epoch": 12.479892761394101,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.011168962064472113,
      "loss": 0.8129,
      "num_input_tokens_seen": 48653088,
      "step": 83790
    },
    {
      "epoch": 12.480637473935062,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.011167077104574718,
      "loss": 0.807,
      "num_input_tokens_seen": 48656096,
      "step": 83795
    },
    {
      "epoch": 12.48138218647602,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.011165192209430665,
      "loss": 0.8058,
      "num_input_tokens_seen": 48658976,
      "step": 83800
    },
    {
      "epoch": 12.48212689901698,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.011163307379071785,
      "loss": 0.8099,
      "num_input_tokens_seen": 48661920,
      "step": 83805
    },
    {
      "epoch": 12.482871611557938,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.011161422613529936,
      "loss": 0.8072,
      "num_input_tokens_seen": 48664864,
      "step": 83810
    },
    {
      "epoch": 12.483616324098898,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.011159537912836946,
      "loss": 0.8024,
      "num_input_tokens_seen": 48667520,
      "step": 83815
    },
    {
      "epoch": 12.484361036639857,
      "grad_norm": 0.029296875,
      "learning_rate": 0.011157653277024662,
      "loss": 0.8083,
      "num_input_tokens_seen": 48670336,
      "step": 83820
    },
    {
      "epoch": 12.485105749180816,
      "grad_norm": 0.037109375,
      "learning_rate": 0.011155768706124922,
      "loss": 0.8143,
      "num_input_tokens_seen": 48673184,
      "step": 83825
    },
    {
      "epoch": 12.485850461721775,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.011153884200169556,
      "loss": 0.8112,
      "num_input_tokens_seen": 48676000,
      "step": 83830
    },
    {
      "epoch": 12.486595174262735,
      "grad_norm": 0.048828125,
      "learning_rate": 0.011151999759190415,
      "loss": 0.7937,
      "num_input_tokens_seen": 48678880,
      "step": 83835
    },
    {
      "epoch": 12.487339886803694,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.011150115383219323,
      "loss": 0.8133,
      "num_input_tokens_seen": 48681664,
      "step": 83840
    },
    {
      "epoch": 12.488084599344653,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.011148231072288118,
      "loss": 0.8097,
      "num_input_tokens_seen": 48684320,
      "step": 83845
    },
    {
      "epoch": 12.488829311885612,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.011146346826428634,
      "loss": 0.8019,
      "num_input_tokens_seen": 48687104,
      "step": 83850
    },
    {
      "epoch": 12.489574024426572,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.011144462645672707,
      "loss": 0.8009,
      "num_input_tokens_seen": 48690048,
      "step": 83855
    },
    {
      "epoch": 12.49031873696753,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.011142578530052166,
      "loss": 0.808,
      "num_input_tokens_seen": 48692992,
      "step": 83860
    },
    {
      "epoch": 12.49106344950849,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.011140694479598834,
      "loss": 0.8229,
      "num_input_tokens_seen": 48695968,
      "step": 83865
    },
    {
      "epoch": 12.491808162049448,
      "grad_norm": 0.046875,
      "learning_rate": 0.011138810494344552,
      "loss": 0.8096,
      "num_input_tokens_seen": 48699200,
      "step": 83870
    },
    {
      "epoch": 12.492552874590409,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.011136926574321138,
      "loss": 0.8066,
      "num_input_tokens_seen": 48702144,
      "step": 83875
    },
    {
      "epoch": 12.493297587131368,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.011135042719560428,
      "loss": 0.8205,
      "num_input_tokens_seen": 48705120,
      "step": 83880
    },
    {
      "epoch": 12.494042299672326,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.011133158930094238,
      "loss": 0.8001,
      "num_input_tokens_seen": 48707840,
      "step": 83885
    },
    {
      "epoch": 12.494787012213285,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0111312752059544,
      "loss": 0.8068,
      "num_input_tokens_seen": 48710848,
      "step": 83890
    },
    {
      "epoch": 12.495531724754246,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.011129391547172735,
      "loss": 0.8046,
      "num_input_tokens_seen": 48713792,
      "step": 83895
    },
    {
      "epoch": 12.496276437295204,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.011127507953781068,
      "loss": 0.804,
      "num_input_tokens_seen": 48716896,
      "step": 83900
    },
    {
      "epoch": 12.497021149836163,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.011125624425811219,
      "loss": 0.8104,
      "num_input_tokens_seen": 48719552,
      "step": 83905
    },
    {
      "epoch": 12.497765862377122,
      "grad_norm": 0.0390625,
      "learning_rate": 0.011123740963295005,
      "loss": 0.7924,
      "num_input_tokens_seen": 48722144,
      "step": 83910
    },
    {
      "epoch": 12.498510574918082,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.011121857566264251,
      "loss": 0.7875,
      "num_input_tokens_seen": 48725024,
      "step": 83915
    },
    {
      "epoch": 12.499255287459041,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.011119974234750767,
      "loss": 0.813,
      "num_input_tokens_seen": 48728192,
      "step": 83920
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.011118090968786378,
      "loss": 0.7933,
      "num_input_tokens_seen": 48731296,
      "step": 83925
    },
    {
      "epoch": 12.500744712540959,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.011116207768402891,
      "loss": 0.8013,
      "num_input_tokens_seen": 48734080,
      "step": 83930
    },
    {
      "epoch": 12.501489425081918,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.011114324633632135,
      "loss": 0.8086,
      "num_input_tokens_seen": 48736736,
      "step": 83935
    },
    {
      "epoch": 12.502234137622878,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.011112441564505911,
      "loss": 0.8055,
      "num_input_tokens_seen": 48739584,
      "step": 83940
    },
    {
      "epoch": 12.502978850163837,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.011110558561056036,
      "loss": 0.7881,
      "num_input_tokens_seen": 48742720,
      "step": 83945
    },
    {
      "epoch": 12.503723562704796,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.011108675623314324,
      "loss": 0.8053,
      "num_input_tokens_seen": 48745696,
      "step": 83950
    },
    {
      "epoch": 12.504468275245754,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.011106792751312574,
      "loss": 0.781,
      "num_input_tokens_seen": 48748416,
      "step": 83955
    },
    {
      "epoch": 12.505212987786715,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.011104909945082612,
      "loss": 0.8036,
      "num_input_tokens_seen": 48751296,
      "step": 83960
    },
    {
      "epoch": 12.505957700327674,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.011103027204656231,
      "loss": 0.8077,
      "num_input_tokens_seen": 48754336,
      "step": 83965
    },
    {
      "epoch": 12.506702412868632,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.01110114453006525,
      "loss": 0.7889,
      "num_input_tokens_seen": 48757440,
      "step": 83970
    },
    {
      "epoch": 12.507447125409591,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.011099261921341468,
      "loss": 0.8183,
      "num_input_tokens_seen": 48760384,
      "step": 83975
    },
    {
      "epoch": 12.508191837950552,
      "grad_norm": 0.0625,
      "learning_rate": 0.011097379378516691,
      "loss": 0.8005,
      "num_input_tokens_seen": 48763296,
      "step": 83980
    },
    {
      "epoch": 12.50893655049151,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.011095496901622728,
      "loss": 0.7915,
      "num_input_tokens_seen": 48766208,
      "step": 83985
    },
    {
      "epoch": 12.50968126303247,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.011093614490691368,
      "loss": 0.7957,
      "num_input_tokens_seen": 48769280,
      "step": 83990
    },
    {
      "epoch": 12.510425975573428,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.011091732145754426,
      "loss": 0.8024,
      "num_input_tokens_seen": 48772128,
      "step": 83995
    },
    {
      "epoch": 12.511170688114388,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01108984986684369,
      "loss": 0.8086,
      "num_input_tokens_seen": 48775008,
      "step": 84000
    },
    {
      "epoch": 12.511915400655347,
      "grad_norm": 0.068359375,
      "learning_rate": 0.011087967653990977,
      "loss": 0.7942,
      "num_input_tokens_seen": 48777760,
      "step": 84005
    },
    {
      "epoch": 12.512660113196306,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.011086085507228066,
      "loss": 0.7961,
      "num_input_tokens_seen": 48780288,
      "step": 84010
    },
    {
      "epoch": 12.513404825737265,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.01108420342658677,
      "loss": 0.7965,
      "num_input_tokens_seen": 48783168,
      "step": 84015
    },
    {
      "epoch": 12.514149538278225,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.011082321412098868,
      "loss": 0.8092,
      "num_input_tokens_seen": 48786816,
      "step": 84020
    },
    {
      "epoch": 12.514894250819184,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.011080439463796173,
      "loss": 0.7867,
      "num_input_tokens_seen": 48789696,
      "step": 84025
    },
    {
      "epoch": 12.515638963360143,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.011078557581710469,
      "loss": 0.7953,
      "num_input_tokens_seen": 48792576,
      "step": 84030
    },
    {
      "epoch": 12.516383675901102,
      "grad_norm": 0.076171875,
      "learning_rate": 0.011076675765873541,
      "loss": 0.8104,
      "num_input_tokens_seen": 48795168,
      "step": 84035
    },
    {
      "epoch": 12.517128388442062,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.011074794016317195,
      "loss": 0.8017,
      "num_input_tokens_seen": 48798080,
      "step": 84040
    },
    {
      "epoch": 12.51787310098302,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01107291233307321,
      "loss": 0.7991,
      "num_input_tokens_seen": 48801120,
      "step": 84045
    },
    {
      "epoch": 12.51861781352398,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.011071030716173387,
      "loss": 0.7911,
      "num_input_tokens_seen": 48804064,
      "step": 84050
    },
    {
      "epoch": 12.519362526064938,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.011069149165649498,
      "loss": 0.809,
      "num_input_tokens_seen": 48806848,
      "step": 84055
    },
    {
      "epoch": 12.520107238605899,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.011067267681533346,
      "loss": 0.7942,
      "num_input_tokens_seen": 48809888,
      "step": 84060
    },
    {
      "epoch": 12.520851951146858,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.011065386263856709,
      "loss": 0.8068,
      "num_input_tokens_seen": 48812864,
      "step": 84065
    },
    {
      "epoch": 12.521596663687816,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.011063504912651366,
      "loss": 0.7858,
      "num_input_tokens_seen": 48815808,
      "step": 84070
    },
    {
      "epoch": 12.522341376228775,
      "grad_norm": 0.07421875,
      "learning_rate": 0.011061623627949112,
      "loss": 0.7904,
      "num_input_tokens_seen": 48818592,
      "step": 84075
    },
    {
      "epoch": 12.523086088769734,
      "grad_norm": 0.046875,
      "learning_rate": 0.011059742409781723,
      "loss": 0.8038,
      "num_input_tokens_seen": 48821376,
      "step": 84080
    },
    {
      "epoch": 12.523830801310694,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.011057861258180984,
      "loss": 0.7983,
      "num_input_tokens_seen": 48824224,
      "step": 84085
    },
    {
      "epoch": 12.524575513851653,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.011055980173178664,
      "loss": 0.7838,
      "num_input_tokens_seen": 48827040,
      "step": 84090
    },
    {
      "epoch": 12.525320226392612,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.01105409915480656,
      "loss": 0.7955,
      "num_input_tokens_seen": 48829696,
      "step": 84095
    },
    {
      "epoch": 12.526064938933573,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.011052218203096431,
      "loss": 0.866,
      "num_input_tokens_seen": 48832864,
      "step": 84100
    },
    {
      "epoch": 12.526809651474531,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.01105033731808007,
      "loss": 0.7834,
      "num_input_tokens_seen": 48835584,
      "step": 84105
    },
    {
      "epoch": 12.52755436401549,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.011048456499789246,
      "loss": 0.7902,
      "num_input_tokens_seen": 48838432,
      "step": 84110
    },
    {
      "epoch": 12.528299076556449,
      "grad_norm": 0.2099609375,
      "learning_rate": 0.011046575748255731,
      "loss": 0.8539,
      "num_input_tokens_seen": 48841664,
      "step": 84115
    },
    {
      "epoch": 12.529043789097408,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.011044695063511304,
      "loss": 0.795,
      "num_input_tokens_seen": 48844448,
      "step": 84120
    },
    {
      "epoch": 12.529788501638368,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.011042814445587728,
      "loss": 0.7987,
      "num_input_tokens_seen": 48847456,
      "step": 84125
    },
    {
      "epoch": 12.530533214179327,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.011040933894516783,
      "loss": 0.7979,
      "num_input_tokens_seen": 48850464,
      "step": 84130
    },
    {
      "epoch": 12.531277926720286,
      "grad_norm": 0.02734375,
      "learning_rate": 0.011039053410330232,
      "loss": 0.7932,
      "num_input_tokens_seen": 48853248,
      "step": 84135
    },
    {
      "epoch": 12.532022639261244,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.011037172993059853,
      "loss": 0.7866,
      "num_input_tokens_seen": 48856032,
      "step": 84140
    },
    {
      "epoch": 12.532767351802205,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.01103529264273741,
      "loss": 0.8069,
      "num_input_tokens_seen": 48859072,
      "step": 84145
    },
    {
      "epoch": 12.533512064343164,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.011033412359394668,
      "loss": 0.8076,
      "num_input_tokens_seen": 48862176,
      "step": 84150
    },
    {
      "epoch": 12.534256776884122,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.011031532143063391,
      "loss": 0.8068,
      "num_input_tokens_seen": 48865184,
      "step": 84155
    },
    {
      "epoch": 12.535001489425081,
      "grad_norm": 0.033203125,
      "learning_rate": 0.01102965199377534,
      "loss": 0.7965,
      "num_input_tokens_seen": 48868096,
      "step": 84160
    },
    {
      "epoch": 12.535746201966042,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.01102777191156229,
      "loss": 0.8527,
      "num_input_tokens_seen": 48870784,
      "step": 84165
    },
    {
      "epoch": 12.536490914507,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.011025891896455994,
      "loss": 0.7919,
      "num_input_tokens_seen": 48873408,
      "step": 84170
    },
    {
      "epoch": 12.53723562704796,
      "grad_norm": 0.033203125,
      "learning_rate": 0.011024011948488215,
      "loss": 0.7917,
      "num_input_tokens_seen": 48876256,
      "step": 84175
    },
    {
      "epoch": 12.537980339588918,
      "grad_norm": 0.04296875,
      "learning_rate": 0.011022132067690711,
      "loss": 0.7972,
      "num_input_tokens_seen": 48878944,
      "step": 84180
    },
    {
      "epoch": 12.538725052129879,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.01102025225409525,
      "loss": 0.7918,
      "num_input_tokens_seen": 48882080,
      "step": 84185
    },
    {
      "epoch": 12.539469764670837,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.01101837250773358,
      "loss": 0.8067,
      "num_input_tokens_seen": 48884896,
      "step": 84190
    },
    {
      "epoch": 12.540214477211796,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.011016492828637451,
      "loss": 0.8065,
      "num_input_tokens_seen": 48887840,
      "step": 84195
    },
    {
      "epoch": 12.540959189752755,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.011014613216838637,
      "loss": 0.7942,
      "num_input_tokens_seen": 48890560,
      "step": 84200
    },
    {
      "epoch": 12.541703902293715,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.011012733672368876,
      "loss": 0.7947,
      "num_input_tokens_seen": 48893536,
      "step": 84205
    },
    {
      "epoch": 12.542448614834674,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.011010854195259934,
      "loss": 0.799,
      "num_input_tokens_seen": 48896704,
      "step": 84210
    },
    {
      "epoch": 12.543193327375633,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.011008974785543546,
      "loss": 0.781,
      "num_input_tokens_seen": 48899616,
      "step": 84215
    },
    {
      "epoch": 12.543938039916592,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.011007095443251482,
      "loss": 0.8006,
      "num_input_tokens_seen": 48902496,
      "step": 84220
    },
    {
      "epoch": 12.544682752457552,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.011005216168415481,
      "loss": 0.7841,
      "num_input_tokens_seen": 48905472,
      "step": 84225
    },
    {
      "epoch": 12.545427464998511,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.011003336961067285,
      "loss": 0.7968,
      "num_input_tokens_seen": 48908704,
      "step": 84230
    },
    {
      "epoch": 12.54617217753947,
      "grad_norm": 0.048828125,
      "learning_rate": 0.011001457821238658,
      "loss": 0.7884,
      "num_input_tokens_seen": 48911584,
      "step": 84235
    },
    {
      "epoch": 12.546916890080428,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.010999578748961332,
      "loss": 0.7869,
      "num_input_tokens_seen": 48914144,
      "step": 84240
    },
    {
      "epoch": 12.547661602621389,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.010997699744267063,
      "loss": 0.8032,
      "num_input_tokens_seen": 48916832,
      "step": 84245
    },
    {
      "epoch": 12.548406315162348,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.010995820807187579,
      "loss": 0.7969,
      "num_input_tokens_seen": 48919552,
      "step": 84250
    },
    {
      "epoch": 12.549151027703306,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.010993941937754643,
      "loss": 0.7842,
      "num_input_tokens_seen": 48922560,
      "step": 84255
    },
    {
      "epoch": 12.549895740244265,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.010992063135999977,
      "loss": 0.8074,
      "num_input_tokens_seen": 48925312,
      "step": 84260
    },
    {
      "epoch": 12.550640452785224,
      "grad_norm": 0.0625,
      "learning_rate": 0.010990184401955339,
      "loss": 0.804,
      "num_input_tokens_seen": 48927840,
      "step": 84265
    },
    {
      "epoch": 12.551385165326185,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.010988305735652458,
      "loss": 0.7994,
      "num_input_tokens_seen": 48930688,
      "step": 84270
    },
    {
      "epoch": 12.552129877867143,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.010986427137123073,
      "loss": 0.7866,
      "num_input_tokens_seen": 48933888,
      "step": 84275
    },
    {
      "epoch": 12.552874590408102,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.010984548606398924,
      "loss": 0.8102,
      "num_input_tokens_seen": 48936640,
      "step": 84280
    },
    {
      "epoch": 12.553619302949063,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.01098267014351174,
      "loss": 0.7971,
      "num_input_tokens_seen": 48939776,
      "step": 84285
    },
    {
      "epoch": 12.554364015490021,
      "grad_norm": 0.064453125,
      "learning_rate": 0.010980791748493268,
      "loss": 0.8227,
      "num_input_tokens_seen": 48942624,
      "step": 84290
    },
    {
      "epoch": 12.55510872803098,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.010978913421375226,
      "loss": 0.779,
      "num_input_tokens_seen": 48945440,
      "step": 84295
    },
    {
      "epoch": 12.555853440571939,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.010977035162189364,
      "loss": 0.8083,
      "num_input_tokens_seen": 48948224,
      "step": 84300
    },
    {
      "epoch": 12.556598153112898,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0109751569709674,
      "loss": 0.788,
      "num_input_tokens_seen": 48950944,
      "step": 84305
    },
    {
      "epoch": 12.557342865653858,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.01097327884774107,
      "loss": 0.7938,
      "num_input_tokens_seen": 48953568,
      "step": 84310
    },
    {
      "epoch": 12.558087578194817,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.010971400792542102,
      "loss": 0.812,
      "num_input_tokens_seen": 48956256,
      "step": 84315
    },
    {
      "epoch": 12.558832290735776,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.010969522805402218,
      "loss": 0.8021,
      "num_input_tokens_seen": 48959072,
      "step": 84320
    },
    {
      "epoch": 12.559577003276734,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.010967644886353153,
      "loss": 0.7855,
      "num_input_tokens_seen": 48961792,
      "step": 84325
    },
    {
      "epoch": 12.560321715817695,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.010965767035426625,
      "loss": 0.797,
      "num_input_tokens_seen": 48964768,
      "step": 84330
    },
    {
      "epoch": 12.561066428358654,
      "grad_norm": 0.052734375,
      "learning_rate": 0.01096388925265437,
      "loss": 0.793,
      "num_input_tokens_seen": 48967872,
      "step": 84335
    },
    {
      "epoch": 12.561811140899612,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0109620115380681,
      "loss": 0.7954,
      "num_input_tokens_seen": 48970816,
      "step": 84340
    },
    {
      "epoch": 12.562555853440571,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.010960133891699543,
      "loss": 0.8,
      "num_input_tokens_seen": 48973760,
      "step": 84345
    },
    {
      "epoch": 12.563300565981532,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.010958256313580418,
      "loss": 0.8084,
      "num_input_tokens_seen": 48976576,
      "step": 84350
    },
    {
      "epoch": 12.56404527852249,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.010956378803742442,
      "loss": 0.7976,
      "num_input_tokens_seen": 48979520,
      "step": 84355
    },
    {
      "epoch": 12.56478999106345,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01095450136221734,
      "loss": 0.8051,
      "num_input_tokens_seen": 48982208,
      "step": 84360
    },
    {
      "epoch": 12.565534703604408,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.010952623989036818,
      "loss": 0.7985,
      "num_input_tokens_seen": 48985024,
      "step": 84365
    },
    {
      "epoch": 12.566279416145369,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.010950746684232609,
      "loss": 0.7914,
      "num_input_tokens_seen": 48987904,
      "step": 84370
    },
    {
      "epoch": 12.567024128686327,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.010948869447836415,
      "loss": 0.8088,
      "num_input_tokens_seen": 48990976,
      "step": 84375
    },
    {
      "epoch": 12.567768841227286,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.010946992279879959,
      "loss": 0.7853,
      "num_input_tokens_seen": 48994048,
      "step": 84380
    },
    {
      "epoch": 12.568513553768245,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.010945115180394947,
      "loss": 0.776,
      "num_input_tokens_seen": 48996896,
      "step": 84385
    },
    {
      "epoch": 12.569258266309205,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.010943238149413087,
      "loss": 0.7846,
      "num_input_tokens_seen": 48999552,
      "step": 84390
    },
    {
      "epoch": 12.570002978850164,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.010941361186966102,
      "loss": 0.785,
      "num_input_tokens_seen": 49002400,
      "step": 84395
    },
    {
      "epoch": 12.570747691391123,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.010939484293085689,
      "loss": 0.7924,
      "num_input_tokens_seen": 49005056,
      "step": 84400
    },
    {
      "epoch": 12.571492403932082,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.010937607467803567,
      "loss": 0.7893,
      "num_input_tokens_seen": 49007968,
      "step": 84405
    },
    {
      "epoch": 12.572237116473042,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.010935730711151436,
      "loss": 0.793,
      "num_input_tokens_seen": 49010720,
      "step": 84410
    },
    {
      "epoch": 12.572981829014001,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.010933854023161006,
      "loss": 0.7789,
      "num_input_tokens_seen": 49013376,
      "step": 84415
    },
    {
      "epoch": 12.57372654155496,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.010931977403863974,
      "loss": 0.8128,
      "num_input_tokens_seen": 49015904,
      "step": 84420
    },
    {
      "epoch": 12.574471254095918,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.010930100853292057,
      "loss": 0.8033,
      "num_input_tokens_seen": 49018592,
      "step": 84425
    },
    {
      "epoch": 12.575215966636879,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.010928224371476947,
      "loss": 0.789,
      "num_input_tokens_seen": 49021376,
      "step": 84430
    },
    {
      "epoch": 12.575960679177838,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.010926347958450341,
      "loss": 0.789,
      "num_input_tokens_seen": 49024192,
      "step": 84435
    },
    {
      "epoch": 12.576705391718797,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.010924471614243954,
      "loss": 0.7806,
      "num_input_tokens_seen": 49027296,
      "step": 84440
    },
    {
      "epoch": 12.577450104259755,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.010922595338889476,
      "loss": 0.782,
      "num_input_tokens_seen": 49030112,
      "step": 84445
    },
    {
      "epoch": 12.578194816800714,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.010920719132418604,
      "loss": 0.7849,
      "num_input_tokens_seen": 49032832,
      "step": 84450
    },
    {
      "epoch": 12.578939529341675,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.010918842994863034,
      "loss": 0.7709,
      "num_input_tokens_seen": 49036064,
      "step": 84455
    },
    {
      "epoch": 12.579684241882633,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.010916966926254467,
      "loss": 0.7859,
      "num_input_tokens_seen": 49038944,
      "step": 84460
    },
    {
      "epoch": 12.580428954423592,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.010915090926624595,
      "loss": 0.8235,
      "num_input_tokens_seen": 49041856,
      "step": 84465
    },
    {
      "epoch": 12.58117366696455,
      "grad_norm": 0.05078125,
      "learning_rate": 0.010913214996005106,
      "loss": 0.8058,
      "num_input_tokens_seen": 49044832,
      "step": 84470
    },
    {
      "epoch": 12.581918379505511,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0109113391344277,
      "loss": 0.7756,
      "num_input_tokens_seen": 49047872,
      "step": 84475
    },
    {
      "epoch": 12.58266309204647,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.010909463341924063,
      "loss": 0.8146,
      "num_input_tokens_seen": 49050688,
      "step": 84480
    },
    {
      "epoch": 12.583407804587429,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.010907587618525885,
      "loss": 0.7943,
      "num_input_tokens_seen": 49053376,
      "step": 84485
    },
    {
      "epoch": 12.584152517128388,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.01090571196426485,
      "loss": 0.8025,
      "num_input_tokens_seen": 49056096,
      "step": 84490
    },
    {
      "epoch": 12.584897229669348,
      "grad_norm": 0.044921875,
      "learning_rate": 0.010903836379172657,
      "loss": 0.7897,
      "num_input_tokens_seen": 49058944,
      "step": 84495
    },
    {
      "epoch": 12.585641942210307,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.010901960863280982,
      "loss": 0.8101,
      "num_input_tokens_seen": 49062080,
      "step": 84500
    },
    {
      "epoch": 12.586386654751266,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.010900085416621512,
      "loss": 0.8065,
      "num_input_tokens_seen": 49064736,
      "step": 84505
    },
    {
      "epoch": 12.587131367292224,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.010898210039225937,
      "loss": 0.8177,
      "num_input_tokens_seen": 49067584,
      "step": 84510
    },
    {
      "epoch": 12.587876079833185,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.01089633473112593,
      "loss": 0.8051,
      "num_input_tokens_seen": 49070400,
      "step": 84515
    },
    {
      "epoch": 12.588620792374144,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.01089445949235318,
      "loss": 0.7864,
      "num_input_tokens_seen": 49073312,
      "step": 84520
    },
    {
      "epoch": 12.589365504915103,
      "grad_norm": 0.041015625,
      "learning_rate": 0.010892584322939358,
      "loss": 0.8627,
      "num_input_tokens_seen": 49076224,
      "step": 84525
    },
    {
      "epoch": 12.590110217456061,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.010890709222916156,
      "loss": 0.7953,
      "num_input_tokens_seen": 49079072,
      "step": 84530
    },
    {
      "epoch": 12.590854929997022,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.010888834192315242,
      "loss": 0.7931,
      "num_input_tokens_seen": 49082016,
      "step": 84535
    },
    {
      "epoch": 12.59159964253798,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0108869592311683,
      "loss": 0.7874,
      "num_input_tokens_seen": 49085216,
      "step": 84540
    },
    {
      "epoch": 12.59234435507894,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.010885084339506997,
      "loss": 0.7939,
      "num_input_tokens_seen": 49087968,
      "step": 84545
    },
    {
      "epoch": 12.593089067619898,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.010883209517363013,
      "loss": 0.8191,
      "num_input_tokens_seen": 49091232,
      "step": 84550
    },
    {
      "epoch": 12.593833780160859,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.01088133476476802,
      "loss": 0.8146,
      "num_input_tokens_seen": 49094368,
      "step": 84555
    },
    {
      "epoch": 12.594578492701817,
      "grad_norm": 0.03125,
      "learning_rate": 0.010879460081753688,
      "loss": 0.7991,
      "num_input_tokens_seen": 49097184,
      "step": 84560
    },
    {
      "epoch": 12.595323205242776,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.010877585468351693,
      "loss": 0.7902,
      "num_input_tokens_seen": 49100032,
      "step": 84565
    },
    {
      "epoch": 12.596067917783735,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.0108757109245937,
      "loss": 0.8008,
      "num_input_tokens_seen": 49102848,
      "step": 84570
    },
    {
      "epoch": 12.596812630324695,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.010873836450511383,
      "loss": 0.7928,
      "num_input_tokens_seen": 49105632,
      "step": 84575
    },
    {
      "epoch": 12.597557342865654,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0108719620461364,
      "loss": 0.789,
      "num_input_tokens_seen": 49108352,
      "step": 84580
    },
    {
      "epoch": 12.598302055406613,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.010870087711500425,
      "loss": 0.8012,
      "num_input_tokens_seen": 49111264,
      "step": 84585
    },
    {
      "epoch": 12.599046767947572,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.010868213446635125,
      "loss": 0.7969,
      "num_input_tokens_seen": 49114144,
      "step": 84590
    },
    {
      "epoch": 12.599791480488532,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.01086633925157215,
      "loss": 0.818,
      "num_input_tokens_seen": 49117120,
      "step": 84595
    },
    {
      "epoch": 12.600536193029491,
      "grad_norm": 0.0234375,
      "learning_rate": 0.01086446512634318,
      "loss": 0.7997,
      "num_input_tokens_seen": 49119904,
      "step": 84600
    },
    {
      "epoch": 12.60128090557045,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.010862591070979866,
      "loss": 0.7926,
      "num_input_tokens_seen": 49123008,
      "step": 84605
    },
    {
      "epoch": 12.602025618111409,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.010860717085513873,
      "loss": 0.7846,
      "num_input_tokens_seen": 49125920,
      "step": 84610
    },
    {
      "epoch": 12.602770330652369,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.010858843169976852,
      "loss": 0.7814,
      "num_input_tokens_seen": 49128608,
      "step": 84615
    },
    {
      "epoch": 12.603515043193328,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.010856969324400473,
      "loss": 0.7969,
      "num_input_tokens_seen": 49131456,
      "step": 84620
    },
    {
      "epoch": 12.604259755734287,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.010855095548816378,
      "loss": 0.7891,
      "num_input_tokens_seen": 49134144,
      "step": 84625
    },
    {
      "epoch": 12.605004468275245,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.01085322184325624,
      "loss": 0.7925,
      "num_input_tokens_seen": 49136832,
      "step": 84630
    },
    {
      "epoch": 12.605749180816204,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.010851348207751703,
      "loss": 0.8018,
      "num_input_tokens_seen": 49139488,
      "step": 84635
    },
    {
      "epoch": 12.606493893357165,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.01084947464233442,
      "loss": 0.8052,
      "num_input_tokens_seen": 49142432,
      "step": 84640
    },
    {
      "epoch": 12.607238605898123,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.010847601147036045,
      "loss": 0.7969,
      "num_input_tokens_seen": 49145568,
      "step": 84645
    },
    {
      "epoch": 12.607983318439082,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.010845727721888224,
      "loss": 0.813,
      "num_input_tokens_seen": 49148416,
      "step": 84650
    },
    {
      "epoch": 12.608728030980041,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.010843854366922617,
      "loss": 0.7947,
      "num_input_tokens_seen": 49151200,
      "step": 84655
    },
    {
      "epoch": 12.609472743521001,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.01084198108217086,
      "loss": 0.7725,
      "num_input_tokens_seen": 49154048,
      "step": 84660
    },
    {
      "epoch": 12.61021745606196,
      "grad_norm": 0.13671875,
      "learning_rate": 0.010840107867664612,
      "loss": 0.853,
      "num_input_tokens_seen": 49156768,
      "step": 84665
    },
    {
      "epoch": 12.610962168602919,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.010838234723435515,
      "loss": 0.7957,
      "num_input_tokens_seen": 49159328,
      "step": 84670
    },
    {
      "epoch": 12.611706881143878,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.01083636164951521,
      "loss": 0.7981,
      "num_input_tokens_seen": 49162368,
      "step": 84675
    },
    {
      "epoch": 12.612451593684838,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.010834488645935343,
      "loss": 0.8172,
      "num_input_tokens_seen": 49165056,
      "step": 84680
    },
    {
      "epoch": 12.613196306225797,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.010832615712727554,
      "loss": 0.7824,
      "num_input_tokens_seen": 49167840,
      "step": 84685
    },
    {
      "epoch": 12.613941018766756,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.010830742849923492,
      "loss": 0.8016,
      "num_input_tokens_seen": 49170720,
      "step": 84690
    },
    {
      "epoch": 12.614685731307715,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.010828870057554785,
      "loss": 0.7895,
      "num_input_tokens_seen": 49173600,
      "step": 84695
    },
    {
      "epoch": 12.615430443848675,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.010826997335653087,
      "loss": 0.8075,
      "num_input_tokens_seen": 49176512,
      "step": 84700
    },
    {
      "epoch": 12.616175156389634,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.010825124684250023,
      "loss": 0.7801,
      "num_input_tokens_seen": 49179360,
      "step": 84705
    },
    {
      "epoch": 12.616919868930593,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.010823252103377238,
      "loss": 0.8038,
      "num_input_tokens_seen": 49182528,
      "step": 84710
    },
    {
      "epoch": 12.617664581471551,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.010821379593066362,
      "loss": 0.7916,
      "num_input_tokens_seen": 49185312,
      "step": 84715
    },
    {
      "epoch": 12.618409294012512,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.010819507153349027,
      "loss": 0.773,
      "num_input_tokens_seen": 49188192,
      "step": 84720
    },
    {
      "epoch": 12.61915400655347,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.010817634784256876,
      "loss": 0.7906,
      "num_input_tokens_seen": 49191232,
      "step": 84725
    },
    {
      "epoch": 12.61989871909443,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.010815762485821527,
      "loss": 0.854,
      "num_input_tokens_seen": 49194112,
      "step": 84730
    },
    {
      "epoch": 12.620643431635388,
      "grad_norm": 0.041015625,
      "learning_rate": 0.010813890258074624,
      "loss": 0.7834,
      "num_input_tokens_seen": 49196960,
      "step": 84735
    },
    {
      "epoch": 12.621388144176349,
      "grad_norm": 0.02734375,
      "learning_rate": 0.010812018101047788,
      "loss": 0.8044,
      "num_input_tokens_seen": 49199808,
      "step": 84740
    },
    {
      "epoch": 12.622132856717307,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.010810146014772655,
      "loss": 0.8074,
      "num_input_tokens_seen": 49202976,
      "step": 84745
    },
    {
      "epoch": 12.622877569258266,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.010808273999280844,
      "loss": 0.7901,
      "num_input_tokens_seen": 49205888,
      "step": 84750
    },
    {
      "epoch": 12.623622281799225,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.010806402054603977,
      "loss": 0.8034,
      "num_input_tokens_seen": 49208672,
      "step": 84755
    },
    {
      "epoch": 12.624366994340185,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.010804530180773694,
      "loss": 0.7946,
      "num_input_tokens_seen": 49211360,
      "step": 84760
    },
    {
      "epoch": 12.625111706881144,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.010802658377821602,
      "loss": 0.8049,
      "num_input_tokens_seen": 49214144,
      "step": 84765
    },
    {
      "epoch": 12.625856419422103,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.010800786645779336,
      "loss": 0.7959,
      "num_input_tokens_seen": 49217216,
      "step": 84770
    },
    {
      "epoch": 12.626601131963062,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.01079891498467851,
      "loss": 0.7878,
      "num_input_tokens_seen": 49219936,
      "step": 84775
    },
    {
      "epoch": 12.62734584450402,
      "grad_norm": 0.0390625,
      "learning_rate": 0.010797043394550749,
      "loss": 0.7916,
      "num_input_tokens_seen": 49222912,
      "step": 84780
    },
    {
      "epoch": 12.628090557044981,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.01079517187542766,
      "loss": 0.7947,
      "num_input_tokens_seen": 49225632,
      "step": 84785
    },
    {
      "epoch": 12.62883526958594,
      "grad_norm": 0.076171875,
      "learning_rate": 0.010793300427340875,
      "loss": 0.8144,
      "num_input_tokens_seen": 49228416,
      "step": 84790
    },
    {
      "epoch": 12.629579982126899,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.010791429050322003,
      "loss": 0.8061,
      "num_input_tokens_seen": 49231392,
      "step": 84795
    },
    {
      "epoch": 12.63032469466786,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.010789557744402656,
      "loss": 0.7969,
      "num_input_tokens_seen": 49234080,
      "step": 84800
    },
    {
      "epoch": 12.631069407208818,
      "grad_norm": 0.03125,
      "learning_rate": 0.010787686509614456,
      "loss": 0.7865,
      "num_input_tokens_seen": 49237024,
      "step": 84805
    },
    {
      "epoch": 12.631814119749777,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.01078581534598901,
      "loss": 0.7954,
      "num_input_tokens_seen": 49240000,
      "step": 84810
    },
    {
      "epoch": 12.632558832290735,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.010783944253557933,
      "loss": 0.7997,
      "num_input_tokens_seen": 49242752,
      "step": 84815
    },
    {
      "epoch": 12.633303544831694,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.010782073232352826,
      "loss": 0.8036,
      "num_input_tokens_seen": 49245632,
      "step": 84820
    },
    {
      "epoch": 12.634048257372655,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.010780202282405312,
      "loss": 0.7985,
      "num_input_tokens_seen": 49248480,
      "step": 84825
    },
    {
      "epoch": 12.634792969913613,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.010778331403746991,
      "loss": 0.7918,
      "num_input_tokens_seen": 49251584,
      "step": 84830
    },
    {
      "epoch": 12.635537682454572,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.010776460596409465,
      "loss": 0.7904,
      "num_input_tokens_seen": 49254432,
      "step": 84835
    },
    {
      "epoch": 12.636282394995531,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.01077458986042435,
      "loss": 0.7997,
      "num_input_tokens_seen": 49257280,
      "step": 84840
    },
    {
      "epoch": 12.637027107536491,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.010772719195823241,
      "loss": 0.8117,
      "num_input_tokens_seen": 49260544,
      "step": 84845
    },
    {
      "epoch": 12.63777182007745,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.010770848602637748,
      "loss": 0.8057,
      "num_input_tokens_seen": 49263232,
      "step": 84850
    },
    {
      "epoch": 12.638516532618409,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.010768978080899464,
      "loss": 0.7825,
      "num_input_tokens_seen": 49266464,
      "step": 84855
    },
    {
      "epoch": 12.639261245159368,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.01076710763064,
      "loss": 0.7975,
      "num_input_tokens_seen": 49269344,
      "step": 84860
    },
    {
      "epoch": 12.640005957700328,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.01076523725189095,
      "loss": 0.7867,
      "num_input_tokens_seen": 49272096,
      "step": 84865
    },
    {
      "epoch": 12.640750670241287,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.010763366944683913,
      "loss": 0.8184,
      "num_input_tokens_seen": 49274944,
      "step": 84870
    },
    {
      "epoch": 12.641495382782246,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.010761496709050484,
      "loss": 0.8212,
      "num_input_tokens_seen": 49277440,
      "step": 84875
    },
    {
      "epoch": 12.642240095323205,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.01075962654502226,
      "loss": 0.8061,
      "num_input_tokens_seen": 49280128,
      "step": 84880
    },
    {
      "epoch": 12.642984807864165,
      "grad_norm": 0.048828125,
      "learning_rate": 0.01075775645263084,
      "loss": 0.8088,
      "num_input_tokens_seen": 49283040,
      "step": 84885
    },
    {
      "epoch": 12.643729520405124,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.010755886431907804,
      "loss": 0.7744,
      "num_input_tokens_seen": 49285632,
      "step": 84890
    },
    {
      "epoch": 12.644474232946083,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.010754016482884762,
      "loss": 0.7759,
      "num_input_tokens_seen": 49288448,
      "step": 84895
    },
    {
      "epoch": 12.645218945487041,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.01075214660559329,
      "loss": 0.8027,
      "num_input_tokens_seen": 49291552,
      "step": 84900
    },
    {
      "epoch": 12.645963658028002,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.01075027680006499,
      "loss": 0.8063,
      "num_input_tokens_seen": 49294528,
      "step": 84905
    },
    {
      "epoch": 12.64670837056896,
      "grad_norm": 0.04296875,
      "learning_rate": 0.010748407066331444,
      "loss": 0.8076,
      "num_input_tokens_seen": 49297600,
      "step": 84910
    },
    {
      "epoch": 12.64745308310992,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.01074653740442423,
      "loss": 0.788,
      "num_input_tokens_seen": 49300512,
      "step": 84915
    },
    {
      "epoch": 12.648197795650878,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.010744667814374952,
      "loss": 0.801,
      "num_input_tokens_seen": 49303168,
      "step": 84920
    },
    {
      "epoch": 12.648942508191839,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.01074279829621518,
      "loss": 0.7872,
      "num_input_tokens_seen": 49305888,
      "step": 84925
    },
    {
      "epoch": 12.649687220732797,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.010740928849976513,
      "loss": 0.7884,
      "num_input_tokens_seen": 49308640,
      "step": 84930
    },
    {
      "epoch": 12.650431933273756,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.010739059475690519,
      "loss": 0.8162,
      "num_input_tokens_seen": 49311264,
      "step": 84935
    },
    {
      "epoch": 12.651176645814715,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.010737190173388787,
      "loss": 0.8006,
      "num_input_tokens_seen": 49313952,
      "step": 84940
    },
    {
      "epoch": 12.651921358355676,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.010735320943102887,
      "loss": 0.8129,
      "num_input_tokens_seen": 49316928,
      "step": 84945
    },
    {
      "epoch": 12.652666070896634,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.010733451784864415,
      "loss": 0.8576,
      "num_input_tokens_seen": 49319840,
      "step": 84950
    },
    {
      "epoch": 12.653410783437593,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.010731582698704936,
      "loss": 0.8088,
      "num_input_tokens_seen": 49322976,
      "step": 84955
    },
    {
      "epoch": 12.654155495978552,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.010729713684656023,
      "loss": 0.7929,
      "num_input_tokens_seen": 49326336,
      "step": 84960
    },
    {
      "epoch": 12.65490020851951,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.010727844742749265,
      "loss": 0.8022,
      "num_input_tokens_seen": 49329344,
      "step": 84965
    },
    {
      "epoch": 12.655644921060471,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.010725975873016224,
      "loss": 0.7845,
      "num_input_tokens_seen": 49332352,
      "step": 84970
    },
    {
      "epoch": 12.65638963360143,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.01072410707548848,
      "loss": 0.7947,
      "num_input_tokens_seen": 49335328,
      "step": 84975
    },
    {
      "epoch": 12.657134346142389,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.010722238350197593,
      "loss": 0.8014,
      "num_input_tokens_seen": 49338016,
      "step": 84980
    },
    {
      "epoch": 12.657879058683347,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.010720369697175151,
      "loss": 0.7765,
      "num_input_tokens_seen": 49340864,
      "step": 84985
    },
    {
      "epoch": 12.658623771224308,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.010718501116452712,
      "loss": 0.8065,
      "num_input_tokens_seen": 49343616,
      "step": 84990
    },
    {
      "epoch": 12.659368483765267,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.010716632608061837,
      "loss": 0.7964,
      "num_input_tokens_seen": 49346816,
      "step": 84995
    },
    {
      "epoch": 12.660113196306225,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.010714764172034109,
      "loss": 0.7874,
      "num_input_tokens_seen": 49349856,
      "step": 85000
    },
    {
      "epoch": 12.660857908847184,
      "grad_norm": 0.0234375,
      "learning_rate": 0.010712895808401082,
      "loss": 0.8053,
      "num_input_tokens_seen": 49352768,
      "step": 85005
    },
    {
      "epoch": 12.661602621388145,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.010711027517194325,
      "loss": 0.8052,
      "num_input_tokens_seen": 49355648,
      "step": 85010
    },
    {
      "epoch": 12.662347333929103,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.010709159298445394,
      "loss": 0.8027,
      "num_input_tokens_seen": 49358624,
      "step": 85015
    },
    {
      "epoch": 12.663092046470062,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.01070729115218586,
      "loss": 0.8087,
      "num_input_tokens_seen": 49361600,
      "step": 85020
    },
    {
      "epoch": 12.663836759011021,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.010705423078447275,
      "loss": 0.8045,
      "num_input_tokens_seen": 49364576,
      "step": 85025
    },
    {
      "epoch": 12.664581471551982,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.010703555077261208,
      "loss": 0.7863,
      "num_input_tokens_seen": 49367584,
      "step": 85030
    },
    {
      "epoch": 12.66532618409294,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.01070168714865921,
      "loss": 0.791,
      "num_input_tokens_seen": 49370592,
      "step": 85035
    },
    {
      "epoch": 12.666070896633899,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.010699819292672838,
      "loss": 0.801,
      "num_input_tokens_seen": 49373344,
      "step": 85040
    },
    {
      "epoch": 12.666815609174858,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.01069795150933365,
      "loss": 0.7799,
      "num_input_tokens_seen": 49376416,
      "step": 85045
    },
    {
      "epoch": 12.667560321715818,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.01069608379867319,
      "loss": 0.7873,
      "num_input_tokens_seen": 49379456,
      "step": 85050
    },
    {
      "epoch": 12.668305034256777,
      "grad_norm": 0.05859375,
      "learning_rate": 0.010694216160723032,
      "loss": 0.7846,
      "num_input_tokens_seen": 49382496,
      "step": 85055
    },
    {
      "epoch": 12.669049746797736,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.010692348595514706,
      "loss": 0.8322,
      "num_input_tokens_seen": 49385248,
      "step": 85060
    },
    {
      "epoch": 12.669794459338695,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.010690481103079779,
      "loss": 0.8045,
      "num_input_tokens_seen": 49388448,
      "step": 85065
    },
    {
      "epoch": 12.670539171879655,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.010688613683449795,
      "loss": 0.8075,
      "num_input_tokens_seen": 49391168,
      "step": 85070
    },
    {
      "epoch": 12.671283884420614,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.010686746336656295,
      "loss": 0.8115,
      "num_input_tokens_seen": 49394016,
      "step": 85075
    },
    {
      "epoch": 12.672028596961573,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.010684879062730837,
      "loss": 0.8094,
      "num_input_tokens_seen": 49396704,
      "step": 85080
    },
    {
      "epoch": 12.672773309502531,
      "grad_norm": 0.04296875,
      "learning_rate": 0.010683011861704957,
      "loss": 0.7961,
      "num_input_tokens_seen": 49399616,
      "step": 85085
    },
    {
      "epoch": 12.673518022043492,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.010681144733610207,
      "loss": 0.8101,
      "num_input_tokens_seen": 49402208,
      "step": 85090
    },
    {
      "epoch": 12.67426273458445,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.010679277678478124,
      "loss": 0.7983,
      "num_input_tokens_seen": 49405216,
      "step": 85095
    },
    {
      "epoch": 12.67500744712541,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.010677410696340256,
      "loss": 0.7828,
      "num_input_tokens_seen": 49408128,
      "step": 85100
    },
    {
      "epoch": 12.675752159666368,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.01067554378722814,
      "loss": 0.7999,
      "num_input_tokens_seen": 49410912,
      "step": 85105
    },
    {
      "epoch": 12.676496872207329,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.01067367695117332,
      "loss": 0.7996,
      "num_input_tokens_seen": 49413568,
      "step": 85110
    },
    {
      "epoch": 12.677241584748288,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.01067181018820733,
      "loss": 0.7992,
      "num_input_tokens_seen": 49416352,
      "step": 85115
    },
    {
      "epoch": 12.677986297289246,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.010669943498361702,
      "loss": 0.7896,
      "num_input_tokens_seen": 49419200,
      "step": 85120
    },
    {
      "epoch": 12.678731009830205,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.010668076881667982,
      "loss": 0.7961,
      "num_input_tokens_seen": 49422336,
      "step": 85125
    },
    {
      "epoch": 12.679475722371166,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.010666210338157697,
      "loss": 0.7963,
      "num_input_tokens_seen": 49425056,
      "step": 85130
    },
    {
      "epoch": 12.680220434912124,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.01066434386786239,
      "loss": 0.785,
      "num_input_tokens_seen": 49428096,
      "step": 85135
    },
    {
      "epoch": 12.680965147453083,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.010662477470813581,
      "loss": 0.7942,
      "num_input_tokens_seen": 49431040,
      "step": 85140
    },
    {
      "epoch": 12.681709859994042,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.01066061114704281,
      "loss": 0.8158,
      "num_input_tokens_seen": 49433888,
      "step": 85145
    },
    {
      "epoch": 12.682454572535,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.010658744896581605,
      "loss": 0.8046,
      "num_input_tokens_seen": 49436864,
      "step": 85150
    },
    {
      "epoch": 12.683199285075961,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.010656878719461487,
      "loss": 0.7757,
      "num_input_tokens_seen": 49439968,
      "step": 85155
    },
    {
      "epoch": 12.68394399761692,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.010655012615713996,
      "loss": 0.7897,
      "num_input_tokens_seen": 49442944,
      "step": 85160
    },
    {
      "epoch": 12.684688710157879,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.01065314658537064,
      "loss": 0.8051,
      "num_input_tokens_seen": 49445888,
      "step": 85165
    },
    {
      "epoch": 12.685433422698837,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.010651280628462965,
      "loss": 0.81,
      "num_input_tokens_seen": 49449216,
      "step": 85170
    },
    {
      "epoch": 12.686178135239798,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.010649414745022479,
      "loss": 0.7999,
      "num_input_tokens_seen": 49452160,
      "step": 85175
    },
    {
      "epoch": 12.686922847780757,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.010647548935080714,
      "loss": 0.7996,
      "num_input_tokens_seen": 49455008,
      "step": 85180
    },
    {
      "epoch": 12.687667560321715,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.010645683198669178,
      "loss": 0.7917,
      "num_input_tokens_seen": 49457824,
      "step": 85185
    },
    {
      "epoch": 12.688412272862674,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.010643817535819404,
      "loss": 0.7847,
      "num_input_tokens_seen": 49460736,
      "step": 85190
    },
    {
      "epoch": 12.689156985403635,
      "grad_norm": 0.048828125,
      "learning_rate": 0.010641951946562906,
      "loss": 0.7908,
      "num_input_tokens_seen": 49463744,
      "step": 85195
    },
    {
      "epoch": 12.689901697944594,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.010640086430931199,
      "loss": 0.7836,
      "num_input_tokens_seen": 49466528,
      "step": 85200
    },
    {
      "epoch": 12.690646410485552,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0106382209889558,
      "loss": 0.7742,
      "num_input_tokens_seen": 49469632,
      "step": 85205
    },
    {
      "epoch": 12.691391123026511,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.010636355620668224,
      "loss": 0.7914,
      "num_input_tokens_seen": 49472608,
      "step": 85210
    },
    {
      "epoch": 12.692135835567472,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.010634490326099987,
      "loss": 0.8143,
      "num_input_tokens_seen": 49475712,
      "step": 85215
    },
    {
      "epoch": 12.69288054810843,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.010632625105282595,
      "loss": 0.8546,
      "num_input_tokens_seen": 49478688,
      "step": 85220
    },
    {
      "epoch": 12.69362526064939,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.010630759958247568,
      "loss": 0.8072,
      "num_input_tokens_seen": 49481632,
      "step": 85225
    },
    {
      "epoch": 12.694369973190348,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.010628894885026406,
      "loss": 0.8075,
      "num_input_tokens_seen": 49484672,
      "step": 85230
    },
    {
      "epoch": 12.695114685731308,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.010627029885650629,
      "loss": 0.7934,
      "num_input_tokens_seen": 49487680,
      "step": 85235
    },
    {
      "epoch": 12.695859398272267,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.010625164960151735,
      "loss": 0.793,
      "num_input_tokens_seen": 49490752,
      "step": 85240
    },
    {
      "epoch": 12.696604110813226,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.010623300108561228,
      "loss": 0.7825,
      "num_input_tokens_seen": 49493536,
      "step": 85245
    },
    {
      "epoch": 12.697348823354185,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.010621435330910622,
      "loss": 0.8019,
      "num_input_tokens_seen": 49496640,
      "step": 85250
    },
    {
      "epoch": 12.698093535895145,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.01061957062723141,
      "loss": 0.7871,
      "num_input_tokens_seen": 49499616,
      "step": 85255
    },
    {
      "epoch": 12.698838248436104,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.010617705997555106,
      "loss": 0.8073,
      "num_input_tokens_seen": 49502400,
      "step": 85260
    },
    {
      "epoch": 12.699582960977063,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.010615841441913205,
      "loss": 0.7675,
      "num_input_tokens_seen": 49505120,
      "step": 85265
    },
    {
      "epoch": 12.700327673518021,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.010613976960337206,
      "loss": 0.7859,
      "num_input_tokens_seen": 49508032,
      "step": 85270
    },
    {
      "epoch": 12.701072386058982,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.010612112552858611,
      "loss": 0.7825,
      "num_input_tokens_seen": 49510944,
      "step": 85275
    },
    {
      "epoch": 12.70181709859994,
      "grad_norm": 0.056640625,
      "learning_rate": 0.010610248219508909,
      "loss": 0.8194,
      "num_input_tokens_seen": 49513632,
      "step": 85280
    },
    {
      "epoch": 12.7025618111409,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.010608383960319606,
      "loss": 0.8,
      "num_input_tokens_seen": 49516320,
      "step": 85285
    },
    {
      "epoch": 12.703306523681858,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.010606519775322187,
      "loss": 0.7761,
      "num_input_tokens_seen": 49519232,
      "step": 85290
    },
    {
      "epoch": 12.704051236222817,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.010604655664548159,
      "loss": 0.8307,
      "num_input_tokens_seen": 49521952,
      "step": 85295
    },
    {
      "epoch": 12.704795948763778,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.010602791628029,
      "loss": 0.7945,
      "num_input_tokens_seen": 49525088,
      "step": 85300
    },
    {
      "epoch": 12.705540661304736,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.010600927665796215,
      "loss": 0.8077,
      "num_input_tokens_seen": 49528032,
      "step": 85305
    },
    {
      "epoch": 12.706285373845695,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.010599063777881276,
      "loss": 0.8145,
      "num_input_tokens_seen": 49530880,
      "step": 85310
    },
    {
      "epoch": 12.707030086386656,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.01059719996431569,
      "loss": 0.7787,
      "num_input_tokens_seen": 49533728,
      "step": 85315
    },
    {
      "epoch": 12.707774798927614,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.010595336225130933,
      "loss": 0.7928,
      "num_input_tokens_seen": 49536416,
      "step": 85320
    },
    {
      "epoch": 12.708519511468573,
      "grad_norm": 0.03515625,
      "learning_rate": 0.01059347256035849,
      "loss": 0.795,
      "num_input_tokens_seen": 49539488,
      "step": 85325
    },
    {
      "epoch": 12.709264224009532,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.010591608970029856,
      "loss": 0.8101,
      "num_input_tokens_seen": 49542272,
      "step": 85330
    },
    {
      "epoch": 12.71000893655049,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.010589745454176503,
      "loss": 0.794,
      "num_input_tokens_seen": 49545024,
      "step": 85335
    },
    {
      "epoch": 12.710753649091451,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.010587882012829921,
      "loss": 0.8009,
      "num_input_tokens_seen": 49548160,
      "step": 85340
    },
    {
      "epoch": 12.71149836163241,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.010586018646021583,
      "loss": 0.7943,
      "num_input_tokens_seen": 49551008,
      "step": 85345
    },
    {
      "epoch": 12.712243074173369,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.010584155353782982,
      "loss": 0.803,
      "num_input_tokens_seen": 49553600,
      "step": 85350
    },
    {
      "epoch": 12.712987786714327,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.010582292136145584,
      "loss": 0.7984,
      "num_input_tokens_seen": 49556832,
      "step": 85355
    },
    {
      "epoch": 12.713732499255288,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.010580428993140868,
      "loss": 0.7779,
      "num_input_tokens_seen": 49560320,
      "step": 85360
    },
    {
      "epoch": 12.714477211796247,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.010578565924800316,
      "loss": 0.788,
      "num_input_tokens_seen": 49563264,
      "step": 85365
    },
    {
      "epoch": 12.715221924337206,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.010576702931155396,
      "loss": 0.7755,
      "num_input_tokens_seen": 49566656,
      "step": 85370
    },
    {
      "epoch": 12.715966636878164,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.01057484001223759,
      "loss": 0.8025,
      "num_input_tokens_seen": 49569312,
      "step": 85375
    },
    {
      "epoch": 12.716711349419125,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.010572977168078356,
      "loss": 0.7796,
      "num_input_tokens_seen": 49572096,
      "step": 85380
    },
    {
      "epoch": 12.717456061960084,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.01057111439870918,
      "loss": 0.7754,
      "num_input_tokens_seen": 49574976,
      "step": 85385
    },
    {
      "epoch": 12.718200774501042,
      "grad_norm": 0.07421875,
      "learning_rate": 0.010569251704161519,
      "loss": 0.7725,
      "num_input_tokens_seen": 49577632,
      "step": 85390
    },
    {
      "epoch": 12.718945487042001,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.010567389084466854,
      "loss": 0.8485,
      "num_input_tokens_seen": 49580576,
      "step": 85395
    },
    {
      "epoch": 12.719690199582962,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.010565526539656644,
      "loss": 0.7883,
      "num_input_tokens_seen": 49583584,
      "step": 85400
    },
    {
      "epoch": 12.72043491212392,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.010563664069762353,
      "loss": 0.7568,
      "num_input_tokens_seen": 49586560,
      "step": 85405
    },
    {
      "epoch": 12.72117962466488,
      "grad_norm": 0.0703125,
      "learning_rate": 0.010561801674815454,
      "loss": 0.795,
      "num_input_tokens_seen": 49589056,
      "step": 85410
    },
    {
      "epoch": 12.721924337205838,
      "grad_norm": 0.072265625,
      "learning_rate": 0.010559939354847395,
      "loss": 0.7942,
      "num_input_tokens_seen": 49592096,
      "step": 85415
    },
    {
      "epoch": 12.722669049746798,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.010558077109889655,
      "loss": 0.8615,
      "num_input_tokens_seen": 49595072,
      "step": 85420
    },
    {
      "epoch": 12.723413762287757,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.010556214939973685,
      "loss": 0.8015,
      "num_input_tokens_seen": 49597792,
      "step": 85425
    },
    {
      "epoch": 12.724158474828716,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.010554352845130949,
      "loss": 0.7806,
      "num_input_tokens_seen": 49600832,
      "step": 85430
    },
    {
      "epoch": 12.724903187369675,
      "grad_norm": 0.0859375,
      "learning_rate": 0.010552490825392906,
      "loss": 0.8155,
      "num_input_tokens_seen": 49603488,
      "step": 85435
    },
    {
      "epoch": 12.725647899910635,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.010550628880791005,
      "loss": 0.7792,
      "num_input_tokens_seen": 49606720,
      "step": 85440
    },
    {
      "epoch": 12.726392612451594,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.010548767011356713,
      "loss": 0.8078,
      "num_input_tokens_seen": 49609664,
      "step": 85445
    },
    {
      "epoch": 12.727137324992553,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.01054690521712147,
      "loss": 0.7716,
      "num_input_tokens_seen": 49612480,
      "step": 85450
    },
    {
      "epoch": 12.727882037533512,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.010545043498116746,
      "loss": 0.7967,
      "num_input_tokens_seen": 49615488,
      "step": 85455
    },
    {
      "epoch": 12.728626750074472,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.010543181854373975,
      "loss": 0.7942,
      "num_input_tokens_seen": 49618176,
      "step": 85460
    },
    {
      "epoch": 12.72937146261543,
      "grad_norm": 0.052734375,
      "learning_rate": 0.010541320285924624,
      "loss": 0.8144,
      "num_input_tokens_seen": 49620864,
      "step": 85465
    },
    {
      "epoch": 12.73011617515639,
      "grad_norm": 0.0625,
      "learning_rate": 0.010539458792800133,
      "loss": 0.8029,
      "num_input_tokens_seen": 49623648,
      "step": 85470
    },
    {
      "epoch": 12.730860887697348,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.010537597375031956,
      "loss": 0.7846,
      "num_input_tokens_seen": 49626912,
      "step": 85475
    },
    {
      "epoch": 12.731605600238307,
      "grad_norm": 0.0703125,
      "learning_rate": 0.010535736032651535,
      "loss": 0.8095,
      "num_input_tokens_seen": 49629856,
      "step": 85480
    },
    {
      "epoch": 12.732350312779268,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.01053387476569031,
      "loss": 0.8164,
      "num_input_tokens_seen": 49632768,
      "step": 85485
    },
    {
      "epoch": 12.733095025320226,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.010532013574179735,
      "loss": 0.8023,
      "num_input_tokens_seen": 49635680,
      "step": 85490
    },
    {
      "epoch": 12.733839737861185,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.010530152458151247,
      "loss": 0.7934,
      "num_input_tokens_seen": 49638336,
      "step": 85495
    },
    {
      "epoch": 12.734584450402146,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.010528291417636296,
      "loss": 0.8036,
      "num_input_tokens_seen": 49641216,
      "step": 85500
    },
    {
      "epoch": 12.735329162943104,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.010526430452666314,
      "loss": 0.8109,
      "num_input_tokens_seen": 49644544,
      "step": 85505
    },
    {
      "epoch": 12.736073875484063,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.010524569563272745,
      "loss": 0.793,
      "num_input_tokens_seen": 49647488,
      "step": 85510
    },
    {
      "epoch": 12.736818588025022,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.010522708749487025,
      "loss": 0.7982,
      "num_input_tokens_seen": 49650112,
      "step": 85515
    },
    {
      "epoch": 12.73756330056598,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.010520848011340584,
      "loss": 0.8061,
      "num_input_tokens_seen": 49652800,
      "step": 85520
    },
    {
      "epoch": 12.738308013106941,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.010518987348864868,
      "loss": 0.813,
      "num_input_tokens_seen": 49655680,
      "step": 85525
    },
    {
      "epoch": 12.7390527256479,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.010517126762091305,
      "loss": 0.7815,
      "num_input_tokens_seen": 49658496,
      "step": 85530
    },
    {
      "epoch": 12.739797438188859,
      "grad_norm": 0.041015625,
      "learning_rate": 0.01051526625105133,
      "loss": 0.841,
      "num_input_tokens_seen": 49661472,
      "step": 85535
    },
    {
      "epoch": 12.740542150729818,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.01051340581577637,
      "loss": 0.8007,
      "num_input_tokens_seen": 49664448,
      "step": 85540
    },
    {
      "epoch": 12.741286863270778,
      "grad_norm": 0.060546875,
      "learning_rate": 0.010511545456297866,
      "loss": 0.8085,
      "num_input_tokens_seen": 49667424,
      "step": 85545
    },
    {
      "epoch": 12.742031575811737,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.010509685172647231,
      "loss": 0.8014,
      "num_input_tokens_seen": 49670400,
      "step": 85550
    },
    {
      "epoch": 12.742776288352696,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.01050782496485591,
      "loss": 0.7981,
      "num_input_tokens_seen": 49673024,
      "step": 85555
    },
    {
      "epoch": 12.743521000893654,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.010505964832955319,
      "loss": 0.8024,
      "num_input_tokens_seen": 49676320,
      "step": 85560
    },
    {
      "epoch": 12.744265713434615,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01050410477697688,
      "loss": 0.8067,
      "num_input_tokens_seen": 49679360,
      "step": 85565
    },
    {
      "epoch": 12.745010425975574,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.010502244796952024,
      "loss": 0.8017,
      "num_input_tokens_seen": 49682176,
      "step": 85570
    },
    {
      "epoch": 12.745755138516532,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.010500384892912168,
      "loss": 0.7916,
      "num_input_tokens_seen": 49684928,
      "step": 85575
    },
    {
      "epoch": 12.746499851057491,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.01049852506488874,
      "loss": 0.8167,
      "num_input_tokens_seen": 49687808,
      "step": 85580
    },
    {
      "epoch": 12.747244563598452,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01049666531291315,
      "loss": 0.7873,
      "num_input_tokens_seen": 49690752,
      "step": 85585
    },
    {
      "epoch": 12.74798927613941,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01049480563701683,
      "loss": 0.7987,
      "num_input_tokens_seen": 49693696,
      "step": 85590
    },
    {
      "epoch": 12.74873398868037,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.01049294603723119,
      "loss": 0.7911,
      "num_input_tokens_seen": 49697024,
      "step": 85595
    },
    {
      "epoch": 12.749478701221328,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.01049108651358764,
      "loss": 0.773,
      "num_input_tokens_seen": 49699712,
      "step": 85600
    },
    {
      "epoch": 12.750223413762289,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.010489227066117606,
      "loss": 0.8121,
      "num_input_tokens_seen": 49702464,
      "step": 85605
    },
    {
      "epoch": 12.750968126303247,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.010487367694852489,
      "loss": 0.7849,
      "num_input_tokens_seen": 49705184,
      "step": 85610
    },
    {
      "epoch": 12.751712838844206,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.010485508399823717,
      "loss": 0.8287,
      "num_input_tokens_seen": 49708192,
      "step": 85615
    },
    {
      "epoch": 12.752457551385165,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.010483649181062683,
      "loss": 0.7967,
      "num_input_tokens_seen": 49710912,
      "step": 85620
    },
    {
      "epoch": 12.753202263926125,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.010481790038600815,
      "loss": 0.7786,
      "num_input_tokens_seen": 49713536,
      "step": 85625
    },
    {
      "epoch": 12.753946976467084,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.010479930972469506,
      "loss": 0.8001,
      "num_input_tokens_seen": 49716320,
      "step": 85630
    },
    {
      "epoch": 12.754691689008043,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.010478071982700175,
      "loss": 0.7741,
      "num_input_tokens_seen": 49719424,
      "step": 85635
    },
    {
      "epoch": 12.755436401549002,
      "grad_norm": 0.068359375,
      "learning_rate": 0.01047621306932422,
      "loss": 0.7986,
      "num_input_tokens_seen": 49722240,
      "step": 85640
    },
    {
      "epoch": 12.756181114089962,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.010474354232373043,
      "loss": 0.8105,
      "num_input_tokens_seen": 49724864,
      "step": 85645
    },
    {
      "epoch": 12.756925826630921,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.010472495471878054,
      "loss": 0.798,
      "num_input_tokens_seen": 49727776,
      "step": 85650
    },
    {
      "epoch": 12.75767053917188,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.010470636787870649,
      "loss": 0.7916,
      "num_input_tokens_seen": 49730752,
      "step": 85655
    },
    {
      "epoch": 12.758415251712838,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.010468778180382238,
      "loss": 0.8048,
      "num_input_tokens_seen": 49733504,
      "step": 85660
    },
    {
      "epoch": 12.759159964253797,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.010466919649444214,
      "loss": 0.8019,
      "num_input_tokens_seen": 49736480,
      "step": 85665
    },
    {
      "epoch": 12.759904676794758,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.010465061195087975,
      "loss": 0.8126,
      "num_input_tokens_seen": 49739520,
      "step": 85670
    },
    {
      "epoch": 12.760649389335716,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.010463202817344919,
      "loss": 0.7794,
      "num_input_tokens_seen": 49742272,
      "step": 85675
    },
    {
      "epoch": 12.761394101876675,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.010461344516246434,
      "loss": 0.8048,
      "num_input_tokens_seen": 49745120,
      "step": 85680
    },
    {
      "epoch": 12.762138814417634,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.010459486291823927,
      "loss": 0.8011,
      "num_input_tokens_seen": 49748096,
      "step": 85685
    },
    {
      "epoch": 12.762883526958595,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.010457628144108777,
      "loss": 0.8027,
      "num_input_tokens_seen": 49750880,
      "step": 85690
    },
    {
      "epoch": 12.763628239499553,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.01045577007313239,
      "loss": 0.799,
      "num_input_tokens_seen": 49753824,
      "step": 85695
    },
    {
      "epoch": 12.764372952040512,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.010453912078926146,
      "loss": 0.7787,
      "num_input_tokens_seen": 49756640,
      "step": 85700
    },
    {
      "epoch": 12.76511766458147,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.010452054161521441,
      "loss": 0.7807,
      "num_input_tokens_seen": 49759648,
      "step": 85705
    },
    {
      "epoch": 12.765862377122431,
      "grad_norm": 0.0390625,
      "learning_rate": 0.01045019632094965,
      "loss": 0.7866,
      "num_input_tokens_seen": 49762592,
      "step": 85710
    },
    {
      "epoch": 12.76660708966339,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.010448338557242176,
      "loss": 0.7928,
      "num_input_tokens_seen": 49765600,
      "step": 85715
    },
    {
      "epoch": 12.767351802204349,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.010446480870430397,
      "loss": 0.7769,
      "num_input_tokens_seen": 49768352,
      "step": 85720
    },
    {
      "epoch": 12.768096514745308,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.010444623260545686,
      "loss": 0.8113,
      "num_input_tokens_seen": 49771520,
      "step": 85725
    },
    {
      "epoch": 12.768841227286268,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.010442765727619444,
      "loss": 0.8108,
      "num_input_tokens_seen": 49774336,
      "step": 85730
    },
    {
      "epoch": 12.769585939827227,
      "grad_norm": 0.29296875,
      "learning_rate": 0.01044090827168304,
      "loss": 0.8468,
      "num_input_tokens_seen": 49777152,
      "step": 85735
    },
    {
      "epoch": 12.770330652368186,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.01043905089276786,
      "loss": 0.8309,
      "num_input_tokens_seen": 49779840,
      "step": 85740
    },
    {
      "epoch": 12.771075364909144,
      "grad_norm": 0.0234375,
      "learning_rate": 0.010437193590905272,
      "loss": 0.7882,
      "num_input_tokens_seen": 49782592,
      "step": 85745
    },
    {
      "epoch": 12.771820077450105,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.010435336366126669,
      "loss": 0.8127,
      "num_input_tokens_seen": 49785760,
      "step": 85750
    },
    {
      "epoch": 12.772564789991064,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.010433479218463418,
      "loss": 0.7896,
      "num_input_tokens_seen": 49788640,
      "step": 85755
    },
    {
      "epoch": 12.773309502532022,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.010431622147946887,
      "loss": 0.7874,
      "num_input_tokens_seen": 49791392,
      "step": 85760
    },
    {
      "epoch": 12.774054215072981,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.010429765154608465,
      "loss": 0.7811,
      "num_input_tokens_seen": 49794208,
      "step": 85765
    },
    {
      "epoch": 12.774798927613942,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.010427908238479512,
      "loss": 0.8149,
      "num_input_tokens_seen": 49797056,
      "step": 85770
    },
    {
      "epoch": 12.7755436401549,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.010426051399591404,
      "loss": 0.8039,
      "num_input_tokens_seen": 49799776,
      "step": 85775
    },
    {
      "epoch": 12.77628835269586,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.010424194637975503,
      "loss": 0.7876,
      "num_input_tokens_seen": 49802688,
      "step": 85780
    },
    {
      "epoch": 12.777033065236818,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.010422337953663192,
      "loss": 0.7719,
      "num_input_tokens_seen": 49805696,
      "step": 85785
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.010420481346685821,
      "loss": 0.8054,
      "num_input_tokens_seen": 49808704,
      "step": 85790
    },
    {
      "epoch": 12.778522490318737,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.010418624817074767,
      "loss": 0.7835,
      "num_input_tokens_seen": 49811712,
      "step": 85795
    },
    {
      "epoch": 12.779267202859696,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.010416768364861394,
      "loss": 0.7846,
      "num_input_tokens_seen": 49814688,
      "step": 85800
    },
    {
      "epoch": 12.780011915400655,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.010414911990077056,
      "loss": 0.8078,
      "num_input_tokens_seen": 49817472,
      "step": 85805
    },
    {
      "epoch": 12.780756627941614,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.010413055692753125,
      "loss": 0.7775,
      "num_input_tokens_seen": 49820384,
      "step": 85810
    },
    {
      "epoch": 12.781501340482574,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.010411199472920947,
      "loss": 0.8085,
      "num_input_tokens_seen": 49823456,
      "step": 85815
    },
    {
      "epoch": 12.782246053023533,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.010409343330611897,
      "loss": 0.8119,
      "num_input_tokens_seen": 49826272,
      "step": 85820
    },
    {
      "epoch": 12.782990765564492,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.010407487265857322,
      "loss": 0.8148,
      "num_input_tokens_seen": 49829216,
      "step": 85825
    },
    {
      "epoch": 12.783735478105452,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.010405631278688586,
      "loss": 0.7931,
      "num_input_tokens_seen": 49832128,
      "step": 85830
    },
    {
      "epoch": 12.784480190646411,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.010403775369137037,
      "loss": 0.7918,
      "num_input_tokens_seen": 49835136,
      "step": 85835
    },
    {
      "epoch": 12.78522490318737,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.010401919537234036,
      "loss": 0.7666,
      "num_input_tokens_seen": 49837984,
      "step": 85840
    },
    {
      "epoch": 12.785969615728328,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.01040006378301093,
      "loss": 0.7752,
      "num_input_tokens_seen": 49840896,
      "step": 85845
    },
    {
      "epoch": 12.786714328269287,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.010398208106499066,
      "loss": 0.8056,
      "num_input_tokens_seen": 49843872,
      "step": 85850
    },
    {
      "epoch": 12.787459040810248,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.010396352507729804,
      "loss": 0.7826,
      "num_input_tokens_seen": 49846944,
      "step": 85855
    },
    {
      "epoch": 12.788203753351207,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.010394496986734484,
      "loss": 0.7803,
      "num_input_tokens_seen": 49849760,
      "step": 85860
    },
    {
      "epoch": 12.788948465892165,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.010392641543544459,
      "loss": 0.8135,
      "num_input_tokens_seen": 49852800,
      "step": 85865
    },
    {
      "epoch": 12.789693178433124,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.010390786178191069,
      "loss": 0.7876,
      "num_input_tokens_seen": 49855712,
      "step": 85870
    },
    {
      "epoch": 12.790437890974085,
      "grad_norm": 0.0194091796875,
      "learning_rate": 0.010388930890705667,
      "loss": 0.8053,
      "num_input_tokens_seen": 49858560,
      "step": 85875
    },
    {
      "epoch": 12.791182603515043,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.010387075681119592,
      "loss": 0.8064,
      "num_input_tokens_seen": 49861344,
      "step": 85880
    },
    {
      "epoch": 12.791927316056002,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.010385220549464178,
      "loss": 0.7979,
      "num_input_tokens_seen": 49864032,
      "step": 85885
    },
    {
      "epoch": 12.79267202859696,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.010383365495770779,
      "loss": 0.8042,
      "num_input_tokens_seen": 49866912,
      "step": 85890
    },
    {
      "epoch": 12.793416741137921,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.010381510520070723,
      "loss": 0.7736,
      "num_input_tokens_seen": 49869504,
      "step": 85895
    },
    {
      "epoch": 12.79416145367888,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.010379655622395359,
      "loss": 0.8059,
      "num_input_tokens_seen": 49872160,
      "step": 85900
    },
    {
      "epoch": 12.794906166219839,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.010377800802776008,
      "loss": 0.806,
      "num_input_tokens_seen": 49875264,
      "step": 85905
    },
    {
      "epoch": 12.795650878760798,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.010375946061244022,
      "loss": 0.7824,
      "num_input_tokens_seen": 49877920,
      "step": 85910
    },
    {
      "epoch": 12.796395591301758,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.010374091397830723,
      "loss": 0.8151,
      "num_input_tokens_seen": 49880672,
      "step": 85915
    },
    {
      "epoch": 12.797140303842717,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.010372236812567453,
      "loss": 0.8094,
      "num_input_tokens_seen": 49883424,
      "step": 85920
    },
    {
      "epoch": 12.797885016383676,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.010370382305485539,
      "loss": 0.7962,
      "num_input_tokens_seen": 49886368,
      "step": 85925
    },
    {
      "epoch": 12.798629728924634,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.010368527876616306,
      "loss": 0.7996,
      "num_input_tokens_seen": 49889280,
      "step": 85930
    },
    {
      "epoch": 12.799374441465595,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.01036667352599109,
      "loss": 0.7918,
      "num_input_tokens_seen": 49892352,
      "step": 85935
    },
    {
      "epoch": 12.800119154006554,
      "grad_norm": 0.033203125,
      "learning_rate": 0.010364819253641213,
      "loss": 0.8002,
      "num_input_tokens_seen": 49895040,
      "step": 85940
    },
    {
      "epoch": 12.800863866547513,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.010362965059598009,
      "loss": 0.7745,
      "num_input_tokens_seen": 49897600,
      "step": 85945
    },
    {
      "epoch": 12.801608579088471,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.01036111094389279,
      "loss": 0.8078,
      "num_input_tokens_seen": 49900832,
      "step": 85950
    },
    {
      "epoch": 12.802353291629432,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.010359256906556893,
      "loss": 0.7848,
      "num_input_tokens_seen": 49903840,
      "step": 85955
    },
    {
      "epoch": 12.80309800417039,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.010357402947621634,
      "loss": 0.8055,
      "num_input_tokens_seen": 49906848,
      "step": 85960
    },
    {
      "epoch": 12.80384271671135,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.01035554906711833,
      "loss": 0.7895,
      "num_input_tokens_seen": 49909536,
      "step": 85965
    },
    {
      "epoch": 12.804587429252308,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.010353695265078308,
      "loss": 0.7836,
      "num_input_tokens_seen": 49912384,
      "step": 85970
    },
    {
      "epoch": 12.805332141793269,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.010351841541532878,
      "loss": 0.7885,
      "num_input_tokens_seen": 49915616,
      "step": 85975
    },
    {
      "epoch": 12.806076854334227,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.010349987896513363,
      "loss": 0.7888,
      "num_input_tokens_seen": 49918624,
      "step": 85980
    },
    {
      "epoch": 12.806821566875186,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.010348134330051073,
      "loss": 0.7817,
      "num_input_tokens_seen": 49921984,
      "step": 85985
    },
    {
      "epoch": 12.807566279416145,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.01034628084217733,
      "loss": 0.8411,
      "num_input_tokens_seen": 49924576,
      "step": 85990
    },
    {
      "epoch": 12.808310991957104,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.010344427432923442,
      "loss": 0.8057,
      "num_input_tokens_seen": 49927296,
      "step": 85995
    },
    {
      "epoch": 12.809055704498064,
      "grad_norm": 0.060546875,
      "learning_rate": 0.010342574102320723,
      "loss": 0.8356,
      "num_input_tokens_seen": 49930304,
      "step": 86000
    },
    {
      "epoch": 12.809800417039023,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.010340720850400479,
      "loss": 0.7937,
      "num_input_tokens_seen": 49933376,
      "step": 86005
    },
    {
      "epoch": 12.810545129579982,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.010338867677194016,
      "loss": 0.7876,
      "num_input_tokens_seen": 49936000,
      "step": 86010
    },
    {
      "epoch": 12.811289842120942,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.010337014582732652,
      "loss": 0.7889,
      "num_input_tokens_seen": 49938976,
      "step": 86015
    },
    {
      "epoch": 12.812034554661901,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.01033516156704768,
      "loss": 0.787,
      "num_input_tokens_seen": 49941888,
      "step": 86020
    },
    {
      "epoch": 12.81277926720286,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.01033330863017042,
      "loss": 0.7909,
      "num_input_tokens_seen": 49944544,
      "step": 86025
    },
    {
      "epoch": 12.813523979743819,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.010331455772132165,
      "loss": 0.7961,
      "num_input_tokens_seen": 49947200,
      "step": 86030
    },
    {
      "epoch": 12.814268692284777,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.010329602992964223,
      "loss": 0.8113,
      "num_input_tokens_seen": 49950112,
      "step": 86035
    },
    {
      "epoch": 12.815013404825738,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.010327750292697892,
      "loss": 0.7942,
      "num_input_tokens_seen": 49952832,
      "step": 86040
    },
    {
      "epoch": 12.815758117366697,
      "grad_norm": 0.044921875,
      "learning_rate": 0.010325897671364463,
      "loss": 0.8003,
      "num_input_tokens_seen": 49955712,
      "step": 86045
    },
    {
      "epoch": 12.816502829907655,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.010324045128995249,
      "loss": 0.8259,
      "num_input_tokens_seen": 49958720,
      "step": 86050
    },
    {
      "epoch": 12.817247542448614,
      "grad_norm": 0.033203125,
      "learning_rate": 0.010322192665621534,
      "loss": 0.8147,
      "num_input_tokens_seen": 49961760,
      "step": 86055
    },
    {
      "epoch": 12.817992254989575,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.010320340281274626,
      "loss": 0.7908,
      "num_input_tokens_seen": 49964576,
      "step": 86060
    },
    {
      "epoch": 12.818736967530533,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.01031848797598581,
      "loss": 0.8338,
      "num_input_tokens_seen": 49967584,
      "step": 86065
    },
    {
      "epoch": 12.819481680071492,
      "grad_norm": 0.04296875,
      "learning_rate": 0.010316635749786385,
      "loss": 0.8054,
      "num_input_tokens_seen": 49970624,
      "step": 86070
    },
    {
      "epoch": 12.820226392612451,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.010314783602707633,
      "loss": 0.794,
      "num_input_tokens_seen": 49973312,
      "step": 86075
    },
    {
      "epoch": 12.820971105153411,
      "grad_norm": 0.041015625,
      "learning_rate": 0.010312931534780856,
      "loss": 0.8129,
      "num_input_tokens_seen": 49976448,
      "step": 86080
    },
    {
      "epoch": 12.82171581769437,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.010311079546037338,
      "loss": 0.7738,
      "num_input_tokens_seen": 49979488,
      "step": 86085
    },
    {
      "epoch": 12.822460530235329,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.010309227636508357,
      "loss": 0.8135,
      "num_input_tokens_seen": 49982240,
      "step": 86090
    },
    {
      "epoch": 12.823205242776288,
      "grad_norm": 0.03125,
      "learning_rate": 0.010307375806225212,
      "loss": 0.8038,
      "num_input_tokens_seen": 49985088,
      "step": 86095
    },
    {
      "epoch": 12.823949955317248,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.010305524055219183,
      "loss": 0.8103,
      "num_input_tokens_seen": 49988192,
      "step": 86100
    },
    {
      "epoch": 12.824694667858207,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.010303672383521558,
      "loss": 0.7982,
      "num_input_tokens_seen": 49991392,
      "step": 86105
    },
    {
      "epoch": 12.825439380399166,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.010301820791163607,
      "loss": 0.8197,
      "num_input_tokens_seen": 49994496,
      "step": 86110
    },
    {
      "epoch": 12.826184092940125,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.010299969278176625,
      "loss": 0.7893,
      "num_input_tokens_seen": 49997408,
      "step": 86115
    },
    {
      "epoch": 12.826928805481085,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.010298117844591888,
      "loss": 0.7873,
      "num_input_tokens_seen": 50000288,
      "step": 86120
    },
    {
      "epoch": 12.827673518022044,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.010296266490440662,
      "loss": 0.8211,
      "num_input_tokens_seen": 50003264,
      "step": 86125
    },
    {
      "epoch": 12.828418230563003,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.01029441521575424,
      "loss": 0.7869,
      "num_input_tokens_seen": 50005920,
      "step": 86130
    },
    {
      "epoch": 12.829162943103961,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.010292564020563888,
      "loss": 0.7794,
      "num_input_tokens_seen": 50008800,
      "step": 86135
    },
    {
      "epoch": 12.829907655644922,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.010290712904900884,
      "loss": 0.7987,
      "num_input_tokens_seen": 50012352,
      "step": 86140
    },
    {
      "epoch": 12.83065236818588,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.010288861868796493,
      "loss": 0.803,
      "num_input_tokens_seen": 50015424,
      "step": 86145
    },
    {
      "epoch": 12.83139708072684,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.010287010912282001,
      "loss": 0.7941,
      "num_input_tokens_seen": 50018176,
      "step": 86150
    },
    {
      "epoch": 12.832141793267798,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.01028516003538866,
      "loss": 0.7931,
      "num_input_tokens_seen": 50021024,
      "step": 86155
    },
    {
      "epoch": 12.832886505808759,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.010283309238147758,
      "loss": 0.7784,
      "num_input_tokens_seen": 50023968,
      "step": 86160
    },
    {
      "epoch": 12.833631218349717,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.01028145852059055,
      "loss": 0.8209,
      "num_input_tokens_seen": 50027104,
      "step": 86165
    },
    {
      "epoch": 12.834375930890676,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.010279607882748303,
      "loss": 0.7865,
      "num_input_tokens_seen": 50030016,
      "step": 86170
    },
    {
      "epoch": 12.835120643431635,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.010277757324652286,
      "loss": 0.7966,
      "num_input_tokens_seen": 50032800,
      "step": 86175
    },
    {
      "epoch": 12.835865355972594,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.010275906846333753,
      "loss": 0.7894,
      "num_input_tokens_seen": 50035872,
      "step": 86180
    },
    {
      "epoch": 12.836610068513554,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01027405644782398,
      "loss": 0.8183,
      "num_input_tokens_seen": 50038592,
      "step": 86185
    },
    {
      "epoch": 12.837354781054513,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.010272206129154214,
      "loss": 0.8156,
      "num_input_tokens_seen": 50041472,
      "step": 86190
    },
    {
      "epoch": 12.838099493595472,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.010270355890355723,
      "loss": 0.8261,
      "num_input_tokens_seen": 50044512,
      "step": 86195
    },
    {
      "epoch": 12.83884420613643,
      "grad_norm": 0.048828125,
      "learning_rate": 0.010268505731459765,
      "loss": 0.7974,
      "num_input_tokens_seen": 50047136,
      "step": 86200
    },
    {
      "epoch": 12.839588918677391,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.010266655652497588,
      "loss": 0.7772,
      "num_input_tokens_seen": 50050272,
      "step": 86205
    },
    {
      "epoch": 12.84033363121835,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.01026480565350046,
      "loss": 0.8072,
      "num_input_tokens_seen": 50053088,
      "step": 86210
    },
    {
      "epoch": 12.841078343759309,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.01026295573449962,
      "loss": 0.7881,
      "num_input_tokens_seen": 50056000,
      "step": 86215
    },
    {
      "epoch": 12.841823056300267,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.010261105895526333,
      "loss": 0.8231,
      "num_input_tokens_seen": 50058880,
      "step": 86220
    },
    {
      "epoch": 12.842567768841228,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.010259256136611846,
      "loss": 0.7778,
      "num_input_tokens_seen": 50061920,
      "step": 86225
    },
    {
      "epoch": 12.843312481382187,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.010257406457787409,
      "loss": 0.7983,
      "num_input_tokens_seen": 50065120,
      "step": 86230
    },
    {
      "epoch": 12.844057193923145,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.010255556859084263,
      "loss": 0.7772,
      "num_input_tokens_seen": 50067936,
      "step": 86235
    },
    {
      "epoch": 12.844801906464104,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.010253707340533668,
      "loss": 0.798,
      "num_input_tokens_seen": 50070784,
      "step": 86240
    },
    {
      "epoch": 12.845546619005065,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.010251857902166865,
      "loss": 0.766,
      "num_input_tokens_seen": 50073504,
      "step": 86245
    },
    {
      "epoch": 12.846291331546023,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.010250008544015088,
      "loss": 0.8052,
      "num_input_tokens_seen": 50076608,
      "step": 86250
    },
    {
      "epoch": 12.847036044086982,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0102481592661096,
      "loss": 0.7888,
      "num_input_tokens_seen": 50079648,
      "step": 86255
    },
    {
      "epoch": 12.847780756627941,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.010246310068481625,
      "loss": 0.8013,
      "num_input_tokens_seen": 50082336,
      "step": 86260
    },
    {
      "epoch": 12.848525469168901,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.010244460951162413,
      "loss": 0.7847,
      "num_input_tokens_seen": 50085248,
      "step": 86265
    },
    {
      "epoch": 12.84927018170986,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.010242611914183195,
      "loss": 0.7802,
      "num_input_tokens_seen": 50087968,
      "step": 86270
    },
    {
      "epoch": 12.850014894250819,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.01024076295757522,
      "loss": 0.7902,
      "num_input_tokens_seen": 50090656,
      "step": 86275
    },
    {
      "epoch": 12.850759606791778,
      "grad_norm": 0.033203125,
      "learning_rate": 0.010238914081369717,
      "loss": 0.8041,
      "num_input_tokens_seen": 50093728,
      "step": 86280
    },
    {
      "epoch": 12.851504319332738,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.010237065285597915,
      "loss": 0.7992,
      "num_input_tokens_seen": 50096384,
      "step": 86285
    },
    {
      "epoch": 12.852249031873697,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.010235216570291061,
      "loss": 0.805,
      "num_input_tokens_seen": 50099040,
      "step": 86290
    },
    {
      "epoch": 12.852993744414656,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.010233367935480377,
      "loss": 0.7604,
      "num_input_tokens_seen": 50102208,
      "step": 86295
    },
    {
      "epoch": 12.853738456955615,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.0102315193811971,
      "loss": 0.8026,
      "num_input_tokens_seen": 50104960,
      "step": 86300
    },
    {
      "epoch": 12.854483169496575,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.01022967090747245,
      "loss": 0.8049,
      "num_input_tokens_seen": 50108160,
      "step": 86305
    },
    {
      "epoch": 12.855227882037534,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.010227822514337669,
      "loss": 0.784,
      "num_input_tokens_seen": 50111072,
      "step": 86310
    },
    {
      "epoch": 12.855972594578493,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.010225974201823971,
      "loss": 0.7824,
      "num_input_tokens_seen": 50113728,
      "step": 86315
    },
    {
      "epoch": 12.856717307119451,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.01022412596996259,
      "loss": 0.8089,
      "num_input_tokens_seen": 50116544,
      "step": 86320
    },
    {
      "epoch": 12.857462019660412,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.01022227781878475,
      "loss": 0.7749,
      "num_input_tokens_seen": 50119328,
      "step": 86325
    },
    {
      "epoch": 12.85820673220137,
      "grad_norm": 0.044921875,
      "learning_rate": 0.010220429748321665,
      "loss": 0.7975,
      "num_input_tokens_seen": 50122016,
      "step": 86330
    },
    {
      "epoch": 12.85895144474233,
      "grad_norm": 0.044921875,
      "learning_rate": 0.010218581758604567,
      "loss": 0.7983,
      "num_input_tokens_seen": 50124800,
      "step": 86335
    },
    {
      "epoch": 12.859696157283288,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.010216733849664662,
      "loss": 0.7831,
      "num_input_tokens_seen": 50127680,
      "step": 86340
    },
    {
      "epoch": 12.860440869824249,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.010214886021533184,
      "loss": 0.779,
      "num_input_tokens_seen": 50131040,
      "step": 86345
    },
    {
      "epoch": 12.861185582365207,
      "grad_norm": 0.04296875,
      "learning_rate": 0.010213038274241337,
      "loss": 0.8371,
      "num_input_tokens_seen": 50134112,
      "step": 86350
    },
    {
      "epoch": 12.861930294906166,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01021119060782035,
      "loss": 0.7915,
      "num_input_tokens_seen": 50137216,
      "step": 86355
    },
    {
      "epoch": 12.862675007447125,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.01020934302230143,
      "loss": 0.7732,
      "num_input_tokens_seen": 50139872,
      "step": 86360
    },
    {
      "epoch": 12.863419719988084,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.010207495517715787,
      "loss": 0.8003,
      "num_input_tokens_seen": 50142592,
      "step": 86365
    },
    {
      "epoch": 12.864164432529044,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.010205648094094643,
      "loss": 0.7842,
      "num_input_tokens_seen": 50145504,
      "step": 86370
    },
    {
      "epoch": 12.864909145070003,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.01020380075146919,
      "loss": 0.8078,
      "num_input_tokens_seen": 50148384,
      "step": 86375
    },
    {
      "epoch": 12.865653857610962,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.010201953489870658,
      "loss": 0.7977,
      "num_input_tokens_seen": 50151072,
      "step": 86380
    },
    {
      "epoch": 12.86639857015192,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.010200106309330237,
      "loss": 0.7758,
      "num_input_tokens_seen": 50154304,
      "step": 86385
    },
    {
      "epoch": 12.867143282692881,
      "grad_norm": 0.04296875,
      "learning_rate": 0.01019825920987915,
      "loss": 0.8082,
      "num_input_tokens_seen": 50157376,
      "step": 86390
    },
    {
      "epoch": 12.86788799523384,
      "grad_norm": 0.05078125,
      "learning_rate": 0.010196412191548591,
      "loss": 0.7716,
      "num_input_tokens_seen": 50160128,
      "step": 86395
    },
    {
      "epoch": 12.868632707774799,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.010194565254369767,
      "loss": 0.8365,
      "num_input_tokens_seen": 50162944,
      "step": 86400
    },
    {
      "epoch": 12.869377420315757,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.01019271839837388,
      "loss": 0.8183,
      "num_input_tokens_seen": 50165824,
      "step": 86405
    },
    {
      "epoch": 12.870122132856718,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.010190871623592123,
      "loss": 0.8114,
      "num_input_tokens_seen": 50168640,
      "step": 86410
    },
    {
      "epoch": 12.870866845397677,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.010189024930055707,
      "loss": 0.8259,
      "num_input_tokens_seen": 50171392,
      "step": 86415
    },
    {
      "epoch": 12.871611557938635,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.010187178317795823,
      "loss": 0.8108,
      "num_input_tokens_seen": 50174400,
      "step": 86420
    },
    {
      "epoch": 12.872356270479594,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.010185331786843672,
      "loss": 0.7816,
      "num_input_tokens_seen": 50177056,
      "step": 86425
    },
    {
      "epoch": 12.873100983020555,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.010183485337230446,
      "loss": 0.7961,
      "num_input_tokens_seen": 50179872,
      "step": 86430
    },
    {
      "epoch": 12.873845695561513,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.010181638968987342,
      "loss": 0.8054,
      "num_input_tokens_seen": 50182464,
      "step": 86435
    },
    {
      "epoch": 12.874590408102472,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.010179792682145543,
      "loss": 0.7869,
      "num_input_tokens_seen": 50185280,
      "step": 86440
    },
    {
      "epoch": 12.875335120643431,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.010177946476736257,
      "loss": 0.7993,
      "num_input_tokens_seen": 50188224,
      "step": 86445
    },
    {
      "epoch": 12.876079833184392,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.010176100352790662,
      "loss": 0.7755,
      "num_input_tokens_seen": 50191104,
      "step": 86450
    },
    {
      "epoch": 12.87682454572535,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.010174254310339945,
      "loss": 0.7894,
      "num_input_tokens_seen": 50193856,
      "step": 86455
    },
    {
      "epoch": 12.877569258266309,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.0101724083494153,
      "loss": 0.804,
      "num_input_tokens_seen": 50196704,
      "step": 86460
    },
    {
      "epoch": 12.878313970807268,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.010170562470047908,
      "loss": 0.7981,
      "num_input_tokens_seen": 50199584,
      "step": 86465
    },
    {
      "epoch": 12.879058683348228,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.010168716672268957,
      "loss": 0.7835,
      "num_input_tokens_seen": 50202464,
      "step": 86470
    },
    {
      "epoch": 12.879803395889187,
      "grad_norm": 0.041015625,
      "learning_rate": 0.010166870956109622,
      "loss": 0.785,
      "num_input_tokens_seen": 50205184,
      "step": 86475
    },
    {
      "epoch": 12.880548108430146,
      "grad_norm": 0.04296875,
      "learning_rate": 0.010165025321601096,
      "loss": 0.7975,
      "num_input_tokens_seen": 50208000,
      "step": 86480
    },
    {
      "epoch": 12.881292820971105,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.010163179768774552,
      "loss": 0.7976,
      "num_input_tokens_seen": 50210752,
      "step": 86485
    },
    {
      "epoch": 12.882037533512065,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.010161334297661166,
      "loss": 0.7859,
      "num_input_tokens_seen": 50213792,
      "step": 86490
    },
    {
      "epoch": 12.882782246053024,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.010159488908292125,
      "loss": 0.8001,
      "num_input_tokens_seen": 50216640,
      "step": 86495
    },
    {
      "epoch": 12.883526958593983,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.010157643600698594,
      "loss": 0.8537,
      "num_input_tokens_seen": 50219520,
      "step": 86500
    },
    {
      "epoch": 12.884271671134941,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.010155798374911758,
      "loss": 0.8173,
      "num_input_tokens_seen": 50222272,
      "step": 86505
    },
    {
      "epoch": 12.8850163836759,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.010153953230962777,
      "loss": 0.7853,
      "num_input_tokens_seen": 50225152,
      "step": 86510
    },
    {
      "epoch": 12.88576109621686,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.010152108168882838,
      "loss": 0.7976,
      "num_input_tokens_seen": 50228160,
      "step": 86515
    },
    {
      "epoch": 12.88650580875782,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.010150263188703101,
      "loss": 0.7875,
      "num_input_tokens_seen": 50231168,
      "step": 86520
    },
    {
      "epoch": 12.887250521298778,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.01014841829045474,
      "loss": 0.8005,
      "num_input_tokens_seen": 50233760,
      "step": 86525
    },
    {
      "epoch": 12.887995233839739,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.010146573474168922,
      "loss": 0.7868,
      "num_input_tokens_seen": 50236544,
      "step": 86530
    },
    {
      "epoch": 12.888739946380698,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.010144728739876811,
      "loss": 0.8006,
      "num_input_tokens_seen": 50239328,
      "step": 86535
    },
    {
      "epoch": 12.889484658921656,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.010142884087609575,
      "loss": 0.7905,
      "num_input_tokens_seen": 50242208,
      "step": 86540
    },
    {
      "epoch": 12.890229371462615,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.010141039517398368,
      "loss": 0.8009,
      "num_input_tokens_seen": 50245248,
      "step": 86545
    },
    {
      "epoch": 12.890974084003574,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.01013919502927437,
      "loss": 0.8002,
      "num_input_tokens_seen": 50248480,
      "step": 86550
    },
    {
      "epoch": 12.891718796544534,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.010137350623268726,
      "loss": 0.7795,
      "num_input_tokens_seen": 50251072,
      "step": 86555
    },
    {
      "epoch": 12.892463509085493,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.010135506299412607,
      "loss": 0.7962,
      "num_input_tokens_seen": 50254208,
      "step": 86560
    },
    {
      "epoch": 12.893208221626452,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.010133662057737159,
      "loss": 0.7999,
      "num_input_tokens_seen": 50257280,
      "step": 86565
    },
    {
      "epoch": 12.89395293416741,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.010131817898273545,
      "loss": 0.8052,
      "num_input_tokens_seen": 50260064,
      "step": 86570
    },
    {
      "epoch": 12.894697646708371,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.010129973821052922,
      "loss": 0.8034,
      "num_input_tokens_seen": 50262816,
      "step": 86575
    },
    {
      "epoch": 12.89544235924933,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.010128129826106438,
      "loss": 0.7831,
      "num_input_tokens_seen": 50265984,
      "step": 86580
    },
    {
      "epoch": 12.896187071790289,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.010126285913465254,
      "loss": 0.8022,
      "num_input_tokens_seen": 50268832,
      "step": 86585
    },
    {
      "epoch": 12.896931784331247,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.010124442083160512,
      "loss": 0.7785,
      "num_input_tokens_seen": 50271648,
      "step": 86590
    },
    {
      "epoch": 12.897676496872208,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.010122598335223372,
      "loss": 0.7873,
      "num_input_tokens_seen": 50274560,
      "step": 86595
    },
    {
      "epoch": 12.898421209413167,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.010120754669684965,
      "loss": 0.8028,
      "num_input_tokens_seen": 50277408,
      "step": 86600
    },
    {
      "epoch": 12.899165921954125,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.010118911086576457,
      "loss": 0.8004,
      "num_input_tokens_seen": 50280320,
      "step": 86605
    },
    {
      "epoch": 12.899910634495084,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.010117067585928988,
      "loss": 0.7924,
      "num_input_tokens_seen": 50282816,
      "step": 86610
    },
    {
      "epoch": 12.900655347036045,
      "grad_norm": 0.03125,
      "learning_rate": 0.01011522416777369,
      "loss": 0.8039,
      "num_input_tokens_seen": 50285504,
      "step": 86615
    },
    {
      "epoch": 12.901400059577004,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.010113380832141722,
      "loss": 0.7877,
      "num_input_tokens_seen": 50288096,
      "step": 86620
    },
    {
      "epoch": 12.902144772117962,
      "grad_norm": 0.0390625,
      "learning_rate": 0.010111537579064215,
      "loss": 0.816,
      "num_input_tokens_seen": 50290912,
      "step": 86625
    },
    {
      "epoch": 12.902889484658921,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.010109694408572316,
      "loss": 0.7988,
      "num_input_tokens_seen": 50293696,
      "step": 86630
    },
    {
      "epoch": 12.903634197199882,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.010107851320697152,
      "loss": 0.8234,
      "num_input_tokens_seen": 50296576,
      "step": 86635
    },
    {
      "epoch": 12.90437890974084,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.010106008315469876,
      "loss": 0.7911,
      "num_input_tokens_seen": 50299424,
      "step": 86640
    },
    {
      "epoch": 12.905123622281799,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.010104165392921614,
      "loss": 0.8107,
      "num_input_tokens_seen": 50302208,
      "step": 86645
    },
    {
      "epoch": 12.905868334822758,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.010102322553083498,
      "loss": 0.7866,
      "num_input_tokens_seen": 50305248,
      "step": 86650
    },
    {
      "epoch": 12.906613047363718,
      "grad_norm": 0.046875,
      "learning_rate": 0.010100479795986668,
      "loss": 0.7917,
      "num_input_tokens_seen": 50308096,
      "step": 86655
    },
    {
      "epoch": 12.907357759904677,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.010098637121662252,
      "loss": 0.793,
      "num_input_tokens_seen": 50310816,
      "step": 86660
    },
    {
      "epoch": 12.908102472445636,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.01009679453014138,
      "loss": 0.7928,
      "num_input_tokens_seen": 50313568,
      "step": 86665
    },
    {
      "epoch": 12.908847184986595,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.010094952021455177,
      "loss": 0.8105,
      "num_input_tokens_seen": 50316736,
      "step": 86670
    },
    {
      "epoch": 12.909591897527555,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.010093109595634779,
      "loss": 0.7879,
      "num_input_tokens_seen": 50319744,
      "step": 86675
    },
    {
      "epoch": 12.910336610068514,
      "grad_norm": 0.12890625,
      "learning_rate": 0.010091267252711303,
      "loss": 0.8059,
      "num_input_tokens_seen": 50322368,
      "step": 86680
    },
    {
      "epoch": 12.911081322609473,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.010089424992715881,
      "loss": 0.7735,
      "num_input_tokens_seen": 50325280,
      "step": 86685
    },
    {
      "epoch": 12.911826035150431,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.010087582815679634,
      "loss": 0.8078,
      "num_input_tokens_seen": 50328224,
      "step": 86690
    },
    {
      "epoch": 12.91257074769139,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.010085740721633681,
      "loss": 0.7833,
      "num_input_tokens_seen": 50331232,
      "step": 86695
    },
    {
      "epoch": 12.91331546023235,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.010083898710609147,
      "loss": 0.7872,
      "num_input_tokens_seen": 50334144,
      "step": 86700
    },
    {
      "epoch": 12.91406017277331,
      "grad_norm": 0.052734375,
      "learning_rate": 0.01008205678263714,
      "loss": 0.7939,
      "num_input_tokens_seen": 50336896,
      "step": 86705
    },
    {
      "epoch": 12.914804885314268,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.010080214937748794,
      "loss": 0.8216,
      "num_input_tokens_seen": 50339968,
      "step": 86710
    },
    {
      "epoch": 12.915549597855229,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.010078373175975206,
      "loss": 0.7853,
      "num_input_tokens_seen": 50342912,
      "step": 86715
    },
    {
      "epoch": 12.916294310396188,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.010076531497347512,
      "loss": 0.7927,
      "num_input_tokens_seen": 50345728,
      "step": 86720
    },
    {
      "epoch": 12.917039022937146,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.010074689901896812,
      "loss": 0.7765,
      "num_input_tokens_seen": 50348928,
      "step": 86725
    },
    {
      "epoch": 12.917783735478105,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.010072848389654217,
      "loss": 0.8018,
      "num_input_tokens_seen": 50351744,
      "step": 86730
    },
    {
      "epoch": 12.918528448019064,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.010071006960650844,
      "loss": 0.7738,
      "num_input_tokens_seen": 50354528,
      "step": 86735
    },
    {
      "epoch": 12.919273160560024,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.010069165614917793,
      "loss": 0.7976,
      "num_input_tokens_seen": 50357472,
      "step": 86740
    },
    {
      "epoch": 12.920017873100983,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.010067324352486183,
      "loss": 0.7931,
      "num_input_tokens_seen": 50360512,
      "step": 86745
    },
    {
      "epoch": 12.920762585641942,
      "grad_norm": 0.056640625,
      "learning_rate": 0.010065483173387109,
      "loss": 0.7786,
      "num_input_tokens_seen": 50363008,
      "step": 86750
    },
    {
      "epoch": 12.9215072981829,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.010063642077651687,
      "loss": 0.7793,
      "num_input_tokens_seen": 50365728,
      "step": 86755
    },
    {
      "epoch": 12.922252010723861,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.010061801065311012,
      "loss": 0.8055,
      "num_input_tokens_seen": 50368640,
      "step": 86760
    },
    {
      "epoch": 12.92299672326482,
      "grad_norm": 0.37890625,
      "learning_rate": 0.010059960136396193,
      "loss": 0.7955,
      "num_input_tokens_seen": 50371904,
      "step": 86765
    },
    {
      "epoch": 12.923741435805779,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.010058119290938326,
      "loss": 0.7667,
      "num_input_tokens_seen": 50374784,
      "step": 86770
    },
    {
      "epoch": 12.924486148346737,
      "grad_norm": 0.037109375,
      "learning_rate": 0.010056278528968506,
      "loss": 0.8254,
      "num_input_tokens_seen": 50377600,
      "step": 86775
    },
    {
      "epoch": 12.925230860887698,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.010054437850517841,
      "loss": 0.8143,
      "num_input_tokens_seen": 50380736,
      "step": 86780
    },
    {
      "epoch": 12.925975573428657,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.010052597255617414,
      "loss": 0.775,
      "num_input_tokens_seen": 50383968,
      "step": 86785
    },
    {
      "epoch": 12.926720285969616,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.010050756744298337,
      "loss": 0.7936,
      "num_input_tokens_seen": 50386720,
      "step": 86790
    },
    {
      "epoch": 12.927464998510574,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.01004891631659169,
      "loss": 0.8066,
      "num_input_tokens_seen": 50389760,
      "step": 86795
    },
    {
      "epoch": 12.928209711051535,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.010047075972528574,
      "loss": 0.7926,
      "num_input_tokens_seen": 50392416,
      "step": 86800
    },
    {
      "epoch": 12.928954423592494,
      "grad_norm": 0.03125,
      "learning_rate": 0.010045235712140076,
      "loss": 0.7969,
      "num_input_tokens_seen": 50395104,
      "step": 86805
    },
    {
      "epoch": 12.929699136133452,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.010043395535457276,
      "loss": 0.8056,
      "num_input_tokens_seen": 50397888,
      "step": 86810
    },
    {
      "epoch": 12.930443848674411,
      "grad_norm": 0.064453125,
      "learning_rate": 0.010041555442511278,
      "loss": 0.8156,
      "num_input_tokens_seen": 50400608,
      "step": 86815
    },
    {
      "epoch": 12.931188561215372,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.010039715433333155,
      "loss": 0.8077,
      "num_input_tokens_seen": 50403680,
      "step": 86820
    },
    {
      "epoch": 12.93193327375633,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.010037875507954009,
      "loss": 0.7919,
      "num_input_tokens_seen": 50406400,
      "step": 86825
    },
    {
      "epoch": 12.93267798629729,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.010036035666404902,
      "loss": 0.7872,
      "num_input_tokens_seen": 50409472,
      "step": 86830
    },
    {
      "epoch": 12.933422698838248,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.010034195908716936,
      "loss": 0.7805,
      "num_input_tokens_seen": 50412320,
      "step": 86835
    },
    {
      "epoch": 12.934167411379208,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.010032356234921175,
      "loss": 0.7968,
      "num_input_tokens_seen": 50415072,
      "step": 86840
    },
    {
      "epoch": 12.934912123920167,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.010030516645048714,
      "loss": 0.8081,
      "num_input_tokens_seen": 50417888,
      "step": 86845
    },
    {
      "epoch": 12.935656836461126,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.010028677139130622,
      "loss": 0.7863,
      "num_input_tokens_seen": 50420832,
      "step": 86850
    },
    {
      "epoch": 12.936401549002085,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.010026837717197971,
      "loss": 0.808,
      "num_input_tokens_seen": 50424064,
      "step": 86855
    },
    {
      "epoch": 12.937146261543045,
      "grad_norm": 0.037109375,
      "learning_rate": 0.01002499837928185,
      "loss": 0.7742,
      "num_input_tokens_seen": 50426944,
      "step": 86860
    },
    {
      "epoch": 12.937890974084004,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.01002315912541332,
      "loss": 0.7781,
      "num_input_tokens_seen": 50429696,
      "step": 86865
    },
    {
      "epoch": 12.938635686624963,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.010021319955623465,
      "loss": 0.7958,
      "num_input_tokens_seen": 50432320,
      "step": 86870
    },
    {
      "epoch": 12.939380399165922,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.010019480869943339,
      "loss": 0.7797,
      "num_input_tokens_seen": 50435072,
      "step": 86875
    },
    {
      "epoch": 12.94012511170688,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.010017641868404028,
      "loss": 0.8066,
      "num_input_tokens_seen": 50437984,
      "step": 86880
    },
    {
      "epoch": 12.94086982424784,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.010015802951036596,
      "loss": 0.8042,
      "num_input_tokens_seen": 50440800,
      "step": 86885
    },
    {
      "epoch": 12.9416145367888,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.010013964117872103,
      "loss": 0.8032,
      "num_input_tokens_seen": 50443808,
      "step": 86890
    },
    {
      "epoch": 12.942359249329758,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.010012125368941619,
      "loss": 0.8192,
      "num_input_tokens_seen": 50447264,
      "step": 86895
    },
    {
      "epoch": 12.943103961870717,
      "grad_norm": 0.203125,
      "learning_rate": 0.010010286704276207,
      "loss": 0.8307,
      "num_input_tokens_seen": 50450144,
      "step": 86900
    },
    {
      "epoch": 12.943848674411678,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.010008448123906933,
      "loss": 0.7737,
      "num_input_tokens_seen": 50452800,
      "step": 86905
    },
    {
      "epoch": 12.944593386952636,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.01000660962786485,
      "loss": 0.8012,
      "num_input_tokens_seen": 50455744,
      "step": 86910
    },
    {
      "epoch": 12.945338099493595,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.010004771216181027,
      "loss": 0.8052,
      "num_input_tokens_seen": 50458624,
      "step": 86915
    },
    {
      "epoch": 12.946082812034554,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.010002932888886514,
      "loss": 0.7871,
      "num_input_tokens_seen": 50461536,
      "step": 86920
    },
    {
      "epoch": 12.946827524575514,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.010001094646012374,
      "loss": 0.8134,
      "num_input_tokens_seen": 50464640,
      "step": 86925
    },
    {
      "epoch": 12.947572237116473,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.009999256487589658,
      "loss": 0.8166,
      "num_input_tokens_seen": 50467552,
      "step": 86930
    },
    {
      "epoch": 12.948316949657432,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.009997418413649421,
      "loss": 0.8004,
      "num_input_tokens_seen": 50470656,
      "step": 86935
    },
    {
      "epoch": 12.94906166219839,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.009995580424222718,
      "loss": 0.8008,
      "num_input_tokens_seen": 50473536,
      "step": 86940
    },
    {
      "epoch": 12.949806374739351,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.009993742519340591,
      "loss": 0.7891,
      "num_input_tokens_seen": 50476480,
      "step": 86945
    },
    {
      "epoch": 12.95055108728031,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.009991904699034103,
      "loss": 0.8074,
      "num_input_tokens_seen": 50479392,
      "step": 86950
    },
    {
      "epoch": 12.951295799821269,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.00999006696333429,
      "loss": 0.7989,
      "num_input_tokens_seen": 50482432,
      "step": 86955
    },
    {
      "epoch": 12.952040512362228,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.00998822931227221,
      "loss": 0.8145,
      "num_input_tokens_seen": 50485184,
      "step": 86960
    },
    {
      "epoch": 12.952785224903188,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.009986391745878898,
      "loss": 0.8004,
      "num_input_tokens_seen": 50488256,
      "step": 86965
    },
    {
      "epoch": 12.953529937444147,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.009984554264185397,
      "loss": 0.8066,
      "num_input_tokens_seen": 50491136,
      "step": 86970
    },
    {
      "epoch": 12.954274649985106,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.009982716867222763,
      "loss": 0.7999,
      "num_input_tokens_seen": 50494048,
      "step": 86975
    },
    {
      "epoch": 12.955019362526064,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.00998087955502202,
      "loss": 0.7969,
      "num_input_tokens_seen": 50496864,
      "step": 86980
    },
    {
      "epoch": 12.955764075067025,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.009979042327614223,
      "loss": 0.7899,
      "num_input_tokens_seen": 50500096,
      "step": 86985
    },
    {
      "epoch": 12.956508787607984,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.009977205185030398,
      "loss": 0.7847,
      "num_input_tokens_seen": 50503136,
      "step": 86990
    },
    {
      "epoch": 12.957253500148942,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.009975368127301591,
      "loss": 0.8068,
      "num_input_tokens_seen": 50506080,
      "step": 86995
    },
    {
      "epoch": 12.957998212689901,
      "grad_norm": 0.046875,
      "learning_rate": 0.009973531154458827,
      "loss": 0.7859,
      "num_input_tokens_seen": 50508800,
      "step": 87000
    },
    {
      "epoch": 12.958742925230862,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.00997169426653315,
      "loss": 0.8181,
      "num_input_tokens_seen": 50511616,
      "step": 87005
    },
    {
      "epoch": 12.95948763777182,
      "grad_norm": 0.048828125,
      "learning_rate": 0.009969857463555587,
      "loss": 0.7936,
      "num_input_tokens_seen": 50514656,
      "step": 87010
    },
    {
      "epoch": 12.96023235031278,
      "grad_norm": 0.01904296875,
      "learning_rate": 0.009968020745557166,
      "loss": 0.8083,
      "num_input_tokens_seen": 50517568,
      "step": 87015
    },
    {
      "epoch": 12.960977062853738,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.009966184112568926,
      "loss": 0.7869,
      "num_input_tokens_seen": 50520352,
      "step": 87020
    },
    {
      "epoch": 12.961721775394697,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.009964347564621884,
      "loss": 0.7896,
      "num_input_tokens_seen": 50523296,
      "step": 87025
    },
    {
      "epoch": 12.962466487935657,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.009962511101747076,
      "loss": 0.7873,
      "num_input_tokens_seen": 50526144,
      "step": 87030
    },
    {
      "epoch": 12.963211200476616,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.009960674723975518,
      "loss": 0.8012,
      "num_input_tokens_seen": 50529120,
      "step": 87035
    },
    {
      "epoch": 12.963955913017575,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.009958838431338244,
      "loss": 0.794,
      "num_input_tokens_seen": 50532064,
      "step": 87040
    },
    {
      "epoch": 12.964700625558535,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.009957002223866267,
      "loss": 0.7943,
      "num_input_tokens_seen": 50534816,
      "step": 87045
    },
    {
      "epoch": 12.965445338099494,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.009955166101590614,
      "loss": 0.802,
      "num_input_tokens_seen": 50537568,
      "step": 87050
    },
    {
      "epoch": 12.966190050640453,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.009953330064542307,
      "loss": 0.7944,
      "num_input_tokens_seen": 50540512,
      "step": 87055
    },
    {
      "epoch": 12.966934763181412,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.009951494112752353,
      "loss": 0.8028,
      "num_input_tokens_seen": 50543360,
      "step": 87060
    },
    {
      "epoch": 12.96767947572237,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.00994965824625178,
      "loss": 0.8184,
      "num_input_tokens_seen": 50546368,
      "step": 87065
    },
    {
      "epoch": 12.96842418826333,
      "grad_norm": 0.044921875,
      "learning_rate": 0.00994782246507159,
      "loss": 0.793,
      "num_input_tokens_seen": 50549344,
      "step": 87070
    },
    {
      "epoch": 12.96916890080429,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.009945986769242812,
      "loss": 0.7839,
      "num_input_tokens_seen": 50552192,
      "step": 87075
    },
    {
      "epoch": 12.969913613345248,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.009944151158796443,
      "loss": 0.7865,
      "num_input_tokens_seen": 50554848,
      "step": 87080
    },
    {
      "epoch": 12.970658325886207,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.009942315633763509,
      "loss": 0.7908,
      "num_input_tokens_seen": 50558080,
      "step": 87085
    },
    {
      "epoch": 12.971403038427168,
      "grad_norm": 0.048828125,
      "learning_rate": 0.009940480194175014,
      "loss": 0.7649,
      "num_input_tokens_seen": 50561280,
      "step": 87090
    },
    {
      "epoch": 12.972147750968126,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.00993864484006196,
      "loss": 0.7882,
      "num_input_tokens_seen": 50564064,
      "step": 87095
    },
    {
      "epoch": 12.972892463509085,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.009936809571455358,
      "loss": 0.7913,
      "num_input_tokens_seen": 50566912,
      "step": 87100
    },
    {
      "epoch": 12.973637176050044,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.009934974388386208,
      "loss": 0.7787,
      "num_input_tokens_seen": 50570048,
      "step": 87105
    },
    {
      "epoch": 12.974381888591004,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.009933139290885523,
      "loss": 0.7912,
      "num_input_tokens_seen": 50572832,
      "step": 87110
    },
    {
      "epoch": 12.975126601131963,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.009931304278984294,
      "loss": 0.8091,
      "num_input_tokens_seen": 50575616,
      "step": 87115
    },
    {
      "epoch": 12.975871313672922,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.009929469352713534,
      "loss": 0.7944,
      "num_input_tokens_seen": 50578560,
      "step": 87120
    },
    {
      "epoch": 12.97661602621388,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.009927634512104232,
      "loss": 0.8312,
      "num_input_tokens_seen": 50581216,
      "step": 87125
    },
    {
      "epoch": 12.977360738754841,
      "grad_norm": 0.0181884765625,
      "learning_rate": 0.009925799757187393,
      "loss": 0.8089,
      "num_input_tokens_seen": 50583968,
      "step": 87130
    },
    {
      "epoch": 12.9781054512958,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.009923965087994011,
      "loss": 0.8004,
      "num_input_tokens_seen": 50586752,
      "step": 87135
    },
    {
      "epoch": 12.978850163836759,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.009922130504555074,
      "loss": 0.7939,
      "num_input_tokens_seen": 50589696,
      "step": 87140
    },
    {
      "epoch": 12.979594876377718,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.009920296006901586,
      "loss": 0.8163,
      "num_input_tokens_seen": 50592544,
      "step": 87145
    },
    {
      "epoch": 12.980339588918678,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.009918461595064527,
      "loss": 0.8656,
      "num_input_tokens_seen": 50595680,
      "step": 87150
    },
    {
      "epoch": 12.981084301459637,
      "grad_norm": 0.064453125,
      "learning_rate": 0.009916627269074902,
      "loss": 0.8094,
      "num_input_tokens_seen": 50598688,
      "step": 87155
    },
    {
      "epoch": 12.981829014000596,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.00991479302896369,
      "loss": 0.7939,
      "num_input_tokens_seen": 50601504,
      "step": 87160
    },
    {
      "epoch": 12.982573726541554,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.009912958874761886,
      "loss": 0.7979,
      "num_input_tokens_seen": 50604640,
      "step": 87165
    },
    {
      "epoch": 12.983318439082515,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00991112480650047,
      "loss": 0.7732,
      "num_input_tokens_seen": 50607680,
      "step": 87170
    },
    {
      "epoch": 12.984063151623474,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.009909290824210422,
      "loss": 0.8293,
      "num_input_tokens_seen": 50611488,
      "step": 87175
    },
    {
      "epoch": 12.984807864164432,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.009907456927922737,
      "loss": 0.7882,
      "num_input_tokens_seen": 50614496,
      "step": 87180
    },
    {
      "epoch": 12.985552576705391,
      "grad_norm": 0.03125,
      "learning_rate": 0.009905623117668386,
      "loss": 0.7937,
      "num_input_tokens_seen": 50617312,
      "step": 87185
    },
    {
      "epoch": 12.986297289246352,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.009903789393478362,
      "loss": 0.8011,
      "num_input_tokens_seen": 50620256,
      "step": 87190
    },
    {
      "epoch": 12.98704200178731,
      "grad_norm": 0.150390625,
      "learning_rate": 0.009901955755383631,
      "loss": 0.8346,
      "num_input_tokens_seen": 50623072,
      "step": 87195
    },
    {
      "epoch": 12.98778671432827,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.009900122203415183,
      "loss": 0.82,
      "num_input_tokens_seen": 50625568,
      "step": 87200
    },
    {
      "epoch": 12.988531426869228,
      "grad_norm": 0.01806640625,
      "learning_rate": 0.009898288737603978,
      "loss": 0.7856,
      "num_input_tokens_seen": 50628800,
      "step": 87205
    },
    {
      "epoch": 12.989276139410187,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.009896455357981006,
      "loss": 0.7967,
      "num_input_tokens_seen": 50631488,
      "step": 87210
    },
    {
      "epoch": 12.990020851951147,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.009894622064577235,
      "loss": 0.78,
      "num_input_tokens_seen": 50634176,
      "step": 87215
    },
    {
      "epoch": 12.990765564492106,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.00989278885742363,
      "loss": 0.8183,
      "num_input_tokens_seen": 50637088,
      "step": 87220
    },
    {
      "epoch": 12.991510277033065,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.009890955736551169,
      "loss": 0.7921,
      "num_input_tokens_seen": 50639968,
      "step": 87225
    },
    {
      "epoch": 12.992254989574025,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.009889122701990816,
      "loss": 0.8235,
      "num_input_tokens_seen": 50642976,
      "step": 87230
    },
    {
      "epoch": 12.992999702114984,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.009887289753773544,
      "loss": 0.7978,
      "num_input_tokens_seen": 50646144,
      "step": 87235
    },
    {
      "epoch": 12.993744414655943,
      "grad_norm": 0.02001953125,
      "learning_rate": 0.00988545689193031,
      "loss": 0.802,
      "num_input_tokens_seen": 50649280,
      "step": 87240
    },
    {
      "epoch": 12.994489127196902,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00988362411649209,
      "loss": 0.7916,
      "num_input_tokens_seen": 50651808,
      "step": 87245
    },
    {
      "epoch": 12.99523383973786,
      "grad_norm": 0.018798828125,
      "learning_rate": 0.009881791427489841,
      "loss": 0.7918,
      "num_input_tokens_seen": 50654912,
      "step": 87250
    },
    {
      "epoch": 12.995978552278821,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.009879958824954519,
      "loss": 0.7858,
      "num_input_tokens_seen": 50657600,
      "step": 87255
    },
    {
      "epoch": 12.99672326481978,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.009878126308917089,
      "loss": 0.7949,
      "num_input_tokens_seen": 50660512,
      "step": 87260
    },
    {
      "epoch": 12.997467977360738,
      "grad_norm": 0.03515625,
      "learning_rate": 0.00987629387940851,
      "loss": 0.7994,
      "num_input_tokens_seen": 50663392,
      "step": 87265
    },
    {
      "epoch": 12.998212689901697,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.009874461536459744,
      "loss": 0.8055,
      "num_input_tokens_seen": 50666592,
      "step": 87270
    },
    {
      "epoch": 12.998957402442658,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.009872629280101733,
      "loss": 0.7956,
      "num_input_tokens_seen": 50669664,
      "step": 87275
    },
    {
      "epoch": 12.999702114983616,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.009870797110365441,
      "loss": 0.8272,
      "num_input_tokens_seen": 50672352,
      "step": 87280
    },
    {
      "epoch": 13.0,
      "eval_loss": 0.799648106098175,
      "eval_runtime": 70.5802,
      "eval_samples_per_second": 42.278,
      "eval_steps_per_second": 10.57,
      "num_input_tokens_seen": 50673016,
      "step": 87282
    },
    {
      "epoch": 13.000446827524575,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.009868965027281821,
      "loss": 0.8028,
      "num_input_tokens_seen": 50675000,
      "step": 87285
    },
    {
      "epoch": 13.001191540065534,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.009867133030881823,
      "loss": 0.8006,
      "num_input_tokens_seen": 50678072,
      "step": 87290
    },
    {
      "epoch": 13.001936252606495,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.009865301121196397,
      "loss": 0.8331,
      "num_input_tokens_seen": 50680952,
      "step": 87295
    },
    {
      "epoch": 13.002680965147453,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.009863469298256482,
      "loss": 0.8074,
      "num_input_tokens_seen": 50683928,
      "step": 87300
    },
    {
      "epoch": 13.003425677688412,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.00986163756209304,
      "loss": 0.791,
      "num_input_tokens_seen": 50687096,
      "step": 87305
    },
    {
      "epoch": 13.00417039022937,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.009859805912737002,
      "loss": 0.7843,
      "num_input_tokens_seen": 50689848,
      "step": 87310
    },
    {
      "epoch": 13.004915102770331,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.009857974350219328,
      "loss": 0.7911,
      "num_input_tokens_seen": 50692632,
      "step": 87315
    },
    {
      "epoch": 13.00565981531129,
      "grad_norm": 0.03125,
      "learning_rate": 0.009856142874570943,
      "loss": 0.7854,
      "num_input_tokens_seen": 50695576,
      "step": 87320
    },
    {
      "epoch": 13.006404527852249,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.009854311485822801,
      "loss": 0.7974,
      "num_input_tokens_seen": 50698872,
      "step": 87325
    },
    {
      "epoch": 13.007149240393208,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.009852480184005839,
      "loss": 0.8212,
      "num_input_tokens_seen": 50701752,
      "step": 87330
    },
    {
      "epoch": 13.007893952934168,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.009850648969150985,
      "loss": 0.7837,
      "num_input_tokens_seen": 50704632,
      "step": 87335
    },
    {
      "epoch": 13.008638665475127,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.009848817841289189,
      "loss": 0.7953,
      "num_input_tokens_seen": 50707448,
      "step": 87340
    },
    {
      "epoch": 13.009383378016086,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.009846986800451374,
      "loss": 0.8014,
      "num_input_tokens_seen": 50710424,
      "step": 87345
    },
    {
      "epoch": 13.010128090557044,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.009845155846668485,
      "loss": 0.799,
      "num_input_tokens_seen": 50713400,
      "step": 87350
    },
    {
      "epoch": 13.010872803098005,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.009843324979971448,
      "loss": 0.7929,
      "num_input_tokens_seen": 50716568,
      "step": 87355
    },
    {
      "epoch": 13.011617515638964,
      "grad_norm": 0.017578125,
      "learning_rate": 0.009841494200391194,
      "loss": 0.7909,
      "num_input_tokens_seen": 50719352,
      "step": 87360
    },
    {
      "epoch": 13.012362228179922,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.009839663507958651,
      "loss": 0.8108,
      "num_input_tokens_seen": 50722136,
      "step": 87365
    },
    {
      "epoch": 13.013106940720881,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.00983783290270475,
      "loss": 0.8084,
      "num_input_tokens_seen": 50725176,
      "step": 87370
    },
    {
      "epoch": 13.013851653261842,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.009836002384660418,
      "loss": 0.823,
      "num_input_tokens_seen": 50727992,
      "step": 87375
    },
    {
      "epoch": 13.0145963658028,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.009834171953856573,
      "loss": 0.8186,
      "num_input_tokens_seen": 50730648,
      "step": 87380
    },
    {
      "epoch": 13.01534107834376,
      "grad_norm": 0.0390625,
      "learning_rate": 0.009832341610324145,
      "loss": 0.8014,
      "num_input_tokens_seen": 50733592,
      "step": 87385
    },
    {
      "epoch": 13.016085790884718,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.009830511354094051,
      "loss": 0.8024,
      "num_input_tokens_seen": 50736536,
      "step": 87390
    },
    {
      "epoch": 13.016830503425677,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.00982868118519722,
      "loss": 0.7949,
      "num_input_tokens_seen": 50739512,
      "step": 87395
    },
    {
      "epoch": 13.017575215966637,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.009826851103664554,
      "loss": 0.8237,
      "num_input_tokens_seen": 50742168,
      "step": 87400
    },
    {
      "epoch": 13.018319928507596,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.009825021109526988,
      "loss": 0.7924,
      "num_input_tokens_seen": 50745048,
      "step": 87405
    },
    {
      "epoch": 13.019064641048555,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.009823191202815431,
      "loss": 0.8076,
      "num_input_tokens_seen": 50747992,
      "step": 87410
    },
    {
      "epoch": 13.019809353589514,
      "grad_norm": 0.03125,
      "learning_rate": 0.009821361383560793,
      "loss": 0.8093,
      "num_input_tokens_seen": 50750744,
      "step": 87415
    },
    {
      "epoch": 13.020554066130474,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.009819531651793995,
      "loss": 0.8087,
      "num_input_tokens_seen": 50753624,
      "step": 87420
    },
    {
      "epoch": 13.021298778671433,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.009817702007545942,
      "loss": 0.8037,
      "num_input_tokens_seen": 50756184,
      "step": 87425
    },
    {
      "epoch": 13.022043491212392,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.009815872450847552,
      "loss": 0.7922,
      "num_input_tokens_seen": 50759288,
      "step": 87430
    },
    {
      "epoch": 13.02278820375335,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.009814042981729717,
      "loss": 0.8037,
      "num_input_tokens_seen": 50762456,
      "step": 87435
    },
    {
      "epoch": 13.023532916294311,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.009812213600223364,
      "loss": 0.8061,
      "num_input_tokens_seen": 50765368,
      "step": 87440
    },
    {
      "epoch": 13.02427762883527,
      "grad_norm": 0.03125,
      "learning_rate": 0.009810384306359384,
      "loss": 0.79,
      "num_input_tokens_seen": 50768280,
      "step": 87445
    },
    {
      "epoch": 13.025022341376228,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.009808555100168693,
      "loss": 0.7905,
      "num_input_tokens_seen": 50771000,
      "step": 87450
    },
    {
      "epoch": 13.025767053917187,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.009806725981682186,
      "loss": 0.7916,
      "num_input_tokens_seen": 50773688,
      "step": 87455
    },
    {
      "epoch": 13.026511766458148,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.009804896950930762,
      "loss": 0.7958,
      "num_input_tokens_seen": 50776632,
      "step": 87460
    },
    {
      "epoch": 13.027256478999107,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.009803068007945328,
      "loss": 0.809,
      "num_input_tokens_seen": 50779384,
      "step": 87465
    },
    {
      "epoch": 13.028001191540065,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.00980123915275677,
      "loss": 0.8002,
      "num_input_tokens_seen": 50782328,
      "step": 87470
    },
    {
      "epoch": 13.028745904081024,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.009799410385396001,
      "loss": 0.7674,
      "num_input_tokens_seen": 50784984,
      "step": 87475
    },
    {
      "epoch": 13.029490616621985,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0097975817058939,
      "loss": 0.8022,
      "num_input_tokens_seen": 50787928,
      "step": 87480
    },
    {
      "epoch": 13.030235329162943,
      "grad_norm": 0.037109375,
      "learning_rate": 0.009795753114281379,
      "loss": 0.7846,
      "num_input_tokens_seen": 50790680,
      "step": 87485
    },
    {
      "epoch": 13.030980041703902,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.009793924610589316,
      "loss": 0.7977,
      "num_input_tokens_seen": 50793752,
      "step": 87490
    },
    {
      "epoch": 13.03172475424486,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.009792096194848603,
      "loss": 0.7908,
      "num_input_tokens_seen": 50796856,
      "step": 87495
    },
    {
      "epoch": 13.032469466785821,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.009790267867090132,
      "loss": 0.8011,
      "num_input_tokens_seen": 50799512,
      "step": 87500
    },
    {
      "epoch": 13.03321417932678,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.009788439627344788,
      "loss": 0.784,
      "num_input_tokens_seen": 50802616,
      "step": 87505
    },
    {
      "epoch": 13.033958891867739,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.009786611475643468,
      "loss": 0.8131,
      "num_input_tokens_seen": 50805816,
      "step": 87510
    },
    {
      "epoch": 13.034703604408698,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.009784783412017037,
      "loss": 0.7856,
      "num_input_tokens_seen": 50808472,
      "step": 87515
    },
    {
      "epoch": 13.035448316949658,
      "grad_norm": 0.019775390625,
      "learning_rate": 0.0097829554364964,
      "loss": 0.8022,
      "num_input_tokens_seen": 50811448,
      "step": 87520
    },
    {
      "epoch": 13.036193029490617,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.009781127549112423,
      "loss": 0.7958,
      "num_input_tokens_seen": 50814520,
      "step": 87525
    },
    {
      "epoch": 13.036937742031576,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.009779299749895996,
      "loss": 0.8019,
      "num_input_tokens_seen": 50817272,
      "step": 87530
    },
    {
      "epoch": 13.037682454572534,
      "grad_norm": 0.03125,
      "learning_rate": 0.009777472038877992,
      "loss": 0.8044,
      "num_input_tokens_seen": 50820248,
      "step": 87535
    },
    {
      "epoch": 13.038427167113495,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.009775644416089286,
      "loss": 0.7887,
      "num_input_tokens_seen": 50823288,
      "step": 87540
    },
    {
      "epoch": 13.039171879654454,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.009773816881560761,
      "loss": 0.8122,
      "num_input_tokens_seen": 50826200,
      "step": 87545
    },
    {
      "epoch": 13.039916592195413,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.00977198943532329,
      "loss": 0.8493,
      "num_input_tokens_seen": 50829272,
      "step": 87550
    },
    {
      "epoch": 13.040661304736371,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.009770162077407739,
      "loss": 0.7952,
      "num_input_tokens_seen": 50831864,
      "step": 87555
    },
    {
      "epoch": 13.041406017277332,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.009768334807844987,
      "loss": 0.8015,
      "num_input_tokens_seen": 50834904,
      "step": 87560
    },
    {
      "epoch": 13.04215072981829,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.009766507626665904,
      "loss": 0.7981,
      "num_input_tokens_seen": 50837720,
      "step": 87565
    },
    {
      "epoch": 13.04289544235925,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.009764680533901355,
      "loss": 0.8004,
      "num_input_tokens_seen": 50840568,
      "step": 87570
    },
    {
      "epoch": 13.043640154900208,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.009762853529582201,
      "loss": 0.8112,
      "num_input_tokens_seen": 50843640,
      "step": 87575
    },
    {
      "epoch": 13.044384867441167,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.009761026613739319,
      "loss": 0.8016,
      "num_input_tokens_seen": 50846712,
      "step": 87580
    },
    {
      "epoch": 13.045129579982127,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.009759199786403569,
      "loss": 0.7854,
      "num_input_tokens_seen": 50849752,
      "step": 87585
    },
    {
      "epoch": 13.045874292523086,
      "grad_norm": 0.04296875,
      "learning_rate": 0.009757373047605806,
      "loss": 0.8028,
      "num_input_tokens_seen": 50852664,
      "step": 87590
    },
    {
      "epoch": 13.046619005064045,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0097555463973769,
      "loss": 0.8058,
      "num_input_tokens_seen": 50855352,
      "step": 87595
    },
    {
      "epoch": 13.047363717605004,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.009753719835747713,
      "loss": 0.8016,
      "num_input_tokens_seen": 50858168,
      "step": 87600
    },
    {
      "epoch": 13.048108430145964,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.009751893362749089,
      "loss": 0.7897,
      "num_input_tokens_seen": 50861304,
      "step": 87605
    },
    {
      "epoch": 13.048853142686923,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.009750066978411898,
      "loss": 0.8121,
      "num_input_tokens_seen": 50864152,
      "step": 87610
    },
    {
      "epoch": 13.049597855227882,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.00974824068276699,
      "loss": 0.8084,
      "num_input_tokens_seen": 50867224,
      "step": 87615
    },
    {
      "epoch": 13.05034256776884,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.009746414475845216,
      "loss": 0.8076,
      "num_input_tokens_seen": 50870200,
      "step": 87620
    },
    {
      "epoch": 13.051087280309801,
      "grad_norm": 0.041015625,
      "learning_rate": 0.009744588357677431,
      "loss": 0.8088,
      "num_input_tokens_seen": 50873112,
      "step": 87625
    },
    {
      "epoch": 13.05183199285076,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.009742762328294479,
      "loss": 0.7904,
      "num_input_tokens_seen": 50876088,
      "step": 87630
    },
    {
      "epoch": 13.052576705391719,
      "grad_norm": 0.0184326171875,
      "learning_rate": 0.009740936387727222,
      "loss": 0.7877,
      "num_input_tokens_seen": 50879160,
      "step": 87635
    },
    {
      "epoch": 13.053321417932677,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.009739110536006491,
      "loss": 0.8142,
      "num_input_tokens_seen": 50882232,
      "step": 87640
    },
    {
      "epoch": 13.054066130473638,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.009737284773163149,
      "loss": 0.7785,
      "num_input_tokens_seen": 50885848,
      "step": 87645
    },
    {
      "epoch": 13.054810843014597,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.009735459099228027,
      "loss": 0.8272,
      "num_input_tokens_seen": 50888856,
      "step": 87650
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.009733633514231977,
      "loss": 0.8134,
      "num_input_tokens_seen": 50891640,
      "step": 87655
    },
    {
      "epoch": 13.056300268096514,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.009731808018205836,
      "loss": 0.7769,
      "num_input_tokens_seen": 50894296,
      "step": 87660
    },
    {
      "epoch": 13.057044980637475,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.00972998261118044,
      "loss": 0.7923,
      "num_input_tokens_seen": 50897208,
      "step": 87665
    },
    {
      "epoch": 13.057789693178433,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.009728157293186634,
      "loss": 0.8063,
      "num_input_tokens_seen": 50900088,
      "step": 87670
    },
    {
      "epoch": 13.058534405719392,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00972633206425525,
      "loss": 0.7901,
      "num_input_tokens_seen": 50903096,
      "step": 87675
    },
    {
      "epoch": 13.059279118260351,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.009724506924417129,
      "loss": 0.7996,
      "num_input_tokens_seen": 50906072,
      "step": 87680
    },
    {
      "epoch": 13.060023830801311,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0097226818737031,
      "loss": 0.825,
      "num_input_tokens_seen": 50908888,
      "step": 87685
    },
    {
      "epoch": 13.06076854334227,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.009720856912144001,
      "loss": 0.7888,
      "num_input_tokens_seen": 50911640,
      "step": 87690
    },
    {
      "epoch": 13.061513255883229,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.009719032039770658,
      "loss": 0.7757,
      "num_input_tokens_seen": 50914552,
      "step": 87695
    },
    {
      "epoch": 13.062257968424188,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.009717207256613895,
      "loss": 0.8117,
      "num_input_tokens_seen": 50917400,
      "step": 87700
    },
    {
      "epoch": 13.063002680965148,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.009715382562704553,
      "loss": 0.8056,
      "num_input_tokens_seen": 50920216,
      "step": 87705
    },
    {
      "epoch": 13.063747393506107,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.009713557958073443,
      "loss": 0.8194,
      "num_input_tokens_seen": 50923256,
      "step": 87710
    },
    {
      "epoch": 13.064492106047066,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.009711733442751409,
      "loss": 0.7801,
      "num_input_tokens_seen": 50926040,
      "step": 87715
    },
    {
      "epoch": 13.065236818588025,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.009709909016769257,
      "loss": 0.7855,
      "num_input_tokens_seen": 50928952,
      "step": 87720
    },
    {
      "epoch": 13.065981531128985,
      "grad_norm": 0.0186767578125,
      "learning_rate": 0.00970808468015782,
      "loss": 0.7999,
      "num_input_tokens_seen": 50931960,
      "step": 87725
    },
    {
      "epoch": 13.066726243669944,
      "grad_norm": 0.048828125,
      "learning_rate": 0.009706260432947906,
      "loss": 0.7859,
      "num_input_tokens_seen": 50934776,
      "step": 87730
    },
    {
      "epoch": 13.067470956210903,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00970443627517035,
      "loss": 0.794,
      "num_input_tokens_seen": 50937592,
      "step": 87735
    },
    {
      "epoch": 13.068215668751861,
      "grad_norm": 0.03125,
      "learning_rate": 0.00970261220685596,
      "loss": 0.8008,
      "num_input_tokens_seen": 50940504,
      "step": 87740
    },
    {
      "epoch": 13.06896038129282,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.009700788228035547,
      "loss": 0.8007,
      "num_input_tokens_seen": 50943192,
      "step": 87745
    },
    {
      "epoch": 13.06970509383378,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.00969896433873994,
      "loss": 0.799,
      "num_input_tokens_seen": 50946136,
      "step": 87750
    },
    {
      "epoch": 13.07044980637474,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.009697140538999935,
      "loss": 0.7766,
      "num_input_tokens_seen": 50949304,
      "step": 87755
    },
    {
      "epoch": 13.071194518915698,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.009695316828846357,
      "loss": 0.7995,
      "num_input_tokens_seen": 50952152,
      "step": 87760
    },
    {
      "epoch": 13.071939231456657,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00969349320831,
      "loss": 0.8038,
      "num_input_tokens_seen": 50955288,
      "step": 87765
    },
    {
      "epoch": 13.072683943997617,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.009691669677421692,
      "loss": 0.8122,
      "num_input_tokens_seen": 50957816,
      "step": 87770
    },
    {
      "epoch": 13.073428656538576,
      "grad_norm": 0.080078125,
      "learning_rate": 0.009689846236212227,
      "loss": 0.806,
      "num_input_tokens_seen": 50960504,
      "step": 87775
    },
    {
      "epoch": 13.074173369079535,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.009688022884712408,
      "loss": 0.7663,
      "num_input_tokens_seen": 50963256,
      "step": 87780
    },
    {
      "epoch": 13.074918081620494,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.00968619962295305,
      "loss": 0.8149,
      "num_input_tokens_seen": 50966296,
      "step": 87785
    },
    {
      "epoch": 13.075662794161454,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.009684376450964946,
      "loss": 0.7904,
      "num_input_tokens_seen": 50969144,
      "step": 87790
    },
    {
      "epoch": 13.076407506702413,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.009682553368778902,
      "loss": 0.7746,
      "num_input_tokens_seen": 50971864,
      "step": 87795
    },
    {
      "epoch": 13.077152219243372,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.009680730376425707,
      "loss": 0.7768,
      "num_input_tokens_seen": 50974776,
      "step": 87800
    },
    {
      "epoch": 13.07789693178433,
      "grad_norm": 0.044921875,
      "learning_rate": 0.009678907473936172,
      "loss": 0.7649,
      "num_input_tokens_seen": 50977432,
      "step": 87805
    },
    {
      "epoch": 13.078641644325291,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.009677084661341082,
      "loss": 0.8152,
      "num_input_tokens_seen": 50980568,
      "step": 87810
    },
    {
      "epoch": 13.07938635686625,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.009675261938671242,
      "loss": 0.791,
      "num_input_tokens_seen": 50983672,
      "step": 87815
    },
    {
      "epoch": 13.080131069407209,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.009673439305957442,
      "loss": 0.7857,
      "num_input_tokens_seen": 50986648,
      "step": 87820
    },
    {
      "epoch": 13.080875781948167,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.009671616763230466,
      "loss": 0.8041,
      "num_input_tokens_seen": 50989816,
      "step": 87825
    },
    {
      "epoch": 13.081620494489128,
      "grad_norm": 0.03515625,
      "learning_rate": 0.009669794310521115,
      "loss": 0.796,
      "num_input_tokens_seen": 50992696,
      "step": 87830
    },
    {
      "epoch": 13.082365207030087,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.009667971947860163,
      "loss": 0.7966,
      "num_input_tokens_seen": 50995608,
      "step": 87835
    },
    {
      "epoch": 13.083109919571045,
      "grad_norm": 0.060546875,
      "learning_rate": 0.009666149675278412,
      "loss": 0.792,
      "num_input_tokens_seen": 50998392,
      "step": 87840
    },
    {
      "epoch": 13.083854632112004,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.009664327492806634,
      "loss": 0.8017,
      "num_input_tokens_seen": 51001336,
      "step": 87845
    },
    {
      "epoch": 13.084599344652965,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.009662505400475628,
      "loss": 0.7869,
      "num_input_tokens_seen": 51004024,
      "step": 87850
    },
    {
      "epoch": 13.085344057193923,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.009660683398316168,
      "loss": 0.8037,
      "num_input_tokens_seen": 51007320,
      "step": 87855
    },
    {
      "epoch": 13.086088769734882,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.009658861486359032,
      "loss": 0.7986,
      "num_input_tokens_seen": 51010456,
      "step": 87860
    },
    {
      "epoch": 13.086833482275841,
      "grad_norm": 0.0201416015625,
      "learning_rate": 0.009657039664635008,
      "loss": 0.8038,
      "num_input_tokens_seen": 51013432,
      "step": 87865
    },
    {
      "epoch": 13.087578194816802,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.009655217933174861,
      "loss": 0.8159,
      "num_input_tokens_seen": 51016280,
      "step": 87870
    },
    {
      "epoch": 13.08832290735776,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.00965339629200938,
      "loss": 0.7815,
      "num_input_tokens_seen": 51019448,
      "step": 87875
    },
    {
      "epoch": 13.089067619898719,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.009651574741169332,
      "loss": 0.8157,
      "num_input_tokens_seen": 51022360,
      "step": 87880
    },
    {
      "epoch": 13.089812332439678,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.009649753280685493,
      "loss": 0.7884,
      "num_input_tokens_seen": 51025080,
      "step": 87885
    },
    {
      "epoch": 13.090557044980638,
      "grad_norm": 0.0191650390625,
      "learning_rate": 0.009647931910588634,
      "loss": 0.7823,
      "num_input_tokens_seen": 51027832,
      "step": 87890
    },
    {
      "epoch": 13.091301757521597,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.00964611063090953,
      "loss": 0.7678,
      "num_input_tokens_seen": 51030808,
      "step": 87895
    },
    {
      "epoch": 13.092046470062556,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.009644289441678944,
      "loss": 0.802,
      "num_input_tokens_seen": 51034104,
      "step": 87900
    },
    {
      "epoch": 13.092791182603515,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00964246834292764,
      "loss": 0.7973,
      "num_input_tokens_seen": 51037048,
      "step": 87905
    },
    {
      "epoch": 13.093535895144473,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.009640647334686394,
      "loss": 0.7861,
      "num_input_tokens_seen": 51040184,
      "step": 87910
    },
    {
      "epoch": 13.094280607685434,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.00963882641698596,
      "loss": 0.7872,
      "num_input_tokens_seen": 51042808,
      "step": 87915
    },
    {
      "epoch": 13.095025320226393,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.009637005589857105,
      "loss": 0.807,
      "num_input_tokens_seen": 51045432,
      "step": 87920
    },
    {
      "epoch": 13.095770032767351,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00963518485333059,
      "loss": 0.7858,
      "num_input_tokens_seen": 51048536,
      "step": 87925
    },
    {
      "epoch": 13.09651474530831,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.009633364207437179,
      "loss": 0.7865,
      "num_input_tokens_seen": 51051448,
      "step": 87930
    },
    {
      "epoch": 13.09725945784927,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.009631543652207624,
      "loss": 0.8005,
      "num_input_tokens_seen": 51054328,
      "step": 87935
    },
    {
      "epoch": 13.09800417039023,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.009629723187672676,
      "loss": 0.8063,
      "num_input_tokens_seen": 51057080,
      "step": 87940
    },
    {
      "epoch": 13.098748882931188,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.009627902813863104,
      "loss": 0.8013,
      "num_input_tokens_seen": 51060120,
      "step": 87945
    },
    {
      "epoch": 13.099493595472147,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.009626082530809653,
      "loss": 0.7989,
      "num_input_tokens_seen": 51063160,
      "step": 87950
    },
    {
      "epoch": 13.100238308013108,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.009624262338543075,
      "loss": 0.7932,
      "num_input_tokens_seen": 51066136,
      "step": 87955
    },
    {
      "epoch": 13.100983020554066,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.009622442237094115,
      "loss": 0.8071,
      "num_input_tokens_seen": 51068888,
      "step": 87960
    },
    {
      "epoch": 13.101727733095025,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.009620622226493536,
      "loss": 0.7774,
      "num_input_tokens_seen": 51072056,
      "step": 87965
    },
    {
      "epoch": 13.102472445635984,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.009618802306772072,
      "loss": 0.8502,
      "num_input_tokens_seen": 51075032,
      "step": 87970
    },
    {
      "epoch": 13.103217158176944,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.009616982477960476,
      "loss": 0.8222,
      "num_input_tokens_seen": 51077944,
      "step": 87975
    },
    {
      "epoch": 13.103961870717903,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.009615162740089495,
      "loss": 0.8084,
      "num_input_tokens_seen": 51080632,
      "step": 87980
    },
    {
      "epoch": 13.104706583258862,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.009613343093189858,
      "loss": 0.7786,
      "num_input_tokens_seen": 51083352,
      "step": 87985
    },
    {
      "epoch": 13.10545129579982,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.009611523537292322,
      "loss": 0.7916,
      "num_input_tokens_seen": 51086360,
      "step": 87990
    },
    {
      "epoch": 13.106196008340781,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.009609704072427611,
      "loss": 0.7948,
      "num_input_tokens_seen": 51089144,
      "step": 87995
    },
    {
      "epoch": 13.10694072088174,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.009607884698626478,
      "loss": 0.7989,
      "num_input_tokens_seen": 51092088,
      "step": 88000
    },
    {
      "epoch": 13.107685433422699,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.009606065415919646,
      "loss": 0.7818,
      "num_input_tokens_seen": 51094968,
      "step": 88005
    },
    {
      "epoch": 13.108430145963657,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.009604246224337865,
      "loss": 0.8104,
      "num_input_tokens_seen": 51098136,
      "step": 88010
    },
    {
      "epoch": 13.109174858504618,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.009602427123911857,
      "loss": 0.7886,
      "num_input_tokens_seen": 51101048,
      "step": 88015
    },
    {
      "epoch": 13.109919571045577,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.009600608114672357,
      "loss": 0.8012,
      "num_input_tokens_seen": 51103736,
      "step": 88020
    },
    {
      "epoch": 13.110664283586535,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.009598789196650095,
      "loss": 0.8089,
      "num_input_tokens_seen": 51106840,
      "step": 88025
    },
    {
      "epoch": 13.111408996127494,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.009596970369875798,
      "loss": 0.817,
      "num_input_tokens_seen": 51109592,
      "step": 88030
    },
    {
      "epoch": 13.112153708668455,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.009595151634380197,
      "loss": 0.7847,
      "num_input_tokens_seen": 51112440,
      "step": 88035
    },
    {
      "epoch": 13.112898421209414,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.009593332990194011,
      "loss": 0.8083,
      "num_input_tokens_seen": 51115128,
      "step": 88040
    },
    {
      "epoch": 13.113643133750372,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.009591514437347976,
      "loss": 0.7943,
      "num_input_tokens_seen": 51117720,
      "step": 88045
    },
    {
      "epoch": 13.114387846291331,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.009589695975872807,
      "loss": 0.8012,
      "num_input_tokens_seen": 51120792,
      "step": 88050
    },
    {
      "epoch": 13.115132558832292,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.009587877605799225,
      "loss": 0.7888,
      "num_input_tokens_seen": 51123576,
      "step": 88055
    },
    {
      "epoch": 13.11587727137325,
      "grad_norm": 0.03125,
      "learning_rate": 0.009586059327157951,
      "loss": 0.801,
      "num_input_tokens_seen": 51126648,
      "step": 88060
    },
    {
      "epoch": 13.116621983914209,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.009584241139979697,
      "loss": 0.7941,
      "num_input_tokens_seen": 51129432,
      "step": 88065
    },
    {
      "epoch": 13.117366696455168,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.009582423044295189,
      "loss": 0.7813,
      "num_input_tokens_seen": 51132344,
      "step": 88070
    },
    {
      "epoch": 13.118111408996128,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.009580605040135134,
      "loss": 0.7984,
      "num_input_tokens_seen": 51135480,
      "step": 88075
    },
    {
      "epoch": 13.118856121537087,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.009578787127530252,
      "loss": 0.7996,
      "num_input_tokens_seen": 51138424,
      "step": 88080
    },
    {
      "epoch": 13.119600834078046,
      "grad_norm": 0.046875,
      "learning_rate": 0.00957696930651125,
      "loss": 0.8044,
      "num_input_tokens_seen": 51141272,
      "step": 88085
    },
    {
      "epoch": 13.120345546619005,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.009575151577108842,
      "loss": 0.7803,
      "num_input_tokens_seen": 51144248,
      "step": 88090
    },
    {
      "epoch": 13.121090259159963,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.009573333939353737,
      "loss": 0.7964,
      "num_input_tokens_seen": 51147256,
      "step": 88095
    },
    {
      "epoch": 13.121834971700924,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00957151639327663,
      "loss": 0.8171,
      "num_input_tokens_seen": 51150392,
      "step": 88100
    },
    {
      "epoch": 13.122579684241883,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.009569698938908244,
      "loss": 0.7824,
      "num_input_tokens_seen": 51153304,
      "step": 88105
    },
    {
      "epoch": 13.123324396782841,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.009567881576279273,
      "loss": 0.8104,
      "num_input_tokens_seen": 51156312,
      "step": 88110
    },
    {
      "epoch": 13.1240691093238,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.009566064305420422,
      "loss": 0.797,
      "num_input_tokens_seen": 51159224,
      "step": 88115
    },
    {
      "epoch": 13.12481382186476,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.009564247126362392,
      "loss": 0.7573,
      "num_input_tokens_seen": 51162232,
      "step": 88120
    },
    {
      "epoch": 13.12555853440572,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.009562430039135884,
      "loss": 0.8213,
      "num_input_tokens_seen": 51164984,
      "step": 88125
    },
    {
      "epoch": 13.126303246946678,
      "grad_norm": 0.03125,
      "learning_rate": 0.009560613043771588,
      "loss": 0.7896,
      "num_input_tokens_seen": 51167672,
      "step": 88130
    },
    {
      "epoch": 13.127047959487637,
      "grad_norm": 0.046875,
      "learning_rate": 0.009558796140300212,
      "loss": 0.7946,
      "num_input_tokens_seen": 51170392,
      "step": 88135
    },
    {
      "epoch": 13.127792672028598,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.009556979328752446,
      "loss": 0.7853,
      "num_input_tokens_seen": 51173048,
      "step": 88140
    },
    {
      "epoch": 13.128537384569556,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.009555162609158976,
      "loss": 0.7887,
      "num_input_tokens_seen": 51175960,
      "step": 88145
    },
    {
      "epoch": 13.129282097110515,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.009553345981550505,
      "loss": 0.7962,
      "num_input_tokens_seen": 51179032,
      "step": 88150
    },
    {
      "epoch": 13.130026809651474,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.009551529445957717,
      "loss": 0.7927,
      "num_input_tokens_seen": 51182008,
      "step": 88155
    },
    {
      "epoch": 13.130771522192434,
      "grad_norm": 0.033203125,
      "learning_rate": 0.009549713002411303,
      "loss": 0.8165,
      "num_input_tokens_seen": 51184952,
      "step": 88160
    },
    {
      "epoch": 13.131516234733393,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.009547896650941943,
      "loss": 0.7894,
      "num_input_tokens_seen": 51187960,
      "step": 88165
    },
    {
      "epoch": 13.132260947274352,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.009546080391580333,
      "loss": 0.8006,
      "num_input_tokens_seen": 51190808,
      "step": 88170
    },
    {
      "epoch": 13.13300565981531,
      "grad_norm": 0.033203125,
      "learning_rate": 0.009544264224357151,
      "loss": 0.7877,
      "num_input_tokens_seen": 51193656,
      "step": 88175
    },
    {
      "epoch": 13.133750372356271,
      "grad_norm": 0.033203125,
      "learning_rate": 0.009542448149303075,
      "loss": 0.8047,
      "num_input_tokens_seen": 51196568,
      "step": 88180
    },
    {
      "epoch": 13.13449508489723,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.009540632166448796,
      "loss": 0.7703,
      "num_input_tokens_seen": 51199160,
      "step": 88185
    },
    {
      "epoch": 13.135239797438189,
      "grad_norm": 0.041015625,
      "learning_rate": 0.009538816275824985,
      "loss": 0.8077,
      "num_input_tokens_seen": 51201944,
      "step": 88190
    },
    {
      "epoch": 13.135984509979147,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.009537000477462327,
      "loss": 0.8484,
      "num_input_tokens_seen": 51204728,
      "step": 88195
    },
    {
      "epoch": 13.136729222520108,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.009535184771391486,
      "loss": 0.8079,
      "num_input_tokens_seen": 51207640,
      "step": 88200
    },
    {
      "epoch": 13.137473935061067,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.00953336915764315,
      "loss": 0.7906,
      "num_input_tokens_seen": 51210360,
      "step": 88205
    },
    {
      "epoch": 13.138218647602026,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.009531553636247985,
      "loss": 0.8094,
      "num_input_tokens_seen": 51212920,
      "step": 88210
    },
    {
      "epoch": 13.138963360142984,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.009529738207236661,
      "loss": 0.8046,
      "num_input_tokens_seen": 51215736,
      "step": 88215
    },
    {
      "epoch": 13.139708072683945,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.009527922870639855,
      "loss": 0.7943,
      "num_input_tokens_seen": 51218872,
      "step": 88220
    },
    {
      "epoch": 13.140452785224904,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.009526107626488228,
      "loss": 0.8229,
      "num_input_tokens_seen": 51221848,
      "step": 88225
    },
    {
      "epoch": 13.141197497765862,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.00952429247481245,
      "loss": 0.8114,
      "num_input_tokens_seen": 51224632,
      "step": 88230
    },
    {
      "epoch": 13.141942210306821,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.009522477415643181,
      "loss": 0.8087,
      "num_input_tokens_seen": 51227608,
      "step": 88235
    },
    {
      "epoch": 13.142686922847782,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.009520662449011093,
      "loss": 0.8101,
      "num_input_tokens_seen": 51230232,
      "step": 88240
    },
    {
      "epoch": 13.14343163538874,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.009518847574946844,
      "loss": 0.8002,
      "num_input_tokens_seen": 51233656,
      "step": 88245
    },
    {
      "epoch": 13.1441763479297,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.009517032793481092,
      "loss": 0.8017,
      "num_input_tokens_seen": 51236696,
      "step": 88250
    },
    {
      "epoch": 13.144921060470658,
      "grad_norm": 0.048828125,
      "learning_rate": 0.009515218104644499,
      "loss": 0.8076,
      "num_input_tokens_seen": 51239768,
      "step": 88255
    },
    {
      "epoch": 13.145665773011617,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.009513403508467725,
      "loss": 0.8137,
      "num_input_tokens_seen": 51242584,
      "step": 88260
    },
    {
      "epoch": 13.146410485552577,
      "grad_norm": 0.037109375,
      "learning_rate": 0.009511589004981422,
      "loss": 0.8225,
      "num_input_tokens_seen": 51245688,
      "step": 88265
    },
    {
      "epoch": 13.147155198093536,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00950977459421624,
      "loss": 0.7997,
      "num_input_tokens_seen": 51248696,
      "step": 88270
    },
    {
      "epoch": 13.147899910634495,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.00950796027620284,
      "loss": 0.7879,
      "num_input_tokens_seen": 51251512,
      "step": 88275
    },
    {
      "epoch": 13.148644623175453,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.009506146050971869,
      "loss": 0.7898,
      "num_input_tokens_seen": 51254200,
      "step": 88280
    },
    {
      "epoch": 13.149389335716414,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.009504331918553977,
      "loss": 0.7853,
      "num_input_tokens_seen": 51257176,
      "step": 88285
    },
    {
      "epoch": 13.150134048257373,
      "grad_norm": 0.025390625,
      "learning_rate": 0.009502517878979806,
      "loss": 0.7945,
      "num_input_tokens_seen": 51260312,
      "step": 88290
    },
    {
      "epoch": 13.150878760798332,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.009500703932280015,
      "loss": 0.8131,
      "num_input_tokens_seen": 51263320,
      "step": 88295
    },
    {
      "epoch": 13.15162347333929,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.009498890078485242,
      "loss": 0.791,
      "num_input_tokens_seen": 51266200,
      "step": 88300
    },
    {
      "epoch": 13.15236818588025,
      "grad_norm": 0.0546875,
      "learning_rate": 0.009497076317626125,
      "loss": 0.7972,
      "num_input_tokens_seen": 51268984,
      "step": 88305
    },
    {
      "epoch": 13.15311289842121,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.009495262649733313,
      "loss": 0.8221,
      "num_input_tokens_seen": 51271800,
      "step": 88310
    },
    {
      "epoch": 13.153857610962168,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.009493449074837443,
      "loss": 0.8047,
      "num_input_tokens_seen": 51274712,
      "step": 88315
    },
    {
      "epoch": 13.154602323503127,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00949163559296916,
      "loss": 0.7791,
      "num_input_tokens_seen": 51278168,
      "step": 88320
    },
    {
      "epoch": 13.155347036044088,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.009489822204159086,
      "loss": 0.7882,
      "num_input_tokens_seen": 51280728,
      "step": 88325
    },
    {
      "epoch": 13.156091748585046,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.009488008908437871,
      "loss": 0.7865,
      "num_input_tokens_seen": 51283800,
      "step": 88330
    },
    {
      "epoch": 13.156836461126005,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.009486195705836139,
      "loss": 0.8006,
      "num_input_tokens_seen": 51286808,
      "step": 88335
    },
    {
      "epoch": 13.157581173666964,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.009484382596384532,
      "loss": 0.7968,
      "num_input_tokens_seen": 51289816,
      "step": 88340
    },
    {
      "epoch": 13.158325886207924,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.009482569580113676,
      "loss": 0.8124,
      "num_input_tokens_seen": 51292696,
      "step": 88345
    },
    {
      "epoch": 13.159070598748883,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.009480756657054194,
      "loss": 0.8109,
      "num_input_tokens_seen": 51295896,
      "step": 88350
    },
    {
      "epoch": 13.159815311289842,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.009478943827236724,
      "loss": 0.7831,
      "num_input_tokens_seen": 51298744,
      "step": 88355
    },
    {
      "epoch": 13.1605600238308,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00947713109069188,
      "loss": 0.7949,
      "num_input_tokens_seen": 51301816,
      "step": 88360
    },
    {
      "epoch": 13.161304736371761,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.009475318447450299,
      "loss": 0.8055,
      "num_input_tokens_seen": 51304664,
      "step": 88365
    },
    {
      "epoch": 13.16204944891272,
      "grad_norm": 0.022705078125,
      "learning_rate": 0.009473505897542592,
      "loss": 0.8031,
      "num_input_tokens_seen": 51307544,
      "step": 88370
    },
    {
      "epoch": 13.162794161453679,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.009471693440999393,
      "loss": 0.7867,
      "num_input_tokens_seen": 51310360,
      "step": 88375
    },
    {
      "epoch": 13.163538873994638,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.009469881077851313,
      "loss": 0.8134,
      "num_input_tokens_seen": 51313720,
      "step": 88380
    },
    {
      "epoch": 13.164283586535598,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.00946806880812897,
      "loss": 0.7904,
      "num_input_tokens_seen": 51316504,
      "step": 88385
    },
    {
      "epoch": 13.165028299076557,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.009466256631862983,
      "loss": 0.8008,
      "num_input_tokens_seen": 51319672,
      "step": 88390
    },
    {
      "epoch": 13.165773011617516,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.009464444549083961,
      "loss": 0.7944,
      "num_input_tokens_seen": 51322616,
      "step": 88395
    },
    {
      "epoch": 13.166517724158474,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.009462632559822529,
      "loss": 0.7962,
      "num_input_tokens_seen": 51325656,
      "step": 88400
    },
    {
      "epoch": 13.167262436699435,
      "grad_norm": 0.041015625,
      "learning_rate": 0.009460820664109287,
      "loss": 0.7892,
      "num_input_tokens_seen": 51328472,
      "step": 88405
    },
    {
      "epoch": 13.168007149240394,
      "grad_norm": 0.056640625,
      "learning_rate": 0.009459008861974852,
      "loss": 0.7984,
      "num_input_tokens_seen": 51331384,
      "step": 88410
    },
    {
      "epoch": 13.168751861781352,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.00945719715344983,
      "loss": 0.7684,
      "num_input_tokens_seen": 51334264,
      "step": 88415
    },
    {
      "epoch": 13.169496574322311,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.009455385538564832,
      "loss": 0.795,
      "num_input_tokens_seen": 51337176,
      "step": 88420
    },
    {
      "epoch": 13.17024128686327,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.009453574017350459,
      "loss": 0.7901,
      "num_input_tokens_seen": 51339928,
      "step": 88425
    },
    {
      "epoch": 13.17098599940423,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.009451762589837312,
      "loss": 0.7892,
      "num_input_tokens_seen": 51342648,
      "step": 88430
    },
    {
      "epoch": 13.17173071194519,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.009449951256056003,
      "loss": 0.8203,
      "num_input_tokens_seen": 51345208,
      "step": 88435
    },
    {
      "epoch": 13.172475424486148,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.009448140016037119,
      "loss": 0.781,
      "num_input_tokens_seen": 51348184,
      "step": 88440
    },
    {
      "epoch": 13.173220137027107,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.009446328869811272,
      "loss": 0.7898,
      "num_input_tokens_seen": 51350808,
      "step": 88445
    },
    {
      "epoch": 13.173964849568067,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.009444517817409053,
      "loss": 0.7982,
      "num_input_tokens_seen": 51353464,
      "step": 88450
    },
    {
      "epoch": 13.174709562109026,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.00944270685886106,
      "loss": 0.7988,
      "num_input_tokens_seen": 51356600,
      "step": 88455
    },
    {
      "epoch": 13.175454274649985,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.009440895994197888,
      "loss": 0.7939,
      "num_input_tokens_seen": 51359352,
      "step": 88460
    },
    {
      "epoch": 13.176198987190944,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.009439085223450123,
      "loss": 0.8107,
      "num_input_tokens_seen": 51362136,
      "step": 88465
    },
    {
      "epoch": 13.176943699731904,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.009437274546648366,
      "loss": 0.7876,
      "num_input_tokens_seen": 51365112,
      "step": 88470
    },
    {
      "epoch": 13.177688412272863,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.009435463963823196,
      "loss": 0.8166,
      "num_input_tokens_seen": 51367960,
      "step": 88475
    },
    {
      "epoch": 13.178433124813822,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.009433653475005214,
      "loss": 0.7823,
      "num_input_tokens_seen": 51370744,
      "step": 88480
    },
    {
      "epoch": 13.17917783735478,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.009431843080224996,
      "loss": 0.823,
      "num_input_tokens_seen": 51373560,
      "step": 88485
    },
    {
      "epoch": 13.17992254989574,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.00943003277951313,
      "loss": 0.788,
      "num_input_tokens_seen": 51376728,
      "step": 88490
    },
    {
      "epoch": 13.1806672624367,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.009428222572900198,
      "loss": 0.8286,
      "num_input_tokens_seen": 51379576,
      "step": 88495
    },
    {
      "epoch": 13.181411974977658,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.009426412460416785,
      "loss": 0.7861,
      "num_input_tokens_seen": 51382488,
      "step": 88500
    },
    {
      "epoch": 13.182156687518617,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.009424602442093468,
      "loss": 0.7818,
      "num_input_tokens_seen": 51385176,
      "step": 88505
    },
    {
      "epoch": 13.182901400059578,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.009422792517960825,
      "loss": 0.7946,
      "num_input_tokens_seen": 51387864,
      "step": 88510
    },
    {
      "epoch": 13.183646112600536,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.009420982688049435,
      "loss": 0.7987,
      "num_input_tokens_seen": 51390584,
      "step": 88515
    },
    {
      "epoch": 13.184390825141495,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.009419172952389872,
      "loss": 0.7932,
      "num_input_tokens_seen": 51393368,
      "step": 88520
    },
    {
      "epoch": 13.185135537682454,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.009417363311012712,
      "loss": 0.8022,
      "num_input_tokens_seen": 51396216,
      "step": 88525
    },
    {
      "epoch": 13.185880250223414,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.00941555376394852,
      "loss": 0.8173,
      "num_input_tokens_seen": 51399288,
      "step": 88530
    },
    {
      "epoch": 13.186624962764373,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.009413744311227875,
      "loss": 0.7913,
      "num_input_tokens_seen": 51402264,
      "step": 88535
    },
    {
      "epoch": 13.187369675305332,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.009411934952881344,
      "loss": 0.8099,
      "num_input_tokens_seen": 51405048,
      "step": 88540
    },
    {
      "epoch": 13.18811438784629,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.009410125688939485,
      "loss": 0.7691,
      "num_input_tokens_seen": 51407800,
      "step": 88545
    },
    {
      "epoch": 13.188859100387251,
      "grad_norm": 0.041015625,
      "learning_rate": 0.00940831651943288,
      "loss": 0.7804,
      "num_input_tokens_seen": 51410424,
      "step": 88550
    },
    {
      "epoch": 13.18960381292821,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.009406507444392078,
      "loss": 0.801,
      "num_input_tokens_seen": 51413464,
      "step": 88555
    },
    {
      "epoch": 13.190348525469169,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.009404698463847653,
      "loss": 0.8079,
      "num_input_tokens_seen": 51416568,
      "step": 88560
    },
    {
      "epoch": 13.191093238010128,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.009402889577830155,
      "loss": 0.7982,
      "num_input_tokens_seen": 51419512,
      "step": 88565
    },
    {
      "epoch": 13.191837950551088,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.009401080786370154,
      "loss": 0.8028,
      "num_input_tokens_seen": 51422424,
      "step": 88570
    },
    {
      "epoch": 13.192582663092047,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.009399272089498199,
      "loss": 0.8001,
      "num_input_tokens_seen": 51425240,
      "step": 88575
    },
    {
      "epoch": 13.193327375633006,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.009397463487244849,
      "loss": 0.8043,
      "num_input_tokens_seen": 51427832,
      "step": 88580
    },
    {
      "epoch": 13.194072088173964,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.009395654979640665,
      "loss": 0.8076,
      "num_input_tokens_seen": 51430552,
      "step": 88585
    },
    {
      "epoch": 13.194816800714925,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.009393846566716189,
      "loss": 0.799,
      "num_input_tokens_seen": 51433496,
      "step": 88590
    },
    {
      "epoch": 13.195561513255884,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.00939203824850198,
      "loss": 0.7939,
      "num_input_tokens_seen": 51436248,
      "step": 88595
    },
    {
      "epoch": 13.196306225796842,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.00939023002502858,
      "loss": 0.7913,
      "num_input_tokens_seen": 51439224,
      "step": 88600
    },
    {
      "epoch": 13.197050938337801,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.009388421896326549,
      "loss": 0.7982,
      "num_input_tokens_seen": 51442168,
      "step": 88605
    },
    {
      "epoch": 13.19779565087876,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.009386613862426422,
      "loss": 0.782,
      "num_input_tokens_seen": 51445048,
      "step": 88610
    },
    {
      "epoch": 13.19854036341972,
      "grad_norm": 0.0546875,
      "learning_rate": 0.009384805923358754,
      "loss": 0.7923,
      "num_input_tokens_seen": 51447896,
      "step": 88615
    },
    {
      "epoch": 13.19928507596068,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.009382998079154077,
      "loss": 0.7887,
      "num_input_tokens_seen": 51450904,
      "step": 88620
    },
    {
      "epoch": 13.200029788501638,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.009381190329842942,
      "loss": 0.7844,
      "num_input_tokens_seen": 51453784,
      "step": 88625
    },
    {
      "epoch": 13.200774501042597,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.009379382675455886,
      "loss": 0.8112,
      "num_input_tokens_seen": 51456600,
      "step": 88630
    },
    {
      "epoch": 13.201519213583557,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00937757511602344,
      "loss": 0.79,
      "num_input_tokens_seen": 51459352,
      "step": 88635
    },
    {
      "epoch": 13.202263926124516,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.00937576765157616,
      "loss": 0.7902,
      "num_input_tokens_seen": 51462424,
      "step": 88640
    },
    {
      "epoch": 13.203008638665475,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.00937396028214456,
      "loss": 0.7989,
      "num_input_tokens_seen": 51464984,
      "step": 88645
    },
    {
      "epoch": 13.203753351206434,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.009372153007759189,
      "loss": 0.8012,
      "num_input_tokens_seen": 51467896,
      "step": 88650
    },
    {
      "epoch": 13.204498063747394,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.009370345828450567,
      "loss": 0.8009,
      "num_input_tokens_seen": 51470648,
      "step": 88655
    },
    {
      "epoch": 13.205242776288353,
      "grad_norm": 0.0390625,
      "learning_rate": 0.009368538744249234,
      "loss": 0.8669,
      "num_input_tokens_seen": 51473976,
      "step": 88660
    },
    {
      "epoch": 13.205987488829312,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.009366731755185717,
      "loss": 0.8084,
      "num_input_tokens_seen": 51476920,
      "step": 88665
    },
    {
      "epoch": 13.20673220137027,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.009364924861290536,
      "loss": 0.8066,
      "num_input_tokens_seen": 51479736,
      "step": 88670
    },
    {
      "epoch": 13.207476913911231,
      "grad_norm": 0.0185546875,
      "learning_rate": 0.009363118062594229,
      "loss": 0.7878,
      "num_input_tokens_seen": 51482392,
      "step": 88675
    },
    {
      "epoch": 13.20822162645219,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.009361311359127308,
      "loss": 0.7755,
      "num_input_tokens_seen": 51485208,
      "step": 88680
    },
    {
      "epoch": 13.208966338993148,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.009359504750920304,
      "loss": 0.8095,
      "num_input_tokens_seen": 51488120,
      "step": 88685
    },
    {
      "epoch": 13.209711051534107,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.00935769823800373,
      "loss": 0.8092,
      "num_input_tokens_seen": 51490968,
      "step": 88690
    },
    {
      "epoch": 13.210455764075068,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.009355891820408116,
      "loss": 0.8027,
      "num_input_tokens_seen": 51493848,
      "step": 88695
    },
    {
      "epoch": 13.211200476616026,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.009354085498163972,
      "loss": 0.8074,
      "num_input_tokens_seen": 51496888,
      "step": 88700
    },
    {
      "epoch": 13.211945189156985,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.009352279271301809,
      "loss": 0.7875,
      "num_input_tokens_seen": 51499864,
      "step": 88705
    },
    {
      "epoch": 13.212689901697944,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.009350473139852156,
      "loss": 0.7815,
      "num_input_tokens_seen": 51502872,
      "step": 88710
    },
    {
      "epoch": 13.213434614238905,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.009348667103845512,
      "loss": 0.807,
      "num_input_tokens_seen": 51505528,
      "step": 88715
    },
    {
      "epoch": 13.214179326779863,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.009346861163312396,
      "loss": 0.7974,
      "num_input_tokens_seen": 51508440,
      "step": 88720
    },
    {
      "epoch": 13.214924039320822,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.00934505531828331,
      "loss": 0.7831,
      "num_input_tokens_seen": 51511384,
      "step": 88725
    },
    {
      "epoch": 13.21566875186178,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.009343249568788774,
      "loss": 0.8216,
      "num_input_tokens_seen": 51514008,
      "step": 88730
    },
    {
      "epoch": 13.216413464402741,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.009341443914859282,
      "loss": 0.803,
      "num_input_tokens_seen": 51516920,
      "step": 88735
    },
    {
      "epoch": 13.2171581769437,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.009339638356525347,
      "loss": 0.7847,
      "num_input_tokens_seen": 51520504,
      "step": 88740
    },
    {
      "epoch": 13.217902889484659,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.009337832893817472,
      "loss": 0.8153,
      "num_input_tokens_seen": 51523512,
      "step": 88745
    },
    {
      "epoch": 13.218647602025618,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.00933602752676615,
      "loss": 0.8086,
      "num_input_tokens_seen": 51526264,
      "step": 88750
    },
    {
      "epoch": 13.219392314566578,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.00933422225540189,
      "loss": 0.8225,
      "num_input_tokens_seen": 51529528,
      "step": 88755
    },
    {
      "epoch": 13.220137027107537,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.00933241707975518,
      "loss": 0.8265,
      "num_input_tokens_seen": 51532056,
      "step": 88760
    },
    {
      "epoch": 13.220881739648496,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.00933061199985653,
      "loss": 0.7902,
      "num_input_tokens_seen": 51534776,
      "step": 88765
    },
    {
      "epoch": 13.221626452189454,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.009328807015736424,
      "loss": 0.8013,
      "num_input_tokens_seen": 51537592,
      "step": 88770
    },
    {
      "epoch": 13.222371164730415,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.009327002127425365,
      "loss": 0.7894,
      "num_input_tokens_seen": 51540248,
      "step": 88775
    },
    {
      "epoch": 13.223115877271374,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.009325197334953836,
      "loss": 0.7918,
      "num_input_tokens_seen": 51542968,
      "step": 88780
    },
    {
      "epoch": 13.223860589812332,
      "grad_norm": 0.0703125,
      "learning_rate": 0.009323392638352329,
      "loss": 0.7905,
      "num_input_tokens_seen": 51545688,
      "step": 88785
    },
    {
      "epoch": 13.224605302353291,
      "grad_norm": 0.13671875,
      "learning_rate": 0.009321588037651337,
      "loss": 0.8362,
      "num_input_tokens_seen": 51548568,
      "step": 88790
    },
    {
      "epoch": 13.22535001489425,
      "grad_norm": 0.03515625,
      "learning_rate": 0.00931978353288134,
      "loss": 0.7918,
      "num_input_tokens_seen": 51551288,
      "step": 88795
    },
    {
      "epoch": 13.22609472743521,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.009317979124072829,
      "loss": 0.8091,
      "num_input_tokens_seen": 51554200,
      "step": 88800
    },
    {
      "epoch": 13.22683943997617,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.009316174811256283,
      "loss": 0.8083,
      "num_input_tokens_seen": 51557048,
      "step": 88805
    },
    {
      "epoch": 13.227584152517128,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.00931437059446219,
      "loss": 0.8028,
      "num_input_tokens_seen": 51559960,
      "step": 88810
    },
    {
      "epoch": 13.228328865058087,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.009312566473721027,
      "loss": 0.7789,
      "num_input_tokens_seen": 51562840,
      "step": 88815
    },
    {
      "epoch": 13.229073577599047,
      "grad_norm": 0.041015625,
      "learning_rate": 0.009310762449063275,
      "loss": 0.798,
      "num_input_tokens_seen": 51565976,
      "step": 88820
    },
    {
      "epoch": 13.229818290140006,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.00930895852051941,
      "loss": 0.8066,
      "num_input_tokens_seen": 51568760,
      "step": 88825
    },
    {
      "epoch": 13.230563002680965,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.009307154688119901,
      "loss": 0.773,
      "num_input_tokens_seen": 51571608,
      "step": 88830
    },
    {
      "epoch": 13.231307715221924,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.009305350951895231,
      "loss": 0.7854,
      "num_input_tokens_seen": 51574520,
      "step": 88835
    },
    {
      "epoch": 13.232052427762884,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.009303547311875864,
      "loss": 0.8061,
      "num_input_tokens_seen": 51577496,
      "step": 88840
    },
    {
      "epoch": 13.232797140303843,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.009301743768092284,
      "loss": 0.8069,
      "num_input_tokens_seen": 51580344,
      "step": 88845
    },
    {
      "epoch": 13.233541852844802,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.009299940320574945,
      "loss": 0.7716,
      "num_input_tokens_seen": 51583352,
      "step": 88850
    },
    {
      "epoch": 13.23428656538576,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.009298136969354323,
      "loss": 0.7796,
      "num_input_tokens_seen": 51586008,
      "step": 88855
    },
    {
      "epoch": 13.235031277926721,
      "grad_norm": 0.03515625,
      "learning_rate": 0.00929633371446088,
      "loss": 0.7979,
      "num_input_tokens_seen": 51588760,
      "step": 88860
    },
    {
      "epoch": 13.23577599046768,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.009294530555925086,
      "loss": 0.7932,
      "num_input_tokens_seen": 51591416,
      "step": 88865
    },
    {
      "epoch": 13.236520703008638,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.009292727493777398,
      "loss": 0.7938,
      "num_input_tokens_seen": 51594552,
      "step": 88870
    },
    {
      "epoch": 13.237265415549597,
      "grad_norm": 0.037109375,
      "learning_rate": 0.009290924528048274,
      "loss": 0.7997,
      "num_input_tokens_seen": 51597528,
      "step": 88875
    },
    {
      "epoch": 13.238010128090558,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.009289121658768185,
      "loss": 0.8022,
      "num_input_tokens_seen": 51600472,
      "step": 88880
    },
    {
      "epoch": 13.238754840631517,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.009287318885967578,
      "loss": 0.7868,
      "num_input_tokens_seen": 51603224,
      "step": 88885
    },
    {
      "epoch": 13.239499553172475,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.009285516209676914,
      "loss": 0.8396,
      "num_input_tokens_seen": 51606200,
      "step": 88890
    },
    {
      "epoch": 13.240244265713434,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.009283713629926638,
      "loss": 0.7788,
      "num_input_tokens_seen": 51609112,
      "step": 88895
    },
    {
      "epoch": 13.240988978254395,
      "grad_norm": 0.0390625,
      "learning_rate": 0.009281911146747219,
      "loss": 0.7941,
      "num_input_tokens_seen": 51612248,
      "step": 88900
    },
    {
      "epoch": 13.241733690795353,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0092801087601691,
      "loss": 0.7896,
      "num_input_tokens_seen": 51614840,
      "step": 88905
    },
    {
      "epoch": 13.242478403336312,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.009278306470222724,
      "loss": 0.7831,
      "num_input_tokens_seen": 51617592,
      "step": 88910
    },
    {
      "epoch": 13.24322311587727,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.009276504276938546,
      "loss": 0.7731,
      "num_input_tokens_seen": 51620600,
      "step": 88915
    },
    {
      "epoch": 13.243967828418231,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.009274702180347014,
      "loss": 0.8044,
      "num_input_tokens_seen": 51623768,
      "step": 88920
    },
    {
      "epoch": 13.24471254095919,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.009272900180478572,
      "loss": 0.781,
      "num_input_tokens_seen": 51626808,
      "step": 88925
    },
    {
      "epoch": 13.245457253500149,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.009271098277363651,
      "loss": 0.8014,
      "num_input_tokens_seen": 51629880,
      "step": 88930
    },
    {
      "epoch": 13.246201966041108,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.009269296471032711,
      "loss": 0.8151,
      "num_input_tokens_seen": 51632696,
      "step": 88935
    },
    {
      "epoch": 13.246946678582066,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.00926749476151618,
      "loss": 0.8112,
      "num_input_tokens_seen": 51635288,
      "step": 88940
    },
    {
      "epoch": 13.247691391123027,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0092656931488445,
      "loss": 0.8464,
      "num_input_tokens_seen": 51638008,
      "step": 88945
    },
    {
      "epoch": 13.248436103663986,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.009263891633048105,
      "loss": 0.8117,
      "num_input_tokens_seen": 51640824,
      "step": 88950
    },
    {
      "epoch": 13.249180816204944,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00926209021415743,
      "loss": 0.8122,
      "num_input_tokens_seen": 51643640,
      "step": 88955
    },
    {
      "epoch": 13.249925528745903,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.009260288892202912,
      "loss": 0.8002,
      "num_input_tokens_seen": 51646296,
      "step": 88960
    },
    {
      "epoch": 13.250670241286864,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.009258487667214974,
      "loss": 0.7882,
      "num_input_tokens_seen": 51649016,
      "step": 88965
    },
    {
      "epoch": 13.251414953827823,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.009256686539224056,
      "loss": 0.7799,
      "num_input_tokens_seen": 51652216,
      "step": 88970
    },
    {
      "epoch": 13.252159666368781,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.00925488550826058,
      "loss": 0.7896,
      "num_input_tokens_seen": 51654904,
      "step": 88975
    },
    {
      "epoch": 13.25290437890974,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.009253084574354978,
      "loss": 0.8123,
      "num_input_tokens_seen": 51657592,
      "step": 88980
    },
    {
      "epoch": 13.2536490914507,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.00925128373753767,
      "loss": 0.8447,
      "num_input_tokens_seen": 51660376,
      "step": 88985
    },
    {
      "epoch": 13.25439380399166,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.009249482997839075,
      "loss": 0.8005,
      "num_input_tokens_seen": 51663480,
      "step": 88990
    },
    {
      "epoch": 13.255138516532618,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.009247682355289626,
      "loss": 0.7902,
      "num_input_tokens_seen": 51666264,
      "step": 88995
    },
    {
      "epoch": 13.255883229073577,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.009245881809919731,
      "loss": 0.7917,
      "num_input_tokens_seen": 51668984,
      "step": 89000
    },
    {
      "epoch": 13.256627941614537,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.00924408136175982,
      "loss": 0.7916,
      "num_input_tokens_seen": 51671832,
      "step": 89005
    },
    {
      "epoch": 13.257372654155496,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.009242281010840304,
      "loss": 0.7804,
      "num_input_tokens_seen": 51674744,
      "step": 89010
    },
    {
      "epoch": 13.258117366696455,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.009240480757191598,
      "loss": 0.8202,
      "num_input_tokens_seen": 51677624,
      "step": 89015
    },
    {
      "epoch": 13.258862079237414,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.009238680600844113,
      "loss": 0.791,
      "num_input_tokens_seen": 51680664,
      "step": 89020
    },
    {
      "epoch": 13.259606791778374,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.009236880541828267,
      "loss": 0.7969,
      "num_input_tokens_seen": 51683512,
      "step": 89025
    },
    {
      "epoch": 13.260351504319333,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.009235080580174469,
      "loss": 0.7968,
      "num_input_tokens_seen": 51686680,
      "step": 89030
    },
    {
      "epoch": 13.261096216860292,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.009233280715913119,
      "loss": 0.8106,
      "num_input_tokens_seen": 51689592,
      "step": 89035
    },
    {
      "epoch": 13.26184092940125,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.009231480949074635,
      "loss": 0.8049,
      "num_input_tokens_seen": 51692696,
      "step": 89040
    },
    {
      "epoch": 13.262585641942211,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.009229681279689416,
      "loss": 0.7872,
      "num_input_tokens_seen": 51695480,
      "step": 89045
    },
    {
      "epoch": 13.26333035448317,
      "grad_norm": 0.03125,
      "learning_rate": 0.009227881707787869,
      "loss": 0.8134,
      "num_input_tokens_seen": 51698328,
      "step": 89050
    },
    {
      "epoch": 13.264075067024129,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.00922608223340039,
      "loss": 0.7919,
      "num_input_tokens_seen": 51701208,
      "step": 89055
    },
    {
      "epoch": 13.264819779565087,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.009224282856557386,
      "loss": 0.8068,
      "num_input_tokens_seen": 51704056,
      "step": 89060
    },
    {
      "epoch": 13.265564492106048,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.009222483577289255,
      "loss": 0.7796,
      "num_input_tokens_seen": 51706840,
      "step": 89065
    },
    {
      "epoch": 13.266309204647007,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.009220684395626384,
      "loss": 0.7792,
      "num_input_tokens_seen": 51709848,
      "step": 89070
    },
    {
      "epoch": 13.267053917187965,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.009218885311599184,
      "loss": 0.7975,
      "num_input_tokens_seen": 51712888,
      "step": 89075
    },
    {
      "epoch": 13.267798629728924,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.009217086325238039,
      "loss": 0.8187,
      "num_input_tokens_seen": 51715832,
      "step": 89080
    },
    {
      "epoch": 13.268543342269885,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.009215287436573344,
      "loss": 0.8119,
      "num_input_tokens_seen": 51718776,
      "step": 89085
    },
    {
      "epoch": 13.269288054810843,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.009213488645635483,
      "loss": 0.7905,
      "num_input_tokens_seen": 51721656,
      "step": 89090
    },
    {
      "epoch": 13.270032767351802,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.009211689952454857,
      "loss": 0.7942,
      "num_input_tokens_seen": 51724248,
      "step": 89095
    },
    {
      "epoch": 13.270777479892761,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.009209891357061838,
      "loss": 0.7922,
      "num_input_tokens_seen": 51727128,
      "step": 89100
    },
    {
      "epoch": 13.271522192433721,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.00920809285948683,
      "loss": 0.782,
      "num_input_tokens_seen": 51730104,
      "step": 89105
    },
    {
      "epoch": 13.27226690497468,
      "grad_norm": 0.05078125,
      "learning_rate": 0.009206294459760204,
      "loss": 0.8047,
      "num_input_tokens_seen": 51732856,
      "step": 89110
    },
    {
      "epoch": 13.273011617515639,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.00920449615791234,
      "loss": 0.7864,
      "num_input_tokens_seen": 51735896,
      "step": 89115
    },
    {
      "epoch": 13.273756330056598,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.009202697953973629,
      "loss": 0.8157,
      "num_input_tokens_seen": 51738872,
      "step": 89120
    },
    {
      "epoch": 13.274501042597556,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.009200899847974436,
      "loss": 0.8247,
      "num_input_tokens_seen": 51741528,
      "step": 89125
    },
    {
      "epoch": 13.275245755138517,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.009199101839945153,
      "loss": 0.7957,
      "num_input_tokens_seen": 51744568,
      "step": 89130
    },
    {
      "epoch": 13.275990467679476,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.009197303929916145,
      "loss": 0.7931,
      "num_input_tokens_seen": 51747608,
      "step": 89135
    },
    {
      "epoch": 13.276735180220435,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.009195506117917791,
      "loss": 0.7877,
      "num_input_tokens_seen": 51750840,
      "step": 89140
    },
    {
      "epoch": 13.277479892761393,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.009193708403980464,
      "loss": 0.7743,
      "num_input_tokens_seen": 51754040,
      "step": 89145
    },
    {
      "epoch": 13.278224605302354,
      "grad_norm": 0.03515625,
      "learning_rate": 0.009191910788134527,
      "loss": 0.7773,
      "num_input_tokens_seen": 51756856,
      "step": 89150
    },
    {
      "epoch": 13.278969317843313,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.00919011327041036,
      "loss": 0.7983,
      "num_input_tokens_seen": 51759992,
      "step": 89155
    },
    {
      "epoch": 13.279714030384271,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.009188315850838319,
      "loss": 0.7933,
      "num_input_tokens_seen": 51762680,
      "step": 89160
    },
    {
      "epoch": 13.28045874292523,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.00918651852944878,
      "loss": 0.8084,
      "num_input_tokens_seen": 51765624,
      "step": 89165
    },
    {
      "epoch": 13.28120345546619,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.009184721306272096,
      "loss": 0.7847,
      "num_input_tokens_seen": 51768792,
      "step": 89170
    },
    {
      "epoch": 13.28194816800715,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.009182924181338639,
      "loss": 0.7849,
      "num_input_tokens_seen": 51771480,
      "step": 89175
    },
    {
      "epoch": 13.282692880548108,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.009181127154678764,
      "loss": 0.7885,
      "num_input_tokens_seen": 51774552,
      "step": 89180
    },
    {
      "epoch": 13.283437593089067,
      "grad_norm": 0.037109375,
      "learning_rate": 0.009179330226322837,
      "loss": 0.7912,
      "num_input_tokens_seen": 51777528,
      "step": 89185
    },
    {
      "epoch": 13.284182305630027,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.009177533396301205,
      "loss": 0.823,
      "num_input_tokens_seen": 51780504,
      "step": 89190
    },
    {
      "epoch": 13.284927018170986,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.009175736664644227,
      "loss": 0.7876,
      "num_input_tokens_seen": 51783416,
      "step": 89195
    },
    {
      "epoch": 13.285671730711945,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.009173940031382263,
      "loss": 0.7907,
      "num_input_tokens_seen": 51786360,
      "step": 89200
    },
    {
      "epoch": 13.286416443252904,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.009172143496545654,
      "loss": 0.8039,
      "num_input_tokens_seen": 51789048,
      "step": 89205
    },
    {
      "epoch": 13.287161155793864,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.009170347060164766,
      "loss": 0.7849,
      "num_input_tokens_seen": 51791992,
      "step": 89210
    },
    {
      "epoch": 13.287905868334823,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.009168550722269934,
      "loss": 0.7637,
      "num_input_tokens_seen": 51794968,
      "step": 89215
    },
    {
      "epoch": 13.288650580875782,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.009166754482891516,
      "loss": 0.7834,
      "num_input_tokens_seen": 51797720,
      "step": 89220
    },
    {
      "epoch": 13.28939529341674,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.00916495834205985,
      "loss": 0.792,
      "num_input_tokens_seen": 51800472,
      "step": 89225
    },
    {
      "epoch": 13.290140005957701,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.009163162299805279,
      "loss": 0.802,
      "num_input_tokens_seen": 51803448,
      "step": 89230
    },
    {
      "epoch": 13.29088471849866,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.00916136635615815,
      "loss": 0.7985,
      "num_input_tokens_seen": 51806616,
      "step": 89235
    },
    {
      "epoch": 13.291629431039619,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0091595705111488,
      "loss": 0.8068,
      "num_input_tokens_seen": 51809432,
      "step": 89240
    },
    {
      "epoch": 13.292374143580577,
      "grad_norm": 0.01953125,
      "learning_rate": 0.009157774764807577,
      "loss": 0.807,
      "num_input_tokens_seen": 51812376,
      "step": 89245
    },
    {
      "epoch": 13.293118856121538,
      "grad_norm": 0.0390625,
      "learning_rate": 0.009155979117164805,
      "loss": 0.7904,
      "num_input_tokens_seen": 51815256,
      "step": 89250
    },
    {
      "epoch": 13.293863568662497,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.00915418356825083,
      "loss": 0.7773,
      "num_input_tokens_seen": 51817912,
      "step": 89255
    },
    {
      "epoch": 13.294608281203455,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.00915238811809598,
      "loss": 0.7849,
      "num_input_tokens_seen": 51820568,
      "step": 89260
    },
    {
      "epoch": 13.295352993744414,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.009150592766730591,
      "loss": 0.8084,
      "num_input_tokens_seen": 51823576,
      "step": 89265
    },
    {
      "epoch": 13.296097706285375,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.009148797514184997,
      "loss": 0.7808,
      "num_input_tokens_seen": 51826552,
      "step": 89270
    },
    {
      "epoch": 13.296842418826333,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.009147002360489515,
      "loss": 0.8069,
      "num_input_tokens_seen": 51829304,
      "step": 89275
    },
    {
      "epoch": 13.297587131367292,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00914520730567448,
      "loss": 0.8222,
      "num_input_tokens_seen": 51832184,
      "step": 89280
    },
    {
      "epoch": 13.298331843908251,
      "grad_norm": 0.041015625,
      "learning_rate": 0.009143412349770218,
      "loss": 0.7994,
      "num_input_tokens_seen": 51834744,
      "step": 89285
    },
    {
      "epoch": 13.299076556449211,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.009141617492807053,
      "loss": 0.7965,
      "num_input_tokens_seen": 51837720,
      "step": 89290
    },
    {
      "epoch": 13.29982126899017,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.009139822734815303,
      "loss": 0.8106,
      "num_input_tokens_seen": 51840664,
      "step": 89295
    },
    {
      "epoch": 13.300565981531129,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.009138028075825297,
      "loss": 0.7849,
      "num_input_tokens_seen": 51843448,
      "step": 89300
    },
    {
      "epoch": 13.301310694072088,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.00913623351586735,
      "loss": 0.7828,
      "num_input_tokens_seen": 51846584,
      "step": 89305
    },
    {
      "epoch": 13.302055406613047,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.009134439054971771,
      "loss": 0.8181,
      "num_input_tokens_seen": 51849624,
      "step": 89310
    },
    {
      "epoch": 13.302800119154007,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00913264469316889,
      "loss": 0.7877,
      "num_input_tokens_seen": 51852632,
      "step": 89315
    },
    {
      "epoch": 13.303544831694966,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.009130850430489005,
      "loss": 0.7823,
      "num_input_tokens_seen": 51855672,
      "step": 89320
    },
    {
      "epoch": 13.304289544235925,
      "grad_norm": 0.033203125,
      "learning_rate": 0.009129056266962443,
      "loss": 0.7845,
      "num_input_tokens_seen": 51858456,
      "step": 89325
    },
    {
      "epoch": 13.305034256776883,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.009127262202619503,
      "loss": 0.787,
      "num_input_tokens_seen": 51861176,
      "step": 89330
    },
    {
      "epoch": 13.305778969317844,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.009125468237490506,
      "loss": 0.7907,
      "num_input_tokens_seen": 51863832,
      "step": 89335
    },
    {
      "epoch": 13.306523681858803,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.009123674371605748,
      "loss": 0.7926,
      "num_input_tokens_seen": 51866872,
      "step": 89340
    },
    {
      "epoch": 13.307268394399761,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.009121880604995544,
      "loss": 0.794,
      "num_input_tokens_seen": 51869976,
      "step": 89345
    },
    {
      "epoch": 13.30801310694072,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00912008693769019,
      "loss": 0.8117,
      "num_input_tokens_seen": 51872856,
      "step": 89350
    },
    {
      "epoch": 13.30875781948168,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.009118293369719989,
      "loss": 0.7885,
      "num_input_tokens_seen": 51875832,
      "step": 89355
    },
    {
      "epoch": 13.30950253202264,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.009116499901115249,
      "loss": 0.7805,
      "num_input_tokens_seen": 51878680,
      "step": 89360
    },
    {
      "epoch": 13.310247244563598,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.009114706531906255,
      "loss": 0.7832,
      "num_input_tokens_seen": 51881752,
      "step": 89365
    },
    {
      "epoch": 13.310991957104557,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00911291326212332,
      "loss": 0.8092,
      "num_input_tokens_seen": 51884760,
      "step": 89370
    },
    {
      "epoch": 13.311736669645517,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.009111120091796729,
      "loss": 0.7771,
      "num_input_tokens_seen": 51887544,
      "step": 89375
    },
    {
      "epoch": 13.312481382186476,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00910932702095678,
      "loss": 0.7807,
      "num_input_tokens_seen": 51890424,
      "step": 89380
    },
    {
      "epoch": 13.313226094727435,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.00910753404963376,
      "loss": 0.8088,
      "num_input_tokens_seen": 51893464,
      "step": 89385
    },
    {
      "epoch": 13.313970807268394,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.009105741177857968,
      "loss": 0.8035,
      "num_input_tokens_seen": 51896600,
      "step": 89390
    },
    {
      "epoch": 13.314715519809354,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.009103948405659689,
      "loss": 0.8088,
      "num_input_tokens_seen": 51899384,
      "step": 89395
    },
    {
      "epoch": 13.315460232350313,
      "grad_norm": 0.033203125,
      "learning_rate": 0.009102155733069205,
      "loss": 0.7996,
      "num_input_tokens_seen": 51902136,
      "step": 89400
    },
    {
      "epoch": 13.316204944891272,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.009100363160116809,
      "loss": 0.7979,
      "num_input_tokens_seen": 51905016,
      "step": 89405
    },
    {
      "epoch": 13.31694965743223,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00909857068683278,
      "loss": 0.8443,
      "num_input_tokens_seen": 51907576,
      "step": 89410
    },
    {
      "epoch": 13.317694369973191,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.009096778313247404,
      "loss": 0.7975,
      "num_input_tokens_seen": 51910264,
      "step": 89415
    },
    {
      "epoch": 13.31843908251415,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.009094986039390952,
      "loss": 0.7974,
      "num_input_tokens_seen": 51913208,
      "step": 89420
    },
    {
      "epoch": 13.319183795055109,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.009093193865293716,
      "loss": 0.807,
      "num_input_tokens_seen": 51916120,
      "step": 89425
    },
    {
      "epoch": 13.319928507596067,
      "grad_norm": 0.03515625,
      "learning_rate": 0.009091401790985965,
      "loss": 0.8121,
      "num_input_tokens_seen": 51919704,
      "step": 89430
    },
    {
      "epoch": 13.320673220137028,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00908960981649797,
      "loss": 0.8071,
      "num_input_tokens_seen": 51922456,
      "step": 89435
    },
    {
      "epoch": 13.321417932677987,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.009087817941860017,
      "loss": 0.7697,
      "num_input_tokens_seen": 51925496,
      "step": 89440
    },
    {
      "epoch": 13.322162645218945,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.009086026167102367,
      "loss": 0.8002,
      "num_input_tokens_seen": 51928216,
      "step": 89445
    },
    {
      "epoch": 13.322907357759904,
      "grad_norm": 0.03125,
      "learning_rate": 0.009084234492255297,
      "loss": 0.7907,
      "num_input_tokens_seen": 51931160,
      "step": 89450
    },
    {
      "epoch": 13.323652070300863,
      "grad_norm": 0.056640625,
      "learning_rate": 0.009082442917349068,
      "loss": 0.8314,
      "num_input_tokens_seen": 51934040,
      "step": 89455
    },
    {
      "epoch": 13.324396782841823,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.009080651442413955,
      "loss": 0.7986,
      "num_input_tokens_seen": 51937272,
      "step": 89460
    },
    {
      "epoch": 13.325141495382782,
      "grad_norm": 0.021484375,
      "learning_rate": 0.009078860067480217,
      "loss": 0.801,
      "num_input_tokens_seen": 51939992,
      "step": 89465
    },
    {
      "epoch": 13.325886207923741,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.009077068792578124,
      "loss": 0.7955,
      "num_input_tokens_seen": 51943096,
      "step": 89470
    },
    {
      "epoch": 13.3266309204647,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.009075277617737935,
      "loss": 0.8039,
      "num_input_tokens_seen": 51945976,
      "step": 89475
    },
    {
      "epoch": 13.32737563300566,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.009073486542989907,
      "loss": 0.7914,
      "num_input_tokens_seen": 51948856,
      "step": 89480
    },
    {
      "epoch": 13.328120345546619,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.009071695568364301,
      "loss": 0.8086,
      "num_input_tokens_seen": 51951800,
      "step": 89485
    },
    {
      "epoch": 13.328865058087578,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.009069904693891368,
      "loss": 0.8016,
      "num_input_tokens_seen": 51954648,
      "step": 89490
    },
    {
      "epoch": 13.329609770628537,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.009068113919601373,
      "loss": 0.7819,
      "num_input_tokens_seen": 51957496,
      "step": 89495
    },
    {
      "epoch": 13.330354483169497,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.00906632324552456,
      "loss": 0.788,
      "num_input_tokens_seen": 51960472,
      "step": 89500
    },
    {
      "epoch": 13.331099195710456,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.00906453267169119,
      "loss": 0.7893,
      "num_input_tokens_seen": 51963288,
      "step": 89505
    },
    {
      "epoch": 13.331843908251415,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.009062742198131512,
      "loss": 0.7908,
      "num_input_tokens_seen": 51965976,
      "step": 89510
    },
    {
      "epoch": 13.332588620792373,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.009060951824875763,
      "loss": 0.7945,
      "num_input_tokens_seen": 51969112,
      "step": 89515
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0090591615519542,
      "loss": 0.7848,
      "num_input_tokens_seen": 51971896,
      "step": 89520
    },
    {
      "epoch": 13.334078045874293,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.00905737137939706,
      "loss": 0.8178,
      "num_input_tokens_seen": 51974712,
      "step": 89525
    },
    {
      "epoch": 13.334822758415251,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.009055581307234597,
      "loss": 0.8076,
      "num_input_tokens_seen": 51977848,
      "step": 89530
    },
    {
      "epoch": 13.33556747095621,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.009053791335497038,
      "loss": 0.7894,
      "num_input_tokens_seen": 51980792,
      "step": 89535
    },
    {
      "epoch": 13.33631218349717,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.009052001464214638,
      "loss": 0.7974,
      "num_input_tokens_seen": 51983544,
      "step": 89540
    },
    {
      "epoch": 13.33705689603813,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.009050211693417624,
      "loss": 0.8029,
      "num_input_tokens_seen": 51986520,
      "step": 89545
    },
    {
      "epoch": 13.337801608579088,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.009048422023136242,
      "loss": 0.7972,
      "num_input_tokens_seen": 51989272,
      "step": 89550
    },
    {
      "epoch": 13.338546321120047,
      "grad_norm": 0.046875,
      "learning_rate": 0.00904663245340072,
      "loss": 0.8066,
      "num_input_tokens_seen": 51992312,
      "step": 89555
    },
    {
      "epoch": 13.339291033661008,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.009044842984241285,
      "loss": 0.815,
      "num_input_tokens_seen": 51995224,
      "step": 89560
    },
    {
      "epoch": 13.340035746201966,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.009043053615688182,
      "loss": 0.7866,
      "num_input_tokens_seen": 51997880,
      "step": 89565
    },
    {
      "epoch": 13.340780458742925,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.009041264347771626,
      "loss": 0.8024,
      "num_input_tokens_seen": 52000696,
      "step": 89570
    },
    {
      "epoch": 13.341525171283884,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.009039475180521861,
      "loss": 0.8096,
      "num_input_tokens_seen": 52003576,
      "step": 89575
    },
    {
      "epoch": 13.342269883824844,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0090376861139691,
      "loss": 0.7948,
      "num_input_tokens_seen": 52006232,
      "step": 89580
    },
    {
      "epoch": 13.343014596365803,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.009035897148143577,
      "loss": 0.7866,
      "num_input_tokens_seen": 52009016,
      "step": 89585
    },
    {
      "epoch": 13.343759308906762,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.00903410828307551,
      "loss": 0.7934,
      "num_input_tokens_seen": 52011896,
      "step": 89590
    },
    {
      "epoch": 13.34450402144772,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.009032319518795113,
      "loss": 0.7839,
      "num_input_tokens_seen": 52014872,
      "step": 89595
    },
    {
      "epoch": 13.345248733988681,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.009030530855332617,
      "loss": 0.7948,
      "num_input_tokens_seen": 52017848,
      "step": 89600
    },
    {
      "epoch": 13.34599344652964,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.009028742292718234,
      "loss": 0.7864,
      "num_input_tokens_seen": 52020664,
      "step": 89605
    },
    {
      "epoch": 13.346738159070599,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.009026953830982183,
      "loss": 0.7872,
      "num_input_tokens_seen": 52023448,
      "step": 89610
    },
    {
      "epoch": 13.347482871611557,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.009025165470154674,
      "loss": 0.7959,
      "num_input_tokens_seen": 52026616,
      "step": 89615
    },
    {
      "epoch": 13.348227584152518,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.009023377210265925,
      "loss": 0.8303,
      "num_input_tokens_seen": 52029464,
      "step": 89620
    },
    {
      "epoch": 13.348972296693477,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.009021589051346136,
      "loss": 0.8106,
      "num_input_tokens_seen": 52032632,
      "step": 89625
    },
    {
      "epoch": 13.349717009234435,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.009019800993425533,
      "loss": 0.775,
      "num_input_tokens_seen": 52035288,
      "step": 89630
    },
    {
      "epoch": 13.350461721775394,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.009018013036534312,
      "loss": 0.8003,
      "num_input_tokens_seen": 52037976,
      "step": 89635
    },
    {
      "epoch": 13.351206434316353,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.00901622518070268,
      "loss": 0.787,
      "num_input_tokens_seen": 52041048,
      "step": 89640
    },
    {
      "epoch": 13.351951146857314,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.009014437425960843,
      "loss": 0.7981,
      "num_input_tokens_seen": 52043704,
      "step": 89645
    },
    {
      "epoch": 13.352695859398272,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.009012649772338998,
      "loss": 0.7826,
      "num_input_tokens_seen": 52046552,
      "step": 89650
    },
    {
      "epoch": 13.353440571939231,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.009010862219867353,
      "loss": 0.8026,
      "num_input_tokens_seen": 52049144,
      "step": 89655
    },
    {
      "epoch": 13.35418528448019,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.009009074768576101,
      "loss": 0.8046,
      "num_input_tokens_seen": 52052056,
      "step": 89660
    },
    {
      "epoch": 13.35492999702115,
      "grad_norm": 0.033203125,
      "learning_rate": 0.009007287418495444,
      "loss": 0.7929,
      "num_input_tokens_seen": 52054872,
      "step": 89665
    },
    {
      "epoch": 13.35567470956211,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.009005500169655579,
      "loss": 0.8296,
      "num_input_tokens_seen": 52057912,
      "step": 89670
    },
    {
      "epoch": 13.356419422103068,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.00900371302208669,
      "loss": 0.7972,
      "num_input_tokens_seen": 52060632,
      "step": 89675
    },
    {
      "epoch": 13.357164134644027,
      "grad_norm": 0.021240234375,
      "learning_rate": 0.00900192597581898,
      "loss": 0.7905,
      "num_input_tokens_seen": 52063256,
      "step": 89680
    },
    {
      "epoch": 13.357908847184987,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.009000139030882629,
      "loss": 0.7953,
      "num_input_tokens_seen": 52066104,
      "step": 89685
    },
    {
      "epoch": 13.358653559725946,
      "grad_norm": 0.041015625,
      "learning_rate": 0.008998352187307837,
      "loss": 0.8392,
      "num_input_tokens_seen": 52068920,
      "step": 89690
    },
    {
      "epoch": 13.359398272266905,
      "grad_norm": 0.0205078125,
      "learning_rate": 0.008996565445124777,
      "loss": 0.8097,
      "num_input_tokens_seen": 52071832,
      "step": 89695
    },
    {
      "epoch": 13.360142984807863,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.00899477880436365,
      "loss": 0.7935,
      "num_input_tokens_seen": 52075000,
      "step": 89700
    },
    {
      "epoch": 13.360887697348824,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.008992992265054628,
      "loss": 0.8026,
      "num_input_tokens_seen": 52077592,
      "step": 89705
    },
    {
      "epoch": 13.361632409889783,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0089912058272279,
      "loss": 0.8093,
      "num_input_tokens_seen": 52080536,
      "step": 89710
    },
    {
      "epoch": 13.362377122430741,
      "grad_norm": 0.033203125,
      "learning_rate": 0.008989419490913643,
      "loss": 0.7779,
      "num_input_tokens_seen": 52083704,
      "step": 89715
    },
    {
      "epoch": 13.3631218349717,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.008987633256142026,
      "loss": 0.7882,
      "num_input_tokens_seen": 52086872,
      "step": 89720
    },
    {
      "epoch": 13.36386654751266,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.008985847122943244,
      "loss": 0.8068,
      "num_input_tokens_seen": 52089816,
      "step": 89725
    },
    {
      "epoch": 13.36461126005362,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.008984061091347454,
      "loss": 0.7965,
      "num_input_tokens_seen": 52092664,
      "step": 89730
    },
    {
      "epoch": 13.365355972594578,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.008982275161384846,
      "loss": 0.7967,
      "num_input_tokens_seen": 52095640,
      "step": 89735
    },
    {
      "epoch": 13.366100685135537,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.008980489333085578,
      "loss": 0.8061,
      "num_input_tokens_seen": 52098360,
      "step": 89740
    },
    {
      "epoch": 13.366845397676498,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.008978703606479826,
      "loss": 0.7829,
      "num_input_tokens_seen": 52101400,
      "step": 89745
    },
    {
      "epoch": 13.367590110217456,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.00897691798159776,
      "loss": 0.8181,
      "num_input_tokens_seen": 52104504,
      "step": 89750
    },
    {
      "epoch": 13.368334822758415,
      "grad_norm": 0.076171875,
      "learning_rate": 0.008975132458469537,
      "loss": 0.8386,
      "num_input_tokens_seen": 52108536,
      "step": 89755
    },
    {
      "epoch": 13.369079535299374,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.00897334703712533,
      "loss": 0.8044,
      "num_input_tokens_seen": 52111288,
      "step": 89760
    },
    {
      "epoch": 13.369824247840334,
      "grad_norm": 0.037109375,
      "learning_rate": 0.008971561717595299,
      "loss": 0.8214,
      "num_input_tokens_seen": 52115224,
      "step": 89765
    },
    {
      "epoch": 13.370568960381293,
      "grad_norm": 0.037109375,
      "learning_rate": 0.008969776499909609,
      "loss": 0.7814,
      "num_input_tokens_seen": 52118104,
      "step": 89770
    },
    {
      "epoch": 13.371313672922252,
      "grad_norm": 0.05078125,
      "learning_rate": 0.008967991384098413,
      "loss": 0.7952,
      "num_input_tokens_seen": 52120888,
      "step": 89775
    },
    {
      "epoch": 13.37205838546321,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.008966206370191873,
      "loss": 0.7815,
      "num_input_tokens_seen": 52123832,
      "step": 89780
    },
    {
      "epoch": 13.372803098004171,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.00896442145822014,
      "loss": 0.7989,
      "num_input_tokens_seen": 52126584,
      "step": 89785
    },
    {
      "epoch": 13.37354781054513,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.00896263664821338,
      "loss": 0.7924,
      "num_input_tokens_seen": 52129560,
      "step": 89790
    },
    {
      "epoch": 13.374292523086089,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.008960851940201735,
      "loss": 0.803,
      "num_input_tokens_seen": 52132280,
      "step": 89795
    },
    {
      "epoch": 13.375037235627047,
      "grad_norm": 0.044921875,
      "learning_rate": 0.008959067334215354,
      "loss": 0.7872,
      "num_input_tokens_seen": 52136056,
      "step": 89800
    },
    {
      "epoch": 13.375781948168008,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.008957282830284395,
      "loss": 0.7884,
      "num_input_tokens_seen": 52139128,
      "step": 89805
    },
    {
      "epoch": 13.376526660708967,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.008955498428439,
      "loss": 0.8154,
      "num_input_tokens_seen": 52141848,
      "step": 89810
    },
    {
      "epoch": 13.377271373249926,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.008953714128709319,
      "loss": 0.8087,
      "num_input_tokens_seen": 52144472,
      "step": 89815
    },
    {
      "epoch": 13.378016085790884,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.008951929931125483,
      "loss": 0.79,
      "num_input_tokens_seen": 52147384,
      "step": 89820
    },
    {
      "epoch": 13.378760798331843,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.008950145835717654,
      "loss": 0.7808,
      "num_input_tokens_seen": 52150136,
      "step": 89825
    },
    {
      "epoch": 13.379505510872804,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.00894836184251596,
      "loss": 0.8024,
      "num_input_tokens_seen": 52153048,
      "step": 89830
    },
    {
      "epoch": 13.380250223413762,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.008946577951550539,
      "loss": 0.7947,
      "num_input_tokens_seen": 52155992,
      "step": 89835
    },
    {
      "epoch": 13.380994935954721,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.008944794162851533,
      "loss": 0.8114,
      "num_input_tokens_seen": 52158808,
      "step": 89840
    },
    {
      "epoch": 13.38173964849568,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.008943010476449074,
      "loss": 0.7926,
      "num_input_tokens_seen": 52161528,
      "step": 89845
    },
    {
      "epoch": 13.38248436103664,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.0089412268923733,
      "loss": 0.8041,
      "num_input_tokens_seen": 52164280,
      "step": 89850
    },
    {
      "epoch": 13.3832290735776,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.008939443410654335,
      "loss": 0.7935,
      "num_input_tokens_seen": 52167480,
      "step": 89855
    },
    {
      "epoch": 13.383973786118558,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.008937660031322318,
      "loss": 0.7885,
      "num_input_tokens_seen": 52170296,
      "step": 89860
    },
    {
      "epoch": 13.384718498659517,
      "grad_norm": 0.048828125,
      "learning_rate": 0.008935876754407369,
      "loss": 0.7801,
      "num_input_tokens_seen": 52173464,
      "step": 89865
    },
    {
      "epoch": 13.385463211200477,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.008934093579939623,
      "loss": 0.8254,
      "num_input_tokens_seen": 52176248,
      "step": 89870
    },
    {
      "epoch": 13.386207923741436,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0089323105079492,
      "loss": 0.797,
      "num_input_tokens_seen": 52179096,
      "step": 89875
    },
    {
      "epoch": 13.386952636282395,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.008930527538466225,
      "loss": 0.813,
      "num_input_tokens_seen": 52182104,
      "step": 89880
    },
    {
      "epoch": 13.387697348823353,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.008928744671520819,
      "loss": 0.8191,
      "num_input_tokens_seen": 52184888,
      "step": 89885
    },
    {
      "epoch": 13.388442061364314,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.008926961907143097,
      "loss": 0.791,
      "num_input_tokens_seen": 52187960,
      "step": 89890
    },
    {
      "epoch": 13.389186773905273,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.008925179245363185,
      "loss": 0.8077,
      "num_input_tokens_seen": 52190904,
      "step": 89895
    },
    {
      "epoch": 13.389931486446232,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.00892339668621119,
      "loss": 0.8139,
      "num_input_tokens_seen": 52193880,
      "step": 89900
    },
    {
      "epoch": 13.39067619898719,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.008921614229717238,
      "loss": 0.8309,
      "num_input_tokens_seen": 52196760,
      "step": 89905
    },
    {
      "epoch": 13.39142091152815,
      "grad_norm": 0.04296875,
      "learning_rate": 0.008919831875911435,
      "loss": 0.7883,
      "num_input_tokens_seen": 52199832,
      "step": 89910
    },
    {
      "epoch": 13.39216562406911,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.008918049624823893,
      "loss": 0.8375,
      "num_input_tokens_seen": 52202488,
      "step": 89915
    },
    {
      "epoch": 13.392910336610068,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.008916267476484722,
      "loss": 0.7896,
      "num_input_tokens_seen": 52205304,
      "step": 89920
    },
    {
      "epoch": 13.393655049151027,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.008914485430924024,
      "loss": 0.8292,
      "num_input_tokens_seen": 52207960,
      "step": 89925
    },
    {
      "epoch": 13.394399761691988,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.008912703488171914,
      "loss": 0.7849,
      "num_input_tokens_seen": 52210904,
      "step": 89930
    },
    {
      "epoch": 13.395144474232946,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.008910921648258489,
      "loss": 0.7907,
      "num_input_tokens_seen": 52213688,
      "step": 89935
    },
    {
      "epoch": 13.395889186773905,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.008909139911213854,
      "loss": 0.802,
      "num_input_tokens_seen": 52216920,
      "step": 89940
    },
    {
      "epoch": 13.396633899314864,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.00890735827706811,
      "loss": 0.8051,
      "num_input_tokens_seen": 52219736,
      "step": 89945
    },
    {
      "epoch": 13.397378611855824,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.008905576745851358,
      "loss": 0.7923,
      "num_input_tokens_seen": 52222776,
      "step": 89950
    },
    {
      "epoch": 13.398123324396783,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.008903795317593693,
      "loss": 0.7874,
      "num_input_tokens_seen": 52225624,
      "step": 89955
    },
    {
      "epoch": 13.398868036937742,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.008902013992325205,
      "loss": 0.801,
      "num_input_tokens_seen": 52228440,
      "step": 89960
    },
    {
      "epoch": 13.3996127494787,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.008900232770075999,
      "loss": 0.7907,
      "num_input_tokens_seen": 52231256,
      "step": 89965
    },
    {
      "epoch": 13.400357462019661,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.008898451650876158,
      "loss": 0.7997,
      "num_input_tokens_seen": 52234072,
      "step": 89970
    },
    {
      "epoch": 13.40110217456062,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.008896670634755778,
      "loss": 0.8084,
      "num_input_tokens_seen": 52237208,
      "step": 89975
    },
    {
      "epoch": 13.401846887101579,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.008894889721744936,
      "loss": 0.8111,
      "num_input_tokens_seen": 52239960,
      "step": 89980
    },
    {
      "epoch": 13.402591599642538,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.008893108911873737,
      "loss": 0.7994,
      "num_input_tokens_seen": 52242744,
      "step": 89985
    },
    {
      "epoch": 13.403336312183498,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.008891328205172246,
      "loss": 0.7906,
      "num_input_tokens_seen": 52245528,
      "step": 89990
    },
    {
      "epoch": 13.404081024724457,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.008889547601670565,
      "loss": 0.7992,
      "num_input_tokens_seen": 52248248,
      "step": 89995
    },
    {
      "epoch": 13.404825737265416,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.008887767101398767,
      "loss": 0.7994,
      "num_input_tokens_seen": 52251256,
      "step": 90000
    },
    {
      "epoch": 13.405570449806374,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.008885986704386927,
      "loss": 0.7918,
      "num_input_tokens_seen": 52254328,
      "step": 90005
    },
    {
      "epoch": 13.406315162347333,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.008884206410665134,
      "loss": 0.7996,
      "num_input_tokens_seen": 52257528,
      "step": 90010
    },
    {
      "epoch": 13.407059874888294,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.008882426220263449,
      "loss": 0.8088,
      "num_input_tokens_seen": 52260760,
      "step": 90015
    },
    {
      "epoch": 13.407804587429252,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.008880646133211961,
      "loss": 0.8156,
      "num_input_tokens_seen": 52263640,
      "step": 90020
    },
    {
      "epoch": 13.408549299970211,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.008878866149540733,
      "loss": 0.7971,
      "num_input_tokens_seen": 52266648,
      "step": 90025
    },
    {
      "epoch": 13.40929401251117,
      "grad_norm": 0.037109375,
      "learning_rate": 0.008877086269279844,
      "loss": 0.8134,
      "num_input_tokens_seen": 52269720,
      "step": 90030
    },
    {
      "epoch": 13.41003872505213,
      "grad_norm": 0.046875,
      "learning_rate": 0.008875306492459363,
      "loss": 0.7828,
      "num_input_tokens_seen": 52272728,
      "step": 90035
    },
    {
      "epoch": 13.41078343759309,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.008873526819109348,
      "loss": 0.8104,
      "num_input_tokens_seen": 52275480,
      "step": 90040
    },
    {
      "epoch": 13.411528150134048,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.008871747249259874,
      "loss": 0.7993,
      "num_input_tokens_seen": 52278520,
      "step": 90045
    },
    {
      "epoch": 13.412272862675007,
      "grad_norm": 0.0206298828125,
      "learning_rate": 0.008869967782940996,
      "loss": 0.7992,
      "num_input_tokens_seen": 52281112,
      "step": 90050
    },
    {
      "epoch": 13.413017575215967,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.008868188420182789,
      "loss": 0.7906,
      "num_input_tokens_seen": 52284184,
      "step": 90055
    },
    {
      "epoch": 13.413762287756926,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.0088664091610153,
      "loss": 0.7948,
      "num_input_tokens_seen": 52287096,
      "step": 90060
    },
    {
      "epoch": 13.414507000297885,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0088646300054686,
      "loss": 0.7995,
      "num_input_tokens_seen": 52290008,
      "step": 90065
    },
    {
      "epoch": 13.415251712838844,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.008862850953572737,
      "loss": 0.7956,
      "num_input_tokens_seen": 52292792,
      "step": 90070
    },
    {
      "epoch": 13.415996425379804,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.008861072005357773,
      "loss": 0.8014,
      "num_input_tokens_seen": 52295768,
      "step": 90075
    },
    {
      "epoch": 13.416741137920763,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.00885929316085376,
      "loss": 0.8059,
      "num_input_tokens_seen": 52298456,
      "step": 90080
    },
    {
      "epoch": 13.417485850461722,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.008857514420090739,
      "loss": 0.8079,
      "num_input_tokens_seen": 52301592,
      "step": 90085
    },
    {
      "epoch": 13.41823056300268,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.008855735783098775,
      "loss": 0.8103,
      "num_input_tokens_seen": 52304344,
      "step": 90090
    },
    {
      "epoch": 13.418975275543641,
      "grad_norm": 0.033203125,
      "learning_rate": 0.008853957249907904,
      "loss": 0.796,
      "num_input_tokens_seen": 52307352,
      "step": 90095
    },
    {
      "epoch": 13.4197199880846,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.008852178820548185,
      "loss": 0.7933,
      "num_input_tokens_seen": 52310104,
      "step": 90100
    },
    {
      "epoch": 13.420464700625558,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.008850400495049657,
      "loss": 0.7926,
      "num_input_tokens_seen": 52313240,
      "step": 90105
    },
    {
      "epoch": 13.421209413166517,
      "grad_norm": 0.044921875,
      "learning_rate": 0.008848622273442363,
      "loss": 0.7996,
      "num_input_tokens_seen": 52316088,
      "step": 90110
    },
    {
      "epoch": 13.421954125707478,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.008846844155756343,
      "loss": 0.8127,
      "num_input_tokens_seen": 52318904,
      "step": 90115
    },
    {
      "epoch": 13.422698838248436,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.008845066142021631,
      "loss": 0.7936,
      "num_input_tokens_seen": 52321912,
      "step": 90120
    },
    {
      "epoch": 13.423443550789395,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.008843288232268277,
      "loss": 0.7999,
      "num_input_tokens_seen": 52324408,
      "step": 90125
    },
    {
      "epoch": 13.424188263330354,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.008841510426526308,
      "loss": 0.7991,
      "num_input_tokens_seen": 52327512,
      "step": 90130
    },
    {
      "epoch": 13.424932975871315,
      "grad_norm": 0.046875,
      "learning_rate": 0.008839732724825763,
      "loss": 0.8001,
      "num_input_tokens_seen": 52330392,
      "step": 90135
    },
    {
      "epoch": 13.425677688412273,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.008837955127196673,
      "loss": 0.7955,
      "num_input_tokens_seen": 52333240,
      "step": 90140
    },
    {
      "epoch": 13.426422400953232,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.00883617763366907,
      "loss": 0.8014,
      "num_input_tokens_seen": 52336472,
      "step": 90145
    },
    {
      "epoch": 13.42716711349419,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.008834400244272974,
      "loss": 0.7939,
      "num_input_tokens_seen": 52339480,
      "step": 90150
    },
    {
      "epoch": 13.42791182603515,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.008832622959038425,
      "loss": 0.8051,
      "num_input_tokens_seen": 52342200,
      "step": 90155
    },
    {
      "epoch": 13.42865653857611,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.008830845777995446,
      "loss": 0.7998,
      "num_input_tokens_seen": 52345336,
      "step": 90160
    },
    {
      "epoch": 13.429401251117069,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.00882906870117405,
      "loss": 0.8018,
      "num_input_tokens_seen": 52348152,
      "step": 90165
    },
    {
      "epoch": 13.430145963658028,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.00882729172860427,
      "loss": 0.7856,
      "num_input_tokens_seen": 52351096,
      "step": 90170
    },
    {
      "epoch": 13.430890676198986,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.008825514860316123,
      "loss": 0.7973,
      "num_input_tokens_seen": 52354104,
      "step": 90175
    },
    {
      "epoch": 13.431635388739947,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.008823738096339628,
      "loss": 0.7961,
      "num_input_tokens_seen": 52356984,
      "step": 90180
    },
    {
      "epoch": 13.432380101280906,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.008821961436704797,
      "loss": 0.7835,
      "num_input_tokens_seen": 52360248,
      "step": 90185
    },
    {
      "epoch": 13.433124813821864,
      "grad_norm": 0.04296875,
      "learning_rate": 0.00882018488144165,
      "loss": 0.8061,
      "num_input_tokens_seen": 52363288,
      "step": 90190
    },
    {
      "epoch": 13.433869526362823,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.008818408430580203,
      "loss": 0.7979,
      "num_input_tokens_seen": 52366168,
      "step": 90195
    },
    {
      "epoch": 13.434614238903784,
      "grad_norm": 0.044921875,
      "learning_rate": 0.008816632084150455,
      "loss": 0.8001,
      "num_input_tokens_seen": 52369240,
      "step": 90200
    },
    {
      "epoch": 13.435358951444742,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.008814855842182428,
      "loss": 0.8114,
      "num_input_tokens_seen": 52371832,
      "step": 90205
    },
    {
      "epoch": 13.436103663985701,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.008813079704706125,
      "loss": 0.7985,
      "num_input_tokens_seen": 52374936,
      "step": 90210
    },
    {
      "epoch": 13.43684837652666,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.008811303671751553,
      "loss": 0.8089,
      "num_input_tokens_seen": 52377720,
      "step": 90215
    },
    {
      "epoch": 13.43759308906762,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.008809527743348712,
      "loss": 0.7915,
      "num_input_tokens_seen": 52380696,
      "step": 90220
    },
    {
      "epoch": 13.43833780160858,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.008807751919527615,
      "loss": 0.7929,
      "num_input_tokens_seen": 52383704,
      "step": 90225
    },
    {
      "epoch": 13.439082514149538,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.008805976200318245,
      "loss": 0.8156,
      "num_input_tokens_seen": 52386552,
      "step": 90230
    },
    {
      "epoch": 13.439827226690497,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.008804200585750621,
      "loss": 0.8002,
      "num_input_tokens_seen": 52389528,
      "step": 90235
    },
    {
      "epoch": 13.440571939231457,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00880242507585473,
      "loss": 0.7888,
      "num_input_tokens_seen": 52392472,
      "step": 90240
    },
    {
      "epoch": 13.441316651772416,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.008800649670660564,
      "loss": 0.7997,
      "num_input_tokens_seen": 52395128,
      "step": 90245
    },
    {
      "epoch": 13.442061364313375,
      "grad_norm": 0.037109375,
      "learning_rate": 0.008798874370198129,
      "loss": 0.7967,
      "num_input_tokens_seen": 52398040,
      "step": 90250
    },
    {
      "epoch": 13.442806076854334,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.008797099174497398,
      "loss": 0.8055,
      "num_input_tokens_seen": 52400696,
      "step": 90255
    },
    {
      "epoch": 13.443550789395294,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00879532408358838,
      "loss": 0.8002,
      "num_input_tokens_seen": 52403416,
      "step": 90260
    },
    {
      "epoch": 13.444295501936253,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.008793549097501052,
      "loss": 0.8,
      "num_input_tokens_seen": 52406072,
      "step": 90265
    },
    {
      "epoch": 13.445040214477212,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.008791774216265405,
      "loss": 0.7848,
      "num_input_tokens_seen": 52408824,
      "step": 90270
    },
    {
      "epoch": 13.44578492701817,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.008789999439911425,
      "loss": 0.7949,
      "num_input_tokens_seen": 52411704,
      "step": 90275
    },
    {
      "epoch": 13.446529639559131,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.008788224768469091,
      "loss": 0.7872,
      "num_input_tokens_seen": 52414520,
      "step": 90280
    },
    {
      "epoch": 13.44727435210009,
      "grad_norm": 0.03515625,
      "learning_rate": 0.008786450201968388,
      "loss": 0.8062,
      "num_input_tokens_seen": 52417432,
      "step": 90285
    },
    {
      "epoch": 13.448019064641048,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.008784675740439287,
      "loss": 0.7997,
      "num_input_tokens_seen": 52420312,
      "step": 90290
    },
    {
      "epoch": 13.448763777182007,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.00878290138391178,
      "loss": 0.8022,
      "num_input_tokens_seen": 52423096,
      "step": 90295
    },
    {
      "epoch": 13.449508489722968,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.008781127132415831,
      "loss": 0.7929,
      "num_input_tokens_seen": 52426232,
      "step": 90300
    },
    {
      "epoch": 13.450253202263927,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.008779352985981424,
      "loss": 0.8132,
      "num_input_tokens_seen": 52428888,
      "step": 90305
    },
    {
      "epoch": 13.450997914804885,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.008777578944638516,
      "loss": 0.7893,
      "num_input_tokens_seen": 52431608,
      "step": 90310
    },
    {
      "epoch": 13.451742627345844,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.008775805008417096,
      "loss": 0.7968,
      "num_input_tokens_seen": 52434744,
      "step": 90315
    },
    {
      "epoch": 13.452487339886805,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.008774031177347125,
      "loss": 0.7941,
      "num_input_tokens_seen": 52437656,
      "step": 90320
    },
    {
      "epoch": 13.453232052427763,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.008772257451458562,
      "loss": 0.7864,
      "num_input_tokens_seen": 52440728,
      "step": 90325
    },
    {
      "epoch": 13.453976764968722,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.008770483830781385,
      "loss": 0.7991,
      "num_input_tokens_seen": 52443640,
      "step": 90330
    },
    {
      "epoch": 13.45472147750968,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.008768710315345552,
      "loss": 0.8013,
      "num_input_tokens_seen": 52446904,
      "step": 90335
    },
    {
      "epoch": 13.45546619005064,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.008766936905181027,
      "loss": 0.7864,
      "num_input_tokens_seen": 52449720,
      "step": 90340
    },
    {
      "epoch": 13.4562109025916,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.00876516360031776,
      "loss": 0.7787,
      "num_input_tokens_seen": 52452344,
      "step": 90345
    },
    {
      "epoch": 13.456955615132559,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.008763390400785727,
      "loss": 0.7989,
      "num_input_tokens_seen": 52455384,
      "step": 90350
    },
    {
      "epoch": 13.457700327673518,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.00876161730661487,
      "loss": 0.8073,
      "num_input_tokens_seen": 52458168,
      "step": 90355
    },
    {
      "epoch": 13.458445040214476,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.008759844317835145,
      "loss": 0.7918,
      "num_input_tokens_seen": 52461048,
      "step": 90360
    },
    {
      "epoch": 13.459189752755437,
      "grad_norm": 0.033203125,
      "learning_rate": 0.008758071434476515,
      "loss": 0.7781,
      "num_input_tokens_seen": 52464056,
      "step": 90365
    },
    {
      "epoch": 13.459934465296396,
      "grad_norm": 0.095703125,
      "learning_rate": 0.008756298656568916,
      "loss": 0.852,
      "num_input_tokens_seen": 52467256,
      "step": 90370
    },
    {
      "epoch": 13.460679177837354,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.008754525984142312,
      "loss": 0.7959,
      "num_input_tokens_seen": 52470232,
      "step": 90375
    },
    {
      "epoch": 13.461423890378313,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.008752753417226639,
      "loss": 0.7864,
      "num_input_tokens_seen": 52473496,
      "step": 90380
    },
    {
      "epoch": 13.462168602919274,
      "grad_norm": 0.03515625,
      "learning_rate": 0.008750980955851851,
      "loss": 0.7874,
      "num_input_tokens_seen": 52476408,
      "step": 90385
    },
    {
      "epoch": 13.462913315460233,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.008749208600047891,
      "loss": 0.7988,
      "num_input_tokens_seen": 52479480,
      "step": 90390
    },
    {
      "epoch": 13.463658028001191,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.008747436349844697,
      "loss": 0.8098,
      "num_input_tokens_seen": 52482296,
      "step": 90395
    },
    {
      "epoch": 13.46440274054215,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.008745664205272213,
      "loss": 0.824,
      "num_input_tokens_seen": 52485400,
      "step": 90400
    },
    {
      "epoch": 13.46514745308311,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.008743892166360368,
      "loss": 0.7939,
      "num_input_tokens_seen": 52488088,
      "step": 90405
    },
    {
      "epoch": 13.46589216562407,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.008742120233139113,
      "loss": 0.791,
      "num_input_tokens_seen": 52491128,
      "step": 90410
    },
    {
      "epoch": 13.466636878165028,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.008740348405638369,
      "loss": 0.7917,
      "num_input_tokens_seen": 52493976,
      "step": 90415
    },
    {
      "epoch": 13.467381590705987,
      "grad_norm": 0.044921875,
      "learning_rate": 0.008738576683888085,
      "loss": 0.8003,
      "num_input_tokens_seen": 52497112,
      "step": 90420
    },
    {
      "epoch": 13.468126303246947,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.008736805067918177,
      "loss": 0.7865,
      "num_input_tokens_seen": 52499992,
      "step": 90425
    },
    {
      "epoch": 13.468871015787906,
      "grad_norm": 0.06640625,
      "learning_rate": 0.008735033557758587,
      "loss": 0.8474,
      "num_input_tokens_seen": 52502872,
      "step": 90430
    },
    {
      "epoch": 13.469615728328865,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.008733262153439239,
      "loss": 0.7966,
      "num_input_tokens_seen": 52506104,
      "step": 90435
    },
    {
      "epoch": 13.470360440869824,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.00873149085499005,
      "loss": 0.7951,
      "num_input_tokens_seen": 52509112,
      "step": 90440
    },
    {
      "epoch": 13.471105153410784,
      "grad_norm": 0.021484375,
      "learning_rate": 0.00872971966244096,
      "loss": 0.8017,
      "num_input_tokens_seen": 52512024,
      "step": 90445
    },
    {
      "epoch": 13.471849865951743,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.008727948575821882,
      "loss": 0.7996,
      "num_input_tokens_seen": 52515256,
      "step": 90450
    },
    {
      "epoch": 13.472594578492702,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.008726177595162733,
      "loss": 0.7966,
      "num_input_tokens_seen": 52518168,
      "step": 90455
    },
    {
      "epoch": 13.47333929103366,
      "grad_norm": 0.0196533203125,
      "learning_rate": 0.008724406720493443,
      "loss": 0.7956,
      "num_input_tokens_seen": 52521048,
      "step": 90460
    },
    {
      "epoch": 13.474084003574621,
      "grad_norm": 0.0546875,
      "learning_rate": 0.008722635951843923,
      "loss": 0.7939,
      "num_input_tokens_seen": 52523960,
      "step": 90465
    },
    {
      "epoch": 13.47482871611558,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.008720865289244081,
      "loss": 0.8017,
      "num_input_tokens_seen": 52527192,
      "step": 90470
    },
    {
      "epoch": 13.475573428656539,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.008719094732723847,
      "loss": 0.8059,
      "num_input_tokens_seen": 52530104,
      "step": 90475
    },
    {
      "epoch": 13.476318141197497,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.008717324282313123,
      "loss": 0.8006,
      "num_input_tokens_seen": 52533112,
      "step": 90480
    },
    {
      "epoch": 13.477062853738458,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.008715553938041812,
      "loss": 0.7931,
      "num_input_tokens_seen": 52535768,
      "step": 90485
    },
    {
      "epoch": 13.477807566279417,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.00871378369993984,
      "loss": 0.7893,
      "num_input_tokens_seen": 52538584,
      "step": 90490
    },
    {
      "epoch": 13.478552278820375,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.008712013568037092,
      "loss": 0.797,
      "num_input_tokens_seen": 52541528,
      "step": 90495
    },
    {
      "epoch": 13.479296991361334,
      "grad_norm": 0.046875,
      "learning_rate": 0.008710243542363494,
      "loss": 0.7881,
      "num_input_tokens_seen": 52544504,
      "step": 90500
    },
    {
      "epoch": 13.480041703902295,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.008708473622948932,
      "loss": 0.7953,
      "num_input_tokens_seen": 52547288,
      "step": 90505
    },
    {
      "epoch": 13.480786416443253,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.008706703809823317,
      "loss": 0.8131,
      "num_input_tokens_seen": 52550200,
      "step": 90510
    },
    {
      "epoch": 13.481531128984212,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.00870493410301655,
      "loss": 0.8021,
      "num_input_tokens_seen": 52553368,
      "step": 90515
    },
    {
      "epoch": 13.482275841525171,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.008703164502558518,
      "loss": 0.7937,
      "num_input_tokens_seen": 52556120,
      "step": 90520
    },
    {
      "epoch": 13.48302055406613,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.008701395008479126,
      "loss": 0.8097,
      "num_input_tokens_seen": 52559256,
      "step": 90525
    },
    {
      "epoch": 13.48376526660709,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.008699625620808256,
      "loss": 0.7784,
      "num_input_tokens_seen": 52562328,
      "step": 90530
    },
    {
      "epoch": 13.484509979148049,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.008697856339575814,
      "loss": 0.7967,
      "num_input_tokens_seen": 52565080,
      "step": 90535
    },
    {
      "epoch": 13.485254691689008,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.008696087164811676,
      "loss": 0.802,
      "num_input_tokens_seen": 52567768,
      "step": 90540
    },
    {
      "epoch": 13.485999404229966,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.008694318096545747,
      "loss": 0.8113,
      "num_input_tokens_seen": 52570392,
      "step": 90545
    },
    {
      "epoch": 13.486744116770927,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.008692549134807897,
      "loss": 0.7818,
      "num_input_tokens_seen": 52573176,
      "step": 90550
    },
    {
      "epoch": 13.487488829311886,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.008690780279628026,
      "loss": 0.8008,
      "num_input_tokens_seen": 52575800,
      "step": 90555
    },
    {
      "epoch": 13.488233541852845,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.00868901153103601,
      "loss": 0.8043,
      "num_input_tokens_seen": 52578872,
      "step": 90560
    },
    {
      "epoch": 13.488978254393803,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.008687242889061723,
      "loss": 0.79,
      "num_input_tokens_seen": 52581656,
      "step": 90565
    },
    {
      "epoch": 13.489722966934764,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.008685474353735057,
      "loss": 0.8051,
      "num_input_tokens_seen": 52584632,
      "step": 90570
    },
    {
      "epoch": 13.490467679475723,
      "grad_norm": 0.0211181640625,
      "learning_rate": 0.00868370592508588,
      "loss": 0.7941,
      "num_input_tokens_seen": 52587512,
      "step": 90575
    },
    {
      "epoch": 13.491212392016681,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.008681937603144077,
      "loss": 0.7885,
      "num_input_tokens_seen": 52590488,
      "step": 90580
    },
    {
      "epoch": 13.49195710455764,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.008680169387939515,
      "loss": 0.7913,
      "num_input_tokens_seen": 52593368,
      "step": 90585
    },
    {
      "epoch": 13.4927018170986,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.00867840127950207,
      "loss": 0.8145,
      "num_input_tokens_seen": 52596152,
      "step": 90590
    },
    {
      "epoch": 13.49344652963956,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.008676633277861603,
      "loss": 0.7953,
      "num_input_tokens_seen": 52599064,
      "step": 90595
    },
    {
      "epoch": 13.494191242180518,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.008674865383047998,
      "loss": 0.8083,
      "num_input_tokens_seen": 52602104,
      "step": 90600
    },
    {
      "epoch": 13.494935954721477,
      "grad_norm": 0.048828125,
      "learning_rate": 0.008673097595091113,
      "loss": 0.789,
      "num_input_tokens_seen": 52605208,
      "step": 90605
    },
    {
      "epoch": 13.495680667262437,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.008671329914020807,
      "loss": 0.7744,
      "num_input_tokens_seen": 52608056,
      "step": 90610
    },
    {
      "epoch": 13.496425379803396,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.008669562339866956,
      "loss": 0.7996,
      "num_input_tokens_seen": 52611032,
      "step": 90615
    },
    {
      "epoch": 13.497170092344355,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.008667794872659411,
      "loss": 0.8105,
      "num_input_tokens_seen": 52613944,
      "step": 90620
    },
    {
      "epoch": 13.497914804885314,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.00866602751242804,
      "loss": 0.7993,
      "num_input_tokens_seen": 52617016,
      "step": 90625
    },
    {
      "epoch": 13.498659517426274,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.00866426025920269,
      "loss": 0.792,
      "num_input_tokens_seen": 52619736,
      "step": 90630
    },
    {
      "epoch": 13.499404229967233,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.008662493113013231,
      "loss": 0.7825,
      "num_input_tokens_seen": 52622648,
      "step": 90635
    },
    {
      "epoch": 13.500148942508192,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.008660726073889511,
      "loss": 0.77,
      "num_input_tokens_seen": 52625688,
      "step": 90640
    },
    {
      "epoch": 13.50089365504915,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.008658959141861372,
      "loss": 0.79,
      "num_input_tokens_seen": 52628376,
      "step": 90645
    },
    {
      "epoch": 13.501638367590111,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.00865719231695868,
      "loss": 0.812,
      "num_input_tokens_seen": 52631480,
      "step": 90650
    },
    {
      "epoch": 13.50238308013107,
      "grad_norm": 0.044921875,
      "learning_rate": 0.008655425599211279,
      "loss": 0.7872,
      "num_input_tokens_seen": 52634328,
      "step": 90655
    },
    {
      "epoch": 13.503127792672029,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.008653658988649013,
      "loss": 0.8165,
      "num_input_tokens_seen": 52637144,
      "step": 90660
    },
    {
      "epoch": 13.503872505212987,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.008651892485301724,
      "loss": 0.7873,
      "num_input_tokens_seen": 52640184,
      "step": 90665
    },
    {
      "epoch": 13.504617217753946,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.008650126089199265,
      "loss": 0.7926,
      "num_input_tokens_seen": 52643288,
      "step": 90670
    },
    {
      "epoch": 13.505361930294907,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.008648359800371463,
      "loss": 0.7767,
      "num_input_tokens_seen": 52646232,
      "step": 90675
    },
    {
      "epoch": 13.506106642835865,
      "grad_norm": 0.03515625,
      "learning_rate": 0.008646593618848175,
      "loss": 0.7762,
      "num_input_tokens_seen": 52649336,
      "step": 90680
    },
    {
      "epoch": 13.506851355376824,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00864482754465923,
      "loss": 0.7966,
      "num_input_tokens_seen": 52651928,
      "step": 90685
    },
    {
      "epoch": 13.507596067917785,
      "grad_norm": 0.134765625,
      "learning_rate": 0.008643061577834458,
      "loss": 0.8564,
      "num_input_tokens_seen": 52654584,
      "step": 90690
    },
    {
      "epoch": 13.508340780458743,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.008641295718403708,
      "loss": 0.8013,
      "num_input_tokens_seen": 52657336,
      "step": 90695
    },
    {
      "epoch": 13.509085492999702,
      "grad_norm": 0.048828125,
      "learning_rate": 0.008639529966396796,
      "loss": 0.7703,
      "num_input_tokens_seen": 52660088,
      "step": 90700
    },
    {
      "epoch": 13.509830205540661,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.008637764321843567,
      "loss": 0.8143,
      "num_input_tokens_seen": 52663288,
      "step": 90705
    },
    {
      "epoch": 13.51057491808162,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.008635998784773839,
      "loss": 0.7862,
      "num_input_tokens_seen": 52666072,
      "step": 90710
    },
    {
      "epoch": 13.51131963062258,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.008634233355217447,
      "loss": 0.7949,
      "num_input_tokens_seen": 52668824,
      "step": 90715
    },
    {
      "epoch": 13.512064343163539,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.008632468033204214,
      "loss": 0.7934,
      "num_input_tokens_seen": 52671416,
      "step": 90720
    },
    {
      "epoch": 13.512809055704498,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.008630702818763962,
      "loss": 0.7903,
      "num_input_tokens_seen": 52674200,
      "step": 90725
    },
    {
      "epoch": 13.513553768245457,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.008628937711926512,
      "loss": 0.786,
      "num_input_tokens_seen": 52677016,
      "step": 90730
    },
    {
      "epoch": 13.514298480786417,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.00862717271272168,
      "loss": 0.7972,
      "num_input_tokens_seen": 52679768,
      "step": 90735
    },
    {
      "epoch": 13.515043193327376,
      "grad_norm": 0.046875,
      "learning_rate": 0.008625407821179293,
      "loss": 0.7913,
      "num_input_tokens_seen": 52682872,
      "step": 90740
    },
    {
      "epoch": 13.515787905868335,
      "grad_norm": 0.046875,
      "learning_rate": 0.008623643037329154,
      "loss": 0.7828,
      "num_input_tokens_seen": 52686072,
      "step": 90745
    },
    {
      "epoch": 13.516532618409293,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.008621878361201093,
      "loss": 0.7713,
      "num_input_tokens_seen": 52688664,
      "step": 90750
    },
    {
      "epoch": 13.517277330950254,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.008620113792824909,
      "loss": 0.7881,
      "num_input_tokens_seen": 52691704,
      "step": 90755
    },
    {
      "epoch": 13.518022043491213,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.008618349332230423,
      "loss": 0.8002,
      "num_input_tokens_seen": 52694648,
      "step": 90760
    },
    {
      "epoch": 13.518766756032171,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.008616584979447438,
      "loss": 0.8076,
      "num_input_tokens_seen": 52697880,
      "step": 90765
    },
    {
      "epoch": 13.51951146857313,
      "grad_norm": 0.056640625,
      "learning_rate": 0.008614820734505757,
      "loss": 0.7738,
      "num_input_tokens_seen": 52700632,
      "step": 90770
    },
    {
      "epoch": 13.52025618111409,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0086130565974352,
      "loss": 0.8046,
      "num_input_tokens_seen": 52703416,
      "step": 90775
    },
    {
      "epoch": 13.52100089365505,
      "grad_norm": 0.0546875,
      "learning_rate": 0.008611292568265554,
      "loss": 0.8092,
      "num_input_tokens_seen": 52706616,
      "step": 90780
    },
    {
      "epoch": 13.521745606196008,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.008609528647026625,
      "loss": 0.7834,
      "num_input_tokens_seen": 52709400,
      "step": 90785
    },
    {
      "epoch": 13.522490318736967,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.008607764833748218,
      "loss": 0.7943,
      "num_input_tokens_seen": 52712120,
      "step": 90790
    },
    {
      "epoch": 13.523235031277927,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00860600112846013,
      "loss": 0.7712,
      "num_input_tokens_seen": 52715000,
      "step": 90795
    },
    {
      "epoch": 13.523979743818886,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.008604237531192153,
      "loss": 0.8012,
      "num_input_tokens_seen": 52717880,
      "step": 90800
    },
    {
      "epoch": 13.524724456359845,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.008602474041974076,
      "loss": 0.7966,
      "num_input_tokens_seen": 52720632,
      "step": 90805
    },
    {
      "epoch": 13.525469168900804,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.008600710660835705,
      "loss": 0.792,
      "num_input_tokens_seen": 52723544,
      "step": 90810
    },
    {
      "epoch": 13.526213881441764,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.008598947387806817,
      "loss": 0.7832,
      "num_input_tokens_seen": 52726424,
      "step": 90815
    },
    {
      "epoch": 13.526958593982723,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.008597184222917213,
      "loss": 0.8021,
      "num_input_tokens_seen": 52729048,
      "step": 90820
    },
    {
      "epoch": 13.527703306523682,
      "grad_norm": 0.041015625,
      "learning_rate": 0.008595421166196668,
      "loss": 0.7757,
      "num_input_tokens_seen": 52731896,
      "step": 90825
    },
    {
      "epoch": 13.52844801906464,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.008593658217674981,
      "loss": 0.8097,
      "num_input_tokens_seen": 52734776,
      "step": 90830
    },
    {
      "epoch": 13.529192731605601,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.008591895377381918,
      "loss": 0.7987,
      "num_input_tokens_seen": 52737592,
      "step": 90835
    },
    {
      "epoch": 13.52993744414656,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.008590132645347278,
      "loss": 0.8007,
      "num_input_tokens_seen": 52740632,
      "step": 90840
    },
    {
      "epoch": 13.530682156687519,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.008588370021600828,
      "loss": 0.809,
      "num_input_tokens_seen": 52743864,
      "step": 90845
    },
    {
      "epoch": 13.531426869228477,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.008586607506172355,
      "loss": 0.821,
      "num_input_tokens_seen": 52747288,
      "step": 90850
    },
    {
      "epoch": 13.532171581769436,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.00858484509909163,
      "loss": 0.7833,
      "num_input_tokens_seen": 52750040,
      "step": 90855
    },
    {
      "epoch": 13.532916294310397,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.00858308280038842,
      "loss": 0.829,
      "num_input_tokens_seen": 52752856,
      "step": 90860
    },
    {
      "epoch": 13.533661006851355,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.008581320610092508,
      "loss": 0.8244,
      "num_input_tokens_seen": 52756024,
      "step": 90865
    },
    {
      "epoch": 13.534405719392314,
      "grad_norm": 0.033203125,
      "learning_rate": 0.008579558528233657,
      "loss": 0.7977,
      "num_input_tokens_seen": 52758744,
      "step": 90870
    },
    {
      "epoch": 13.535150431933273,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.008577796554841643,
      "loss": 0.7957,
      "num_input_tokens_seen": 52761432,
      "step": 90875
    },
    {
      "epoch": 13.535895144474233,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.00857603468994623,
      "loss": 0.7957,
      "num_input_tokens_seen": 52764376,
      "step": 90880
    },
    {
      "epoch": 13.536639857015192,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.008574272933577176,
      "loss": 0.8016,
      "num_input_tokens_seen": 52767032,
      "step": 90885
    },
    {
      "epoch": 13.537384569556151,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.008572511285764257,
      "loss": 0.8048,
      "num_input_tokens_seen": 52769976,
      "step": 90890
    },
    {
      "epoch": 13.53812928209711,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.008570749746537222,
      "loss": 0.7982,
      "num_input_tokens_seen": 52773112,
      "step": 90895
    },
    {
      "epoch": 13.53887399463807,
      "grad_norm": 0.03515625,
      "learning_rate": 0.008568988315925839,
      "loss": 0.7889,
      "num_input_tokens_seen": 52775960,
      "step": 90900
    },
    {
      "epoch": 13.539618707179029,
      "grad_norm": 0.033203125,
      "learning_rate": 0.008567226993959858,
      "loss": 0.7933,
      "num_input_tokens_seen": 52778776,
      "step": 90905
    },
    {
      "epoch": 13.540363419719988,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.008565465780669043,
      "loss": 0.8073,
      "num_input_tokens_seen": 52781464,
      "step": 90910
    },
    {
      "epoch": 13.541108132260947,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.008563704676083144,
      "loss": 0.8034,
      "num_input_tokens_seen": 52784120,
      "step": 90915
    },
    {
      "epoch": 13.541852844801907,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.008561943680231917,
      "loss": 0.7838,
      "num_input_tokens_seen": 52786744,
      "step": 90920
    },
    {
      "epoch": 13.542597557342866,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.008560182793145103,
      "loss": 0.798,
      "num_input_tokens_seen": 52789688,
      "step": 90925
    },
    {
      "epoch": 13.543342269883825,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.008558422014852454,
      "loss": 0.7852,
      "num_input_tokens_seen": 52792472,
      "step": 90930
    },
    {
      "epoch": 13.544086982424783,
      "grad_norm": 0.02099609375,
      "learning_rate": 0.00855666134538372,
      "loss": 0.8066,
      "num_input_tokens_seen": 52795064,
      "step": 90935
    },
    {
      "epoch": 13.544831694965744,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.00855490078476864,
      "loss": 0.7747,
      "num_input_tokens_seen": 52797912,
      "step": 90940
    },
    {
      "epoch": 13.545576407506703,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.008553140333036969,
      "loss": 0.7829,
      "num_input_tokens_seen": 52800792,
      "step": 90945
    },
    {
      "epoch": 13.546321120047661,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.008551379990218428,
      "loss": 0.7807,
      "num_input_tokens_seen": 52803832,
      "step": 90950
    },
    {
      "epoch": 13.54706583258862,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.008549619756342778,
      "loss": 0.8056,
      "num_input_tokens_seen": 52806744,
      "step": 90955
    },
    {
      "epoch": 13.54781054512958,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.008547859631439746,
      "loss": 0.8028,
      "num_input_tokens_seen": 52809592,
      "step": 90960
    },
    {
      "epoch": 13.54855525767054,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.008546099615539062,
      "loss": 0.8031,
      "num_input_tokens_seen": 52812280,
      "step": 90965
    },
    {
      "epoch": 13.549299970211498,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00854433970867047,
      "loss": 0.7793,
      "num_input_tokens_seen": 52815128,
      "step": 90970
    },
    {
      "epoch": 13.550044682752457,
      "grad_norm": 0.04296875,
      "learning_rate": 0.008542579910863693,
      "loss": 0.8014,
      "num_input_tokens_seen": 52817848,
      "step": 90975
    },
    {
      "epoch": 13.550789395293418,
      "grad_norm": 0.0625,
      "learning_rate": 0.008540820222148469,
      "loss": 0.7907,
      "num_input_tokens_seen": 52820984,
      "step": 90980
    },
    {
      "epoch": 13.551534107834376,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.008539060642554526,
      "loss": 0.7808,
      "num_input_tokens_seen": 52823832,
      "step": 90985
    },
    {
      "epoch": 13.552278820375335,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.008537301172111584,
      "loss": 0.8123,
      "num_input_tokens_seen": 52826936,
      "step": 90990
    },
    {
      "epoch": 13.553023532916294,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.008535541810849366,
      "loss": 0.8112,
      "num_input_tokens_seen": 52830104,
      "step": 90995
    },
    {
      "epoch": 13.553768245457253,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.008533782558797606,
      "loss": 0.805,
      "num_input_tokens_seen": 52832728,
      "step": 91000
    },
    {
      "epoch": 13.554512957998213,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.008532023415986016,
      "loss": 0.8073,
      "num_input_tokens_seen": 52835640,
      "step": 91005
    },
    {
      "epoch": 13.555257670539172,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.00853026438244431,
      "loss": 0.7882,
      "num_input_tokens_seen": 52838712,
      "step": 91010
    },
    {
      "epoch": 13.55600238308013,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.008528505458202218,
      "loss": 0.7999,
      "num_input_tokens_seen": 52841400,
      "step": 91015
    },
    {
      "epoch": 13.556747095621091,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.008526746643289444,
      "loss": 0.7964,
      "num_input_tokens_seen": 52844056,
      "step": 91020
    },
    {
      "epoch": 13.55749180816205,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.008524987937735711,
      "loss": 0.7632,
      "num_input_tokens_seen": 52846776,
      "step": 91025
    },
    {
      "epoch": 13.558236520703009,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.008523229341570722,
      "loss": 0.788,
      "num_input_tokens_seen": 52849720,
      "step": 91030
    },
    {
      "epoch": 13.558981233243967,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.008521470854824194,
      "loss": 0.8036,
      "num_input_tokens_seen": 52852632,
      "step": 91035
    },
    {
      "epoch": 13.559725945784926,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.008519712477525832,
      "loss": 0.7965,
      "num_input_tokens_seen": 52855512,
      "step": 91040
    },
    {
      "epoch": 13.560470658325887,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.008517954209705336,
      "loss": 0.7991,
      "num_input_tokens_seen": 52858616,
      "step": 91045
    },
    {
      "epoch": 13.561215370866845,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.008516196051392422,
      "loss": 0.8046,
      "num_input_tokens_seen": 52861528,
      "step": 91050
    },
    {
      "epoch": 13.561960083407804,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.008514438002616786,
      "loss": 0.7842,
      "num_input_tokens_seen": 52864248,
      "step": 91055
    },
    {
      "epoch": 13.562704795948763,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.008512680063408128,
      "loss": 0.7727,
      "num_input_tokens_seen": 52867064,
      "step": 91060
    },
    {
      "epoch": 13.563449508489724,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00851092223379614,
      "loss": 0.7948,
      "num_input_tokens_seen": 52869976,
      "step": 91065
    },
    {
      "epoch": 13.564194221030682,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.008509164513810534,
      "loss": 0.7958,
      "num_input_tokens_seen": 52872888,
      "step": 91070
    },
    {
      "epoch": 13.564938933571641,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.008507406903480988,
      "loss": 0.7725,
      "num_input_tokens_seen": 52875832,
      "step": 91075
    },
    {
      "epoch": 13.5656836461126,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00850564940283721,
      "loss": 0.792,
      "num_input_tokens_seen": 52878584,
      "step": 91080
    },
    {
      "epoch": 13.56642835865356,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.008503892011908884,
      "loss": 0.7615,
      "num_input_tokens_seen": 52881720,
      "step": 91085
    },
    {
      "epoch": 13.567173071194519,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.008502134730725695,
      "loss": 0.7869,
      "num_input_tokens_seen": 52884504,
      "step": 91090
    },
    {
      "epoch": 13.567917783735478,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.008500377559317344,
      "loss": 0.8308,
      "num_input_tokens_seen": 52887512,
      "step": 91095
    },
    {
      "epoch": 13.568662496276437,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0084986204977135,
      "loss": 0.7551,
      "num_input_tokens_seen": 52890360,
      "step": 91100
    },
    {
      "epoch": 13.569407208817397,
      "grad_norm": 0.068359375,
      "learning_rate": 0.00849686354594386,
      "loss": 0.7773,
      "num_input_tokens_seen": 52893400,
      "step": 91105
    },
    {
      "epoch": 13.570151921358356,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.008495106704038103,
      "loss": 0.8737,
      "num_input_tokens_seen": 52896344,
      "step": 91110
    },
    {
      "epoch": 13.570896633899315,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0084933499720259,
      "loss": 0.778,
      "num_input_tokens_seen": 52899352,
      "step": 91115
    },
    {
      "epoch": 13.571641346440273,
      "grad_norm": 0.0625,
      "learning_rate": 0.008491593349936941,
      "loss": 0.7893,
      "num_input_tokens_seen": 52902264,
      "step": 91120
    },
    {
      "epoch": 13.572386058981234,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.008489836837800899,
      "loss": 0.761,
      "num_input_tokens_seen": 52905080,
      "step": 91125
    },
    {
      "epoch": 13.573130771522193,
      "grad_norm": 0.1328125,
      "learning_rate": 0.008488080435647447,
      "loss": 0.8112,
      "num_input_tokens_seen": 52908120,
      "step": 91130
    },
    {
      "epoch": 13.573875484063151,
      "grad_norm": 0.08203125,
      "learning_rate": 0.00848632414350625,
      "loss": 0.7815,
      "num_input_tokens_seen": 52910904,
      "step": 91135
    },
    {
      "epoch": 13.57462019660411,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.008484567961406994,
      "loss": 0.7939,
      "num_input_tokens_seen": 52913624,
      "step": 91140
    },
    {
      "epoch": 13.57536490914507,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.008482811889379334,
      "loss": 0.762,
      "num_input_tokens_seen": 52916504,
      "step": 91145
    },
    {
      "epoch": 13.57610962168603,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.00848105592745295,
      "loss": 0.7982,
      "num_input_tokens_seen": 52919288,
      "step": 91150
    },
    {
      "epoch": 13.576854334226988,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.008479300075657495,
      "loss": 0.7946,
      "num_input_tokens_seen": 52922328,
      "step": 91155
    },
    {
      "epoch": 13.577599046767947,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.008477544334022641,
      "loss": 0.7904,
      "num_input_tokens_seen": 52925240,
      "step": 91160
    },
    {
      "epoch": 13.578343759308908,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.00847578870257805,
      "loss": 0.7666,
      "num_input_tokens_seen": 52928088,
      "step": 91165
    },
    {
      "epoch": 13.579088471849866,
      "grad_norm": 0.037109375,
      "learning_rate": 0.008474033181353369,
      "loss": 0.7961,
      "num_input_tokens_seen": 52931320,
      "step": 91170
    },
    {
      "epoch": 13.579833184390825,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.008472277770378272,
      "loss": 0.8016,
      "num_input_tokens_seen": 52934136,
      "step": 91175
    },
    {
      "epoch": 13.580577896931784,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.008470522469682407,
      "loss": 0.774,
      "num_input_tokens_seen": 52936952,
      "step": 91180
    },
    {
      "epoch": 13.581322609472743,
      "grad_norm": 0.046875,
      "learning_rate": 0.008468767279295429,
      "loss": 0.773,
      "num_input_tokens_seen": 52939672,
      "step": 91185
    },
    {
      "epoch": 13.582067322013703,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.008467012199246985,
      "loss": 0.7911,
      "num_input_tokens_seen": 52942392,
      "step": 91190
    },
    {
      "epoch": 13.582812034554662,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.008465257229566734,
      "loss": 0.7949,
      "num_input_tokens_seen": 52945432,
      "step": 91195
    },
    {
      "epoch": 13.58355674709562,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.008463502370284315,
      "loss": 0.7809,
      "num_input_tokens_seen": 52948632,
      "step": 91200
    },
    {
      "epoch": 13.584301459636581,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.008461747621429384,
      "loss": 0.7854,
      "num_input_tokens_seen": 52951416,
      "step": 91205
    },
    {
      "epoch": 13.58504617217754,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.008459992983031586,
      "loss": 0.7662,
      "num_input_tokens_seen": 52954360,
      "step": 91210
    },
    {
      "epoch": 13.585790884718499,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.008458238455120552,
      "loss": 0.7481,
      "num_input_tokens_seen": 52957144,
      "step": 91215
    },
    {
      "epoch": 13.586535597259457,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.008456484037725935,
      "loss": 0.8065,
      "num_input_tokens_seen": 52960184,
      "step": 91220
    },
    {
      "epoch": 13.587280309800416,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.008454729730877365,
      "loss": 0.8039,
      "num_input_tokens_seen": 52963288,
      "step": 91225
    },
    {
      "epoch": 13.588025022341377,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.008452975534604492,
      "loss": 0.7832,
      "num_input_tokens_seen": 52966136,
      "step": 91230
    },
    {
      "epoch": 13.588769734882336,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.008451221448936933,
      "loss": 0.8241,
      "num_input_tokens_seen": 52969080,
      "step": 91235
    },
    {
      "epoch": 13.589514447423294,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.008449467473904342,
      "loss": 0.7814,
      "num_input_tokens_seen": 52972280,
      "step": 91240
    },
    {
      "epoch": 13.590259159964253,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.008447713609536337,
      "loss": 0.8047,
      "num_input_tokens_seen": 52975288,
      "step": 91245
    },
    {
      "epoch": 13.591003872505214,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.00844595985586255,
      "loss": 0.8443,
      "num_input_tokens_seen": 52977912,
      "step": 91250
    },
    {
      "epoch": 13.591748585046172,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.008444206212912614,
      "loss": 0.7931,
      "num_input_tokens_seen": 52980856,
      "step": 91255
    },
    {
      "epoch": 13.592493297587131,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.008442452680716143,
      "loss": 0.7817,
      "num_input_tokens_seen": 52983576,
      "step": 91260
    },
    {
      "epoch": 13.59323801012809,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.008440699259302775,
      "loss": 0.8058,
      "num_input_tokens_seen": 52986648,
      "step": 91265
    },
    {
      "epoch": 13.59398272266905,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.008438945948702122,
      "loss": 0.829,
      "num_input_tokens_seen": 52989816,
      "step": 91270
    },
    {
      "epoch": 13.59472743521001,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.008437192748943813,
      "loss": 0.7807,
      "num_input_tokens_seen": 52992568,
      "step": 91275
    },
    {
      "epoch": 13.595472147750968,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.008435439660057458,
      "loss": 0.7838,
      "num_input_tokens_seen": 52995384,
      "step": 91280
    },
    {
      "epoch": 13.596216860291927,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.008433686682072682,
      "loss": 0.8222,
      "num_input_tokens_seen": 52998360,
      "step": 91285
    },
    {
      "epoch": 13.596961572832887,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.008431933815019098,
      "loss": 0.7922,
      "num_input_tokens_seen": 53001240,
      "step": 91290
    },
    {
      "epoch": 13.597706285373846,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.008430181058926308,
      "loss": 0.8094,
      "num_input_tokens_seen": 53004088,
      "step": 91295
    },
    {
      "epoch": 13.598450997914805,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.00842842841382394,
      "loss": 0.7876,
      "num_input_tokens_seen": 53006904,
      "step": 91300
    },
    {
      "epoch": 13.599195710455763,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.008426675879741588,
      "loss": 0.8021,
      "num_input_tokens_seen": 53009752,
      "step": 91305
    },
    {
      "epoch": 13.599940422996724,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.008424923456708873,
      "loss": 0.7843,
      "num_input_tokens_seen": 53012920,
      "step": 91310
    },
    {
      "epoch": 13.600685135537683,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.008423171144755392,
      "loss": 0.7852,
      "num_input_tokens_seen": 53015576,
      "step": 91315
    },
    {
      "epoch": 13.601429848078642,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.008421418943910752,
      "loss": 0.8091,
      "num_input_tokens_seen": 53018680,
      "step": 91320
    },
    {
      "epoch": 13.6021745606196,
      "grad_norm": 0.0703125,
      "learning_rate": 0.008419666854204552,
      "loss": 0.8106,
      "num_input_tokens_seen": 53021912,
      "step": 91325
    },
    {
      "epoch": 13.60291927316056,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.008417914875666387,
      "loss": 0.8169,
      "num_input_tokens_seen": 53024472,
      "step": 91330
    },
    {
      "epoch": 13.60366398570152,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.008416163008325867,
      "loss": 0.8041,
      "num_input_tokens_seen": 53027352,
      "step": 91335
    },
    {
      "epoch": 13.604408698242478,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.008414411252212574,
      "loss": 0.8046,
      "num_input_tokens_seen": 53030328,
      "step": 91340
    },
    {
      "epoch": 13.605153410783437,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.008412659607356114,
      "loss": 0.7858,
      "num_input_tokens_seen": 53033240,
      "step": 91345
    },
    {
      "epoch": 13.605898123324398,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.008410908073786071,
      "loss": 0.7981,
      "num_input_tokens_seen": 53036536,
      "step": 91350
    },
    {
      "epoch": 13.606642835865356,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.008409156651532046,
      "loss": 0.8001,
      "num_input_tokens_seen": 53039224,
      "step": 91355
    },
    {
      "epoch": 13.607387548406315,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.008407405340623612,
      "loss": 0.7868,
      "num_input_tokens_seen": 53042680,
      "step": 91360
    },
    {
      "epoch": 13.608132260947274,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.008405654141090374,
      "loss": 0.7808,
      "num_input_tokens_seen": 53045560,
      "step": 91365
    },
    {
      "epoch": 13.608876973488233,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.008403903052961905,
      "loss": 0.8125,
      "num_input_tokens_seen": 53048408,
      "step": 91370
    },
    {
      "epoch": 13.609621686029193,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.008402152076267783,
      "loss": 0.8102,
      "num_input_tokens_seen": 53051704,
      "step": 91375
    },
    {
      "epoch": 13.610366398570152,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.008400401211037604,
      "loss": 0.7773,
      "num_input_tokens_seen": 53054424,
      "step": 91380
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.008398650457300939,
      "loss": 0.8137,
      "num_input_tokens_seen": 53057464,
      "step": 91385
    },
    {
      "epoch": 13.61185582365207,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.008396899815087363,
      "loss": 0.7993,
      "num_input_tokens_seen": 53060536,
      "step": 91390
    },
    {
      "epoch": 13.61260053619303,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.008395149284426449,
      "loss": 0.7834,
      "num_input_tokens_seen": 53063576,
      "step": 91395
    },
    {
      "epoch": 13.613345248733989,
      "grad_norm": 0.056640625,
      "learning_rate": 0.008393398865347781,
      "loss": 0.8052,
      "num_input_tokens_seen": 53066744,
      "step": 91400
    },
    {
      "epoch": 13.614089961274948,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.008391648557880925,
      "loss": 0.7799,
      "num_input_tokens_seen": 53070008,
      "step": 91405
    },
    {
      "epoch": 13.614834673815906,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.008389898362055443,
      "loss": 0.7964,
      "num_input_tokens_seen": 53073080,
      "step": 91410
    },
    {
      "epoch": 13.615579386356867,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.008388148277900916,
      "loss": 0.7846,
      "num_input_tokens_seen": 53076216,
      "step": 91415
    },
    {
      "epoch": 13.616324098897826,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.008386398305446898,
      "loss": 0.803,
      "num_input_tokens_seen": 53079256,
      "step": 91420
    },
    {
      "epoch": 13.617068811438784,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.008384648444722967,
      "loss": 0.8087,
      "num_input_tokens_seen": 53081976,
      "step": 91425
    },
    {
      "epoch": 13.617813523979743,
      "grad_norm": 0.044921875,
      "learning_rate": 0.00838289869575867,
      "loss": 0.8007,
      "num_input_tokens_seen": 53084632,
      "step": 91430
    },
    {
      "epoch": 13.618558236520704,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.008381149058583578,
      "loss": 0.8303,
      "num_input_tokens_seen": 53087672,
      "step": 91435
    },
    {
      "epoch": 13.619302949061662,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.008379399533227246,
      "loss": 0.8005,
      "num_input_tokens_seen": 53090968,
      "step": 91440
    },
    {
      "epoch": 13.620047661602621,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.008377650119719224,
      "loss": 0.7932,
      "num_input_tokens_seen": 53093912,
      "step": 91445
    },
    {
      "epoch": 13.62079237414358,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.008375900818089083,
      "loss": 0.8029,
      "num_input_tokens_seen": 53096728,
      "step": 91450
    },
    {
      "epoch": 13.62153708668454,
      "grad_norm": 0.03125,
      "learning_rate": 0.008374151628366359,
      "loss": 0.812,
      "num_input_tokens_seen": 53099640,
      "step": 91455
    },
    {
      "epoch": 13.6222817992255,
      "grad_norm": 0.03515625,
      "learning_rate": 0.008372402550580613,
      "loss": 0.8099,
      "num_input_tokens_seen": 53102680,
      "step": 91460
    },
    {
      "epoch": 13.623026511766458,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.008370653584761382,
      "loss": 0.79,
      "num_input_tokens_seen": 53105528,
      "step": 91465
    },
    {
      "epoch": 13.623771224307417,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.008368904730938228,
      "loss": 0.7898,
      "num_input_tokens_seen": 53109656,
      "step": 91470
    },
    {
      "epoch": 13.624515936848377,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.008367155989140682,
      "loss": 0.7853,
      "num_input_tokens_seen": 53112440,
      "step": 91475
    },
    {
      "epoch": 13.625260649389336,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.008365407359398301,
      "loss": 0.785,
      "num_input_tokens_seen": 53115160,
      "step": 91480
    },
    {
      "epoch": 13.626005361930295,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00836365884174062,
      "loss": 0.7674,
      "num_input_tokens_seen": 53118136,
      "step": 91485
    },
    {
      "epoch": 13.626750074471254,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.008361910436197172,
      "loss": 0.7968,
      "num_input_tokens_seen": 53120952,
      "step": 91490
    },
    {
      "epoch": 13.627494787012214,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.008360162142797507,
      "loss": 0.7872,
      "num_input_tokens_seen": 53124152,
      "step": 91495
    },
    {
      "epoch": 13.628239499553173,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.008358413961571148,
      "loss": 0.7772,
      "num_input_tokens_seen": 53126904,
      "step": 91500
    },
    {
      "epoch": 13.628984212094132,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.008356665892547637,
      "loss": 0.7852,
      "num_input_tokens_seen": 53129816,
      "step": 91505
    },
    {
      "epoch": 13.62972892463509,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.008354917935756509,
      "loss": 0.7689,
      "num_input_tokens_seen": 53132792,
      "step": 91510
    },
    {
      "epoch": 13.63047363717605,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.008353170091227284,
      "loss": 0.7973,
      "num_input_tokens_seen": 53135640,
      "step": 91515
    },
    {
      "epoch": 13.63121834971701,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.008351422358989493,
      "loss": 0.8207,
      "num_input_tokens_seen": 53138584,
      "step": 91520
    },
    {
      "epoch": 13.631963062257968,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.008349674739072667,
      "loss": 0.7815,
      "num_input_tokens_seen": 53141240,
      "step": 91525
    },
    {
      "epoch": 13.632707774798927,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.00834792723150633,
      "loss": 0.792,
      "num_input_tokens_seen": 53144120,
      "step": 91530
    },
    {
      "epoch": 13.633452487339888,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.008346179836319995,
      "loss": 0.7892,
      "num_input_tokens_seen": 53146808,
      "step": 91535
    },
    {
      "epoch": 13.634197199880846,
      "grad_norm": 0.037109375,
      "learning_rate": 0.008344432553543195,
      "loss": 0.7659,
      "num_input_tokens_seen": 53149720,
      "step": 91540
    },
    {
      "epoch": 13.634941912421805,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.008342685383205433,
      "loss": 0.8016,
      "num_input_tokens_seen": 53153016,
      "step": 91545
    },
    {
      "epoch": 13.635686624962764,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.008340938325336245,
      "loss": 0.7882,
      "num_input_tokens_seen": 53156216,
      "step": 91550
    },
    {
      "epoch": 13.636431337503723,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.00833919137996513,
      "loss": 0.7916,
      "num_input_tokens_seen": 53159160,
      "step": 91555
    },
    {
      "epoch": 13.637176050044683,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00833744454712161,
      "loss": 0.8141,
      "num_input_tokens_seen": 53162136,
      "step": 91560
    },
    {
      "epoch": 13.637920762585642,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.008335697826835194,
      "loss": 0.782,
      "num_input_tokens_seen": 53165112,
      "step": 91565
    },
    {
      "epoch": 13.6386654751266,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.008333951219135384,
      "loss": 0.7586,
      "num_input_tokens_seen": 53167928,
      "step": 91570
    },
    {
      "epoch": 13.63941018766756,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.008332204724051702,
      "loss": 0.773,
      "num_input_tokens_seen": 53171256,
      "step": 91575
    },
    {
      "epoch": 13.64015490020852,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.00833045834161364,
      "loss": 0.8103,
      "num_input_tokens_seen": 53174104,
      "step": 91580
    },
    {
      "epoch": 13.640899612749479,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.008328712071850708,
      "loss": 0.8185,
      "num_input_tokens_seen": 53176920,
      "step": 91585
    },
    {
      "epoch": 13.641644325290438,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.008326965914792398,
      "loss": 0.8043,
      "num_input_tokens_seen": 53179832,
      "step": 91590
    },
    {
      "epoch": 13.642389037831396,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.008325219870468224,
      "loss": 0.8232,
      "num_input_tokens_seen": 53182616,
      "step": 91595
    },
    {
      "epoch": 13.643133750372357,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.008323473938907672,
      "loss": 0.7804,
      "num_input_tokens_seen": 53185624,
      "step": 91600
    },
    {
      "epoch": 13.643878462913316,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.008321728120140245,
      "loss": 0.7834,
      "num_input_tokens_seen": 53188312,
      "step": 91605
    },
    {
      "epoch": 13.644623175454274,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.008319982414195435,
      "loss": 0.7888,
      "num_input_tokens_seen": 53191128,
      "step": 91610
    },
    {
      "epoch": 13.645367887995233,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.008318236821102727,
      "loss": 0.8278,
      "num_input_tokens_seen": 53193976,
      "step": 91615
    },
    {
      "epoch": 13.646112600536194,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.008316491340891623,
      "loss": 0.8033,
      "num_input_tokens_seen": 53196920,
      "step": 91620
    },
    {
      "epoch": 13.646857313077152,
      "grad_norm": 0.046875,
      "learning_rate": 0.008314745973591598,
      "loss": 0.7908,
      "num_input_tokens_seen": 53199608,
      "step": 91625
    },
    {
      "epoch": 13.647602025618111,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.008313000719232152,
      "loss": 0.8593,
      "num_input_tokens_seen": 53202424,
      "step": 91630
    },
    {
      "epoch": 13.64834673815907,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.008311255577842756,
      "loss": 0.7968,
      "num_input_tokens_seen": 53205080,
      "step": 91635
    },
    {
      "epoch": 13.64909145070003,
      "grad_norm": 0.046875,
      "learning_rate": 0.008309510549452906,
      "loss": 0.791,
      "num_input_tokens_seen": 53207960,
      "step": 91640
    },
    {
      "epoch": 13.64983616324099,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.008307765634092074,
      "loss": 0.7938,
      "num_input_tokens_seen": 53210648,
      "step": 91645
    },
    {
      "epoch": 13.650580875781948,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.00830602083178974,
      "loss": 0.7817,
      "num_input_tokens_seen": 53213752,
      "step": 91650
    },
    {
      "epoch": 13.651325588322907,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.008304276142575381,
      "loss": 0.8031,
      "num_input_tokens_seen": 53216728,
      "step": 91655
    },
    {
      "epoch": 13.652070300863867,
      "grad_norm": 0.03125,
      "learning_rate": 0.008302531566478465,
      "loss": 0.7732,
      "num_input_tokens_seen": 53219448,
      "step": 91660
    },
    {
      "epoch": 13.652815013404826,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.00830078710352848,
      "loss": 0.8221,
      "num_input_tokens_seen": 53222392,
      "step": 91665
    },
    {
      "epoch": 13.653559725945785,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.008299042753754877,
      "loss": 0.8026,
      "num_input_tokens_seen": 53225080,
      "step": 91670
    },
    {
      "epoch": 13.654304438486744,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.008297298517187147,
      "loss": 0.7865,
      "num_input_tokens_seen": 53228088,
      "step": 91675
    },
    {
      "epoch": 13.655049151027704,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.008295554393854737,
      "loss": 0.8025,
      "num_input_tokens_seen": 53230936,
      "step": 91680
    },
    {
      "epoch": 13.655793863568663,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.008293810383787127,
      "loss": 0.7988,
      "num_input_tokens_seen": 53234008,
      "step": 91685
    },
    {
      "epoch": 13.656538576109622,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.008292066487013777,
      "loss": 0.8068,
      "num_input_tokens_seen": 53236856,
      "step": 91690
    },
    {
      "epoch": 13.65728328865058,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00829032270356414,
      "loss": 0.7912,
      "num_input_tokens_seen": 53239608,
      "step": 91695
    },
    {
      "epoch": 13.65802800119154,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.008288579033467686,
      "loss": 0.7857,
      "num_input_tokens_seen": 53242488,
      "step": 91700
    },
    {
      "epoch": 13.6587727137325,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.008286835476753863,
      "loss": 0.8051,
      "num_input_tokens_seen": 53245368,
      "step": 91705
    },
    {
      "epoch": 13.659517426273458,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00828509203345214,
      "loss": 0.8202,
      "num_input_tokens_seen": 53248216,
      "step": 91710
    },
    {
      "epoch": 13.660262138814417,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.008283348703591958,
      "loss": 0.7879,
      "num_input_tokens_seen": 53251096,
      "step": 91715
    },
    {
      "epoch": 13.661006851355378,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.008281605487202776,
      "loss": 0.777,
      "num_input_tokens_seen": 53254104,
      "step": 91720
    },
    {
      "epoch": 13.661751563896336,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.008279862384314038,
      "loss": 0.8035,
      "num_input_tokens_seen": 53257048,
      "step": 91725
    },
    {
      "epoch": 13.662496276437295,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.008278119394955192,
      "loss": 0.801,
      "num_input_tokens_seen": 53260280,
      "step": 91730
    },
    {
      "epoch": 13.663240988978254,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.008276376519155691,
      "loss": 0.7967,
      "num_input_tokens_seen": 53263320,
      "step": 91735
    },
    {
      "epoch": 13.663985701519213,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.00827463375694497,
      "loss": 0.7831,
      "num_input_tokens_seen": 53266104,
      "step": 91740
    },
    {
      "epoch": 13.664730414060173,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.008272891108352482,
      "loss": 0.7844,
      "num_input_tokens_seen": 53268920,
      "step": 91745
    },
    {
      "epoch": 13.665475126601132,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.008271148573407659,
      "loss": 0.7879,
      "num_input_tokens_seen": 53271736,
      "step": 91750
    },
    {
      "epoch": 13.66621983914209,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.008269406152139943,
      "loss": 0.8096,
      "num_input_tokens_seen": 53274808,
      "step": 91755
    },
    {
      "epoch": 13.66696455168305,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.008267663844578766,
      "loss": 0.7972,
      "num_input_tokens_seen": 53277592,
      "step": 91760
    },
    {
      "epoch": 13.66770926422401,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.008265921650753572,
      "loss": 0.7898,
      "num_input_tokens_seen": 53280568,
      "step": 91765
    },
    {
      "epoch": 13.668453976764969,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.008264179570693787,
      "loss": 0.8083,
      "num_input_tokens_seen": 53283512,
      "step": 91770
    },
    {
      "epoch": 13.669198689305928,
      "grad_norm": 0.044921875,
      "learning_rate": 0.008262437604428836,
      "loss": 0.7883,
      "num_input_tokens_seen": 53286872,
      "step": 91775
    },
    {
      "epoch": 13.669943401846886,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.008260695751988158,
      "loss": 0.796,
      "num_input_tokens_seen": 53290168,
      "step": 91780
    },
    {
      "epoch": 13.670688114387847,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.008258954013401177,
      "loss": 0.78,
      "num_input_tokens_seen": 53293464,
      "step": 91785
    },
    {
      "epoch": 13.671432826928806,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.008257212388697319,
      "loss": 0.7852,
      "num_input_tokens_seen": 53296344,
      "step": 91790
    },
    {
      "epoch": 13.672177539469764,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.008255470877905997,
      "loss": 0.7931,
      "num_input_tokens_seen": 53299480,
      "step": 91795
    },
    {
      "epoch": 13.672922252010723,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.008253729481056644,
      "loss": 0.8821,
      "num_input_tokens_seen": 53302104,
      "step": 91800
    },
    {
      "epoch": 13.673666964551684,
      "grad_norm": 0.04296875,
      "learning_rate": 0.00825198819817867,
      "loss": 0.8036,
      "num_input_tokens_seen": 53305144,
      "step": 91805
    },
    {
      "epoch": 13.674411677092642,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.008250247029301504,
      "loss": 0.7659,
      "num_input_tokens_seen": 53307864,
      "step": 91810
    },
    {
      "epoch": 13.675156389633601,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.008248505974454555,
      "loss": 0.8119,
      "num_input_tokens_seen": 53310616,
      "step": 91815
    },
    {
      "epoch": 13.67590110217456,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.008246765033667227,
      "loss": 0.8057,
      "num_input_tokens_seen": 53313432,
      "step": 91820
    },
    {
      "epoch": 13.67664581471552,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.008245024206968947,
      "loss": 0.8033,
      "num_input_tokens_seen": 53316056,
      "step": 91825
    },
    {
      "epoch": 13.67739052725648,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.008243283494389114,
      "loss": 0.7797,
      "num_input_tokens_seen": 53318680,
      "step": 91830
    },
    {
      "epoch": 13.678135239797438,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.008241542895957141,
      "loss": 0.7929,
      "num_input_tokens_seen": 53321688,
      "step": 91835
    },
    {
      "epoch": 13.678879952338397,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.008239802411702435,
      "loss": 0.7917,
      "num_input_tokens_seen": 53324152,
      "step": 91840
    },
    {
      "epoch": 13.679624664879357,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.008238062041654396,
      "loss": 0.7903,
      "num_input_tokens_seen": 53327096,
      "step": 91845
    },
    {
      "epoch": 13.680369377420316,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.008236321785842418,
      "loss": 0.7961,
      "num_input_tokens_seen": 53329912,
      "step": 91850
    },
    {
      "epoch": 13.681114089961275,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.008234581644295918,
      "loss": 0.7972,
      "num_input_tokens_seen": 53332632,
      "step": 91855
    },
    {
      "epoch": 13.681858802502234,
      "grad_norm": 0.0224609375,
      "learning_rate": 0.008232841617044285,
      "loss": 0.7927,
      "num_input_tokens_seen": 53335544,
      "step": 91860
    },
    {
      "epoch": 13.682603515043194,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.008231101704116908,
      "loss": 0.7951,
      "num_input_tokens_seen": 53338456,
      "step": 91865
    },
    {
      "epoch": 13.683348227584153,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.008229361905543194,
      "loss": 0.8003,
      "num_input_tokens_seen": 53341464,
      "step": 91870
    },
    {
      "epoch": 13.684092940125112,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.008227622221352522,
      "loss": 0.7916,
      "num_input_tokens_seen": 53344568,
      "step": 91875
    },
    {
      "epoch": 13.68483765266607,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.008225882651574297,
      "loss": 0.7988,
      "num_input_tokens_seen": 53347096,
      "step": 91880
    },
    {
      "epoch": 13.68558236520703,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.008224143196237894,
      "loss": 0.7955,
      "num_input_tokens_seen": 53350008,
      "step": 91885
    },
    {
      "epoch": 13.68632707774799,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00822240385537271,
      "loss": 0.8002,
      "num_input_tokens_seen": 53353208,
      "step": 91890
    },
    {
      "epoch": 13.687071790288948,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.008220664629008124,
      "loss": 0.8181,
      "num_input_tokens_seen": 53356120,
      "step": 91895
    },
    {
      "epoch": 13.687816502829907,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.008218925517173514,
      "loss": 0.8157,
      "num_input_tokens_seen": 53358840,
      "step": 91900
    },
    {
      "epoch": 13.688561215370868,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.00821718651989827,
      "loss": 0.7869,
      "num_input_tokens_seen": 53361880,
      "step": 91905
    },
    {
      "epoch": 13.689305927911827,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.008215447637211768,
      "loss": 0.8012,
      "num_input_tokens_seen": 53365016,
      "step": 91910
    },
    {
      "epoch": 13.690050640452785,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.008213708869143382,
      "loss": 0.8063,
      "num_input_tokens_seen": 53367864,
      "step": 91915
    },
    {
      "epoch": 13.690795352993744,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.008211970215722482,
      "loss": 0.7829,
      "num_input_tokens_seen": 53370776,
      "step": 91920
    },
    {
      "epoch": 13.691540065534703,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.008210231676978449,
      "loss": 0.7992,
      "num_input_tokens_seen": 53373848,
      "step": 91925
    },
    {
      "epoch": 13.692284778075663,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.008208493252940655,
      "loss": 0.8094,
      "num_input_tokens_seen": 53376856,
      "step": 91930
    },
    {
      "epoch": 13.693029490616622,
      "grad_norm": 0.0546875,
      "learning_rate": 0.008206754943638457,
      "loss": 0.7763,
      "num_input_tokens_seen": 53379928,
      "step": 91935
    },
    {
      "epoch": 13.69377420315758,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.008205016749101237,
      "loss": 0.806,
      "num_input_tokens_seen": 53383032,
      "step": 91940
    },
    {
      "epoch": 13.69451891569854,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.008203278669358344,
      "loss": 0.7505,
      "num_input_tokens_seen": 53385784,
      "step": 91945
    },
    {
      "epoch": 13.6952636282395,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.00820154070443916,
      "loss": 0.7881,
      "num_input_tokens_seen": 53388856,
      "step": 91950
    },
    {
      "epoch": 13.696008340780459,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.00819980285437303,
      "loss": 0.7945,
      "num_input_tokens_seen": 53391672,
      "step": 91955
    },
    {
      "epoch": 13.696753053321418,
      "grad_norm": 0.0203857421875,
      "learning_rate": 0.008198065119189323,
      "loss": 0.7982,
      "num_input_tokens_seen": 53394392,
      "step": 91960
    },
    {
      "epoch": 13.697497765862376,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.008196327498917385,
      "loss": 0.793,
      "num_input_tokens_seen": 53397336,
      "step": 91965
    },
    {
      "epoch": 13.698242478403337,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.008194589993586588,
      "loss": 0.7963,
      "num_input_tokens_seen": 53400280,
      "step": 91970
    },
    {
      "epoch": 13.698987190944296,
      "grad_norm": 0.0198974609375,
      "learning_rate": 0.008192852603226275,
      "loss": 0.799,
      "num_input_tokens_seen": 53403224,
      "step": 91975
    },
    {
      "epoch": 13.699731903485254,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.008191115327865801,
      "loss": 0.7939,
      "num_input_tokens_seen": 53406168,
      "step": 91980
    },
    {
      "epoch": 13.700476616026213,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.008189378167534511,
      "loss": 0.8009,
      "num_input_tokens_seen": 53409528,
      "step": 91985
    },
    {
      "epoch": 13.701221328567174,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.008187641122261748,
      "loss": 0.8016,
      "num_input_tokens_seen": 53412504,
      "step": 91990
    },
    {
      "epoch": 13.701966041108133,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.008185904192076873,
      "loss": 0.7616,
      "num_input_tokens_seen": 53415416,
      "step": 91995
    },
    {
      "epoch": 13.702710753649091,
      "grad_norm": 0.064453125,
      "learning_rate": 0.008184167377009213,
      "loss": 0.8036,
      "num_input_tokens_seen": 53418232,
      "step": 92000
    },
    {
      "epoch": 13.70345546619005,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.008182430677088126,
      "loss": 0.7938,
      "num_input_tokens_seen": 53421112,
      "step": 92005
    },
    {
      "epoch": 13.70420017873101,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.008180694092342943,
      "loss": 0.7932,
      "num_input_tokens_seen": 53424120,
      "step": 92010
    },
    {
      "epoch": 13.70494489127197,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.008178957622802996,
      "loss": 0.7752,
      "num_input_tokens_seen": 53426968,
      "step": 92015
    },
    {
      "epoch": 13.705689603812928,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.008177221268497632,
      "loss": 0.7811,
      "num_input_tokens_seen": 53429656,
      "step": 92020
    },
    {
      "epoch": 13.706434316353887,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.008175485029456177,
      "loss": 0.7931,
      "num_input_tokens_seen": 53432696,
      "step": 92025
    },
    {
      "epoch": 13.707179028894847,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.00817374890570797,
      "loss": 0.8222,
      "num_input_tokens_seen": 53435512,
      "step": 92030
    },
    {
      "epoch": 13.707923741435806,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.008172012897282333,
      "loss": 0.786,
      "num_input_tokens_seen": 53438392,
      "step": 92035
    },
    {
      "epoch": 13.708668453976765,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.008170277004208604,
      "loss": 0.7801,
      "num_input_tokens_seen": 53441176,
      "step": 92040
    },
    {
      "epoch": 13.709413166517724,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.008168541226516104,
      "loss": 0.833,
      "num_input_tokens_seen": 53444088,
      "step": 92045
    },
    {
      "epoch": 13.710157879058684,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.008166805564234155,
      "loss": 0.8124,
      "num_input_tokens_seen": 53447096,
      "step": 92050
    },
    {
      "epoch": 13.710902591599643,
      "grad_norm": 0.072265625,
      "learning_rate": 0.008165070017392083,
      "loss": 0.8171,
      "num_input_tokens_seen": 53449912,
      "step": 92055
    },
    {
      "epoch": 13.711647304140602,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0081633345860192,
      "loss": 0.7799,
      "num_input_tokens_seen": 53452920,
      "step": 92060
    },
    {
      "epoch": 13.71239201668156,
      "grad_norm": 0.064453125,
      "learning_rate": 0.008161599270144834,
      "loss": 0.7955,
      "num_input_tokens_seen": 53455704,
      "step": 92065
    },
    {
      "epoch": 13.71313672922252,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.008159864069798295,
      "loss": 0.7841,
      "num_input_tokens_seen": 53458584,
      "step": 92070
    },
    {
      "epoch": 13.71388144176348,
      "grad_norm": 0.044921875,
      "learning_rate": 0.008158128985008908,
      "loss": 0.7979,
      "num_input_tokens_seen": 53461784,
      "step": 92075
    },
    {
      "epoch": 13.714626154304439,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.00815639401580597,
      "loss": 0.7955,
      "num_input_tokens_seen": 53464376,
      "step": 92080
    },
    {
      "epoch": 13.715370866845397,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.008154659162218804,
      "loss": 0.8116,
      "num_input_tokens_seen": 53467544,
      "step": 92085
    },
    {
      "epoch": 13.716115579386356,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.008152924424276715,
      "loss": 0.7918,
      "num_input_tokens_seen": 53470296,
      "step": 92090
    },
    {
      "epoch": 13.716860291927317,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.008151189802009004,
      "loss": 0.8277,
      "num_input_tokens_seen": 53473432,
      "step": 92095
    },
    {
      "epoch": 13.717605004468275,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.008149455295444988,
      "loss": 0.8028,
      "num_input_tokens_seen": 53476600,
      "step": 92100
    },
    {
      "epoch": 13.718349717009234,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.008147720904613952,
      "loss": 0.7703,
      "num_input_tokens_seen": 53479448,
      "step": 92105
    },
    {
      "epoch": 13.719094429550193,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.008145986629545215,
      "loss": 0.8438,
      "num_input_tokens_seen": 53482136,
      "step": 92110
    },
    {
      "epoch": 13.719839142091153,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.008144252470268067,
      "loss": 0.7732,
      "num_input_tokens_seen": 53484984,
      "step": 92115
    },
    {
      "epoch": 13.720583854632112,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00814251842681181,
      "loss": 0.8438,
      "num_input_tokens_seen": 53487800,
      "step": 92120
    },
    {
      "epoch": 13.721328567173071,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.008140784499205724,
      "loss": 0.8073,
      "num_input_tokens_seen": 53490648,
      "step": 92125
    },
    {
      "epoch": 13.72207327971403,
      "grad_norm": 0.052734375,
      "learning_rate": 0.008139050687479119,
      "loss": 0.7998,
      "num_input_tokens_seen": 53493304,
      "step": 92130
    },
    {
      "epoch": 13.72281799225499,
      "grad_norm": 0.041015625,
      "learning_rate": 0.00813731699166128,
      "loss": 0.781,
      "num_input_tokens_seen": 53496248,
      "step": 92135
    },
    {
      "epoch": 13.723562704795949,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.008135583411781487,
      "loss": 0.8017,
      "num_input_tokens_seen": 53499256,
      "step": 92140
    },
    {
      "epoch": 13.724307417336908,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.008133849947869046,
      "loss": 0.7659,
      "num_input_tokens_seen": 53502072,
      "step": 92145
    },
    {
      "epoch": 13.725052129877866,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.008132116599953224,
      "loss": 0.8005,
      "num_input_tokens_seen": 53505176,
      "step": 92150
    },
    {
      "epoch": 13.725796842418827,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.008130383368063319,
      "loss": 0.7954,
      "num_input_tokens_seen": 53508056,
      "step": 92155
    },
    {
      "epoch": 13.726541554959786,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.008128650252228597,
      "loss": 0.8023,
      "num_input_tokens_seen": 53510808,
      "step": 92160
    },
    {
      "epoch": 13.727286267500745,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.008126917252478353,
      "loss": 0.7833,
      "num_input_tokens_seen": 53513784,
      "step": 92165
    },
    {
      "epoch": 13.728030980041703,
      "grad_norm": 0.0390625,
      "learning_rate": 0.008125184368841853,
      "loss": 0.8044,
      "num_input_tokens_seen": 53516472,
      "step": 92170
    },
    {
      "epoch": 13.728775692582664,
      "grad_norm": 0.033203125,
      "learning_rate": 0.00812345160134838,
      "loss": 0.7771,
      "num_input_tokens_seen": 53519288,
      "step": 92175
    },
    {
      "epoch": 13.729520405123623,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.008121718950027196,
      "loss": 0.7901,
      "num_input_tokens_seen": 53522136,
      "step": 92180
    },
    {
      "epoch": 13.730265117664581,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.008119986414907583,
      "loss": 0.8097,
      "num_input_tokens_seen": 53525272,
      "step": 92185
    },
    {
      "epoch": 13.73100983020554,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.00811825399601881,
      "loss": 0.7845,
      "num_input_tokens_seen": 53527928,
      "step": 92190
    },
    {
      "epoch": 13.7317545427465,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.008116521693390135,
      "loss": 0.7679,
      "num_input_tokens_seen": 53530712,
      "step": 92195
    },
    {
      "epoch": 13.73249925528746,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.008114789507050834,
      "loss": 0.8268,
      "num_input_tokens_seen": 53533592,
      "step": 92200
    },
    {
      "epoch": 13.733243967828418,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.008113057437030162,
      "loss": 0.8609,
      "num_input_tokens_seen": 53536568,
      "step": 92205
    },
    {
      "epoch": 13.733988680369377,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.008111325483357391,
      "loss": 0.7982,
      "num_input_tokens_seen": 53539320,
      "step": 92210
    },
    {
      "epoch": 13.734733392910336,
      "grad_norm": 0.03125,
      "learning_rate": 0.008109593646061774,
      "loss": 0.8068,
      "num_input_tokens_seen": 53542200,
      "step": 92215
    },
    {
      "epoch": 13.735478105451296,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.008107861925172564,
      "loss": 0.823,
      "num_input_tokens_seen": 53544952,
      "step": 92220
    },
    {
      "epoch": 13.736222817992255,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.008106130320719029,
      "loss": 0.7975,
      "num_input_tokens_seen": 53547576,
      "step": 92225
    },
    {
      "epoch": 13.736967530533214,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.008104398832730408,
      "loss": 0.8071,
      "num_input_tokens_seen": 53550520,
      "step": 92230
    },
    {
      "epoch": 13.737712243074174,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.008102667461235967,
      "loss": 0.8217,
      "num_input_tokens_seen": 53553560,
      "step": 92235
    },
    {
      "epoch": 13.738456955615133,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.008100936206264949,
      "loss": 0.7881,
      "num_input_tokens_seen": 53556312,
      "step": 92240
    },
    {
      "epoch": 13.739201668156092,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.008099205067846602,
      "loss": 0.798,
      "num_input_tokens_seen": 53559576,
      "step": 92245
    },
    {
      "epoch": 13.73994638069705,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.00809747404601017,
      "loss": 0.8057,
      "num_input_tokens_seen": 53562616,
      "step": 92250
    },
    {
      "epoch": 13.74069109323801,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.008095743140784895,
      "loss": 0.7975,
      "num_input_tokens_seen": 53565624,
      "step": 92255
    },
    {
      "epoch": 13.74143580577897,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.008094012352200027,
      "loss": 0.8119,
      "num_input_tokens_seen": 53568728,
      "step": 92260
    },
    {
      "epoch": 13.742180518319929,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.008092281680284795,
      "loss": 0.8077,
      "num_input_tokens_seen": 53571992,
      "step": 92265
    },
    {
      "epoch": 13.742925230860887,
      "grad_norm": 0.044921875,
      "learning_rate": 0.00809055112506845,
      "loss": 0.8036,
      "num_input_tokens_seen": 53574968,
      "step": 92270
    },
    {
      "epoch": 13.743669943401846,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.008088820686580216,
      "loss": 0.794,
      "num_input_tokens_seen": 53577752,
      "step": 92275
    },
    {
      "epoch": 13.744414655942807,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.008087090364849338,
      "loss": 0.7834,
      "num_input_tokens_seen": 53581016,
      "step": 92280
    },
    {
      "epoch": 13.745159368483765,
      "grad_norm": 0.041015625,
      "learning_rate": 0.008085360159905037,
      "loss": 0.7892,
      "num_input_tokens_seen": 53583832,
      "step": 92285
    },
    {
      "epoch": 13.745904081024724,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.008083630071776551,
      "loss": 0.8111,
      "num_input_tokens_seen": 53586712,
      "step": 92290
    },
    {
      "epoch": 13.746648793565683,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.008081900100493108,
      "loss": 0.8009,
      "num_input_tokens_seen": 53589400,
      "step": 92295
    },
    {
      "epoch": 13.747393506106643,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.008080170246083927,
      "loss": 0.8015,
      "num_input_tokens_seen": 53592088,
      "step": 92300
    },
    {
      "epoch": 13.748138218647602,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.008078440508578241,
      "loss": 0.8161,
      "num_input_tokens_seen": 53594904,
      "step": 92305
    },
    {
      "epoch": 13.748882931188561,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.008076710888005267,
      "loss": 0.8068,
      "num_input_tokens_seen": 53597784,
      "step": 92310
    },
    {
      "epoch": 13.74962764372952,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.008074981384394228,
      "loss": 0.812,
      "num_input_tokens_seen": 53600664,
      "step": 92315
    },
    {
      "epoch": 13.75037235627048,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.008073251997774334,
      "loss": 0.8017,
      "num_input_tokens_seen": 53603800,
      "step": 92320
    },
    {
      "epoch": 13.751117068811439,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.008071522728174811,
      "loss": 0.7806,
      "num_input_tokens_seen": 53606840,
      "step": 92325
    },
    {
      "epoch": 13.751861781352398,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.008069793575624873,
      "loss": 0.7973,
      "num_input_tokens_seen": 53609784,
      "step": 92330
    },
    {
      "epoch": 13.752606493893357,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.008068064540153722,
      "loss": 0.7812,
      "num_input_tokens_seen": 53612856,
      "step": 92335
    },
    {
      "epoch": 13.753351206434317,
      "grad_norm": 0.033203125,
      "learning_rate": 0.00806633562179058,
      "loss": 0.7929,
      "num_input_tokens_seen": 53615512,
      "step": 92340
    },
    {
      "epoch": 13.754095918975276,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.008064606820564644,
      "loss": 0.7977,
      "num_input_tokens_seen": 53618360,
      "step": 92345
    },
    {
      "epoch": 13.754840631516235,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.008062878136505136,
      "loss": 0.8055,
      "num_input_tokens_seen": 53621464,
      "step": 92350
    },
    {
      "epoch": 13.755585344057193,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.008061149569641244,
      "loss": 0.8028,
      "num_input_tokens_seen": 53624344,
      "step": 92355
    },
    {
      "epoch": 13.756330056598154,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00805942112000218,
      "loss": 0.8163,
      "num_input_tokens_seen": 53627256,
      "step": 92360
    },
    {
      "epoch": 13.757074769139113,
      "grad_norm": 0.130859375,
      "learning_rate": 0.008057692787617142,
      "loss": 0.86,
      "num_input_tokens_seen": 53630168,
      "step": 92365
    },
    {
      "epoch": 13.757819481680071,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.008055964572515333,
      "loss": 0.784,
      "num_input_tokens_seen": 53632888,
      "step": 92370
    },
    {
      "epoch": 13.75856419422103,
      "grad_norm": 0.033203125,
      "learning_rate": 0.008054236474725944,
      "loss": 0.7863,
      "num_input_tokens_seen": 53635480,
      "step": 92375
    },
    {
      "epoch": 13.75930890676199,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.008052508494278169,
      "loss": 0.7992,
      "num_input_tokens_seen": 53638776,
      "step": 92380
    },
    {
      "epoch": 13.76005361930295,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.008050780631201203,
      "loss": 0.797,
      "num_input_tokens_seen": 53641592,
      "step": 92385
    },
    {
      "epoch": 13.760798331843908,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00804905288552423,
      "loss": 0.8016,
      "num_input_tokens_seen": 53645048,
      "step": 92390
    },
    {
      "epoch": 13.761543044384867,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.008047325257276448,
      "loss": 0.7918,
      "num_input_tokens_seen": 53647928,
      "step": 92395
    },
    {
      "epoch": 13.762287756925826,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.008045597746487033,
      "loss": 0.7965,
      "num_input_tokens_seen": 53650776,
      "step": 92400
    },
    {
      "epoch": 13.763032469466786,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.008043870353185186,
      "loss": 0.8306,
      "num_input_tokens_seen": 53653720,
      "step": 92405
    },
    {
      "epoch": 13.763777182007745,
      "grad_norm": 0.0189208984375,
      "learning_rate": 0.00804214307740007,
      "loss": 0.7874,
      "num_input_tokens_seen": 53656600,
      "step": 92410
    },
    {
      "epoch": 13.764521894548704,
      "grad_norm": 0.05078125,
      "learning_rate": 0.008040415919160881,
      "loss": 0.8045,
      "num_input_tokens_seen": 53659416,
      "step": 92415
    },
    {
      "epoch": 13.765266607089664,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.008038688878496795,
      "loss": 0.8036,
      "num_input_tokens_seen": 53662168,
      "step": 92420
    },
    {
      "epoch": 13.766011319630623,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.008036961955436977,
      "loss": 0.8099,
      "num_input_tokens_seen": 53665048,
      "step": 92425
    },
    {
      "epoch": 13.766756032171582,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.008035235150010617,
      "loss": 0.783,
      "num_input_tokens_seen": 53667928,
      "step": 92430
    },
    {
      "epoch": 13.76750074471254,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.008033508462246873,
      "loss": 0.7842,
      "num_input_tokens_seen": 53671256,
      "step": 92435
    },
    {
      "epoch": 13.7682454572535,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.00803178189217493,
      "loss": 0.791,
      "num_input_tokens_seen": 53673912,
      "step": 92440
    },
    {
      "epoch": 13.76899016979446,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.008030055439823948,
      "loss": 0.7992,
      "num_input_tokens_seen": 53676632,
      "step": 92445
    },
    {
      "epoch": 13.769734882335419,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.008028329105223097,
      "loss": 0.7916,
      "num_input_tokens_seen": 53679256,
      "step": 92450
    },
    {
      "epoch": 13.770479594876377,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00802660288840154,
      "loss": 0.7877,
      "num_input_tokens_seen": 53682168,
      "step": 92455
    },
    {
      "epoch": 13.771224307417336,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.008024876789388432,
      "loss": 0.8021,
      "num_input_tokens_seen": 53684952,
      "step": 92460
    },
    {
      "epoch": 13.771969019958297,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00802315080821295,
      "loss": 0.808,
      "num_input_tokens_seen": 53687800,
      "step": 92465
    },
    {
      "epoch": 13.772713732499255,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.008021424944904237,
      "loss": 0.7905,
      "num_input_tokens_seen": 53690584,
      "step": 92470
    },
    {
      "epoch": 13.773458445040214,
      "grad_norm": 0.046875,
      "learning_rate": 0.008019699199491463,
      "loss": 0.7868,
      "num_input_tokens_seen": 53693496,
      "step": 92475
    },
    {
      "epoch": 13.774203157581173,
      "grad_norm": 0.0218505859375,
      "learning_rate": 0.008017973572003772,
      "loss": 0.7936,
      "num_input_tokens_seen": 53696216,
      "step": 92480
    },
    {
      "epoch": 13.774947870122134,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.008016248062470324,
      "loss": 0.8131,
      "num_input_tokens_seen": 53698872,
      "step": 92485
    },
    {
      "epoch": 13.775692582663092,
      "grad_norm": 0.037109375,
      "learning_rate": 0.008014522670920262,
      "loss": 0.7836,
      "num_input_tokens_seen": 53701688,
      "step": 92490
    },
    {
      "epoch": 13.776437295204051,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.008012797397382747,
      "loss": 0.7926,
      "num_input_tokens_seen": 53704664,
      "step": 92495
    },
    {
      "epoch": 13.77718200774501,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.00801107224188692,
      "loss": 0.7986,
      "num_input_tokens_seen": 53707480,
      "step": 92500
    },
    {
      "epoch": 13.77792672028597,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.008009347204461921,
      "loss": 0.7866,
      "num_input_tokens_seen": 53710296,
      "step": 92505
    },
    {
      "epoch": 13.778671432826929,
      "grad_norm": 0.037109375,
      "learning_rate": 0.008007622285136897,
      "loss": 0.7833,
      "num_input_tokens_seen": 53713400,
      "step": 92510
    },
    {
      "epoch": 13.779416145367888,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.008005897483940982,
      "loss": 0.782,
      "num_input_tokens_seen": 53716120,
      "step": 92515
    },
    {
      "epoch": 13.780160857908847,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.008004172800903327,
      "loss": 0.7885,
      "num_input_tokens_seen": 53718904,
      "step": 92520
    },
    {
      "epoch": 13.780905570449807,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.008002448236053056,
      "loss": 0.8015,
      "num_input_tokens_seen": 53721784,
      "step": 92525
    },
    {
      "epoch": 13.781650282990766,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.008000723789419316,
      "loss": 0.8199,
      "num_input_tokens_seen": 53724760,
      "step": 92530
    },
    {
      "epoch": 13.782394995531725,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.007998999461031232,
      "loss": 0.7912,
      "num_input_tokens_seen": 53727672,
      "step": 92535
    },
    {
      "epoch": 13.783139708072683,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.007997275250917934,
      "loss": 0.8037,
      "num_input_tokens_seen": 53730360,
      "step": 92540
    },
    {
      "epoch": 13.783884420613644,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.007995551159108558,
      "loss": 0.7858,
      "num_input_tokens_seen": 53733048,
      "step": 92545
    },
    {
      "epoch": 13.784629133154603,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.00799382718563222,
      "loss": 0.7975,
      "num_input_tokens_seen": 53735896,
      "step": 92550
    },
    {
      "epoch": 13.785373845695561,
      "grad_norm": 0.04296875,
      "learning_rate": 0.007992103330518056,
      "loss": 0.7822,
      "num_input_tokens_seen": 53738808,
      "step": 92555
    },
    {
      "epoch": 13.78611855823652,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.007990379593795179,
      "loss": 0.7844,
      "num_input_tokens_seen": 53741688,
      "step": 92560
    },
    {
      "epoch": 13.78686327077748,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.007988655975492719,
      "loss": 0.7922,
      "num_input_tokens_seen": 53744280,
      "step": 92565
    },
    {
      "epoch": 13.78760798331844,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.007986932475639794,
      "loss": 0.8178,
      "num_input_tokens_seen": 53747000,
      "step": 92570
    },
    {
      "epoch": 13.788352695859398,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.007985209094265512,
      "loss": 0.7881,
      "num_input_tokens_seen": 53749976,
      "step": 92575
    },
    {
      "epoch": 13.789097408400357,
      "grad_norm": 0.020263671875,
      "learning_rate": 0.007983485831398993,
      "loss": 0.7736,
      "num_input_tokens_seen": 53752760,
      "step": 92580
    },
    {
      "epoch": 13.789842120941316,
      "grad_norm": 0.021728515625,
      "learning_rate": 0.00798176268706935,
      "loss": 0.7952,
      "num_input_tokens_seen": 53755704,
      "step": 92585
    },
    {
      "epoch": 13.790586833482276,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.007980039661305694,
      "loss": 0.7975,
      "num_input_tokens_seen": 53758840,
      "step": 92590
    },
    {
      "epoch": 13.791331546023235,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.007978316754137127,
      "loss": 0.7682,
      "num_input_tokens_seen": 53761624,
      "step": 92595
    },
    {
      "epoch": 13.792076258564194,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.007976593965592771,
      "loss": 0.7946,
      "num_input_tokens_seen": 53764504,
      "step": 92600
    },
    {
      "epoch": 13.792820971105153,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.007974871295701711,
      "loss": 0.8049,
      "num_input_tokens_seen": 53767224,
      "step": 92605
    },
    {
      "epoch": 13.793565683646113,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.00797314874449307,
      "loss": 0.7943,
      "num_input_tokens_seen": 53770104,
      "step": 92610
    },
    {
      "epoch": 13.794310396187072,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.00797142631199594,
      "loss": 0.8043,
      "num_input_tokens_seen": 53772728,
      "step": 92615
    },
    {
      "epoch": 13.79505510872803,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.007969703998239413,
      "loss": 0.8095,
      "num_input_tokens_seen": 53776024,
      "step": 92620
    },
    {
      "epoch": 13.79579982126899,
      "grad_norm": 0.044921875,
      "learning_rate": 0.007967981803252595,
      "loss": 0.8051,
      "num_input_tokens_seen": 53779064,
      "step": 92625
    },
    {
      "epoch": 13.79654453380995,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.007966259727064572,
      "loss": 0.804,
      "num_input_tokens_seen": 53782200,
      "step": 92630
    },
    {
      "epoch": 13.797289246350909,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.007964537769704448,
      "loss": 0.8027,
      "num_input_tokens_seen": 53785464,
      "step": 92635
    },
    {
      "epoch": 13.798033958891867,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.007962815931201309,
      "loss": 0.7832,
      "num_input_tokens_seen": 53788536,
      "step": 92640
    },
    {
      "epoch": 13.798778671432826,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.007961094211584243,
      "loss": 0.7847,
      "num_input_tokens_seen": 53791512,
      "step": 92645
    },
    {
      "epoch": 13.799523383973787,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00795937261088233,
      "loss": 0.7856,
      "num_input_tokens_seen": 53794264,
      "step": 92650
    },
    {
      "epoch": 13.800268096514746,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.007957651129124665,
      "loss": 0.8088,
      "num_input_tokens_seen": 53797176,
      "step": 92655
    },
    {
      "epoch": 13.801012809055704,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.007955929766340328,
      "loss": 0.7857,
      "num_input_tokens_seen": 53799928,
      "step": 92660
    },
    {
      "epoch": 13.801757521596663,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.007954208522558393,
      "loss": 0.8117,
      "num_input_tokens_seen": 53803096,
      "step": 92665
    },
    {
      "epoch": 13.802502234137624,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.00795248739780795,
      "loss": 0.8159,
      "num_input_tokens_seen": 53805880,
      "step": 92670
    },
    {
      "epoch": 13.803246946678582,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.007950766392118063,
      "loss": 0.762,
      "num_input_tokens_seen": 53808824,
      "step": 92675
    },
    {
      "epoch": 13.803991659219541,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.007949045505517819,
      "loss": 0.8401,
      "num_input_tokens_seen": 53811640,
      "step": 92680
    },
    {
      "epoch": 13.8047363717605,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.007947324738036278,
      "loss": 0.8193,
      "num_input_tokens_seen": 53814616,
      "step": 92685
    },
    {
      "epoch": 13.80548108430146,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.007945604089702527,
      "loss": 0.7966,
      "num_input_tokens_seen": 53817624,
      "step": 92690
    },
    {
      "epoch": 13.80622579684242,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.007943883560545623,
      "loss": 0.789,
      "num_input_tokens_seen": 53820440,
      "step": 92695
    },
    {
      "epoch": 13.806970509383378,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.007942163150594629,
      "loss": 0.7767,
      "num_input_tokens_seen": 53823288,
      "step": 92700
    },
    {
      "epoch": 13.807715221924337,
      "grad_norm": 0.033203125,
      "learning_rate": 0.00794044285987862,
      "loss": 0.803,
      "num_input_tokens_seen": 53826424,
      "step": 92705
    },
    {
      "epoch": 13.808459934465297,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.007938722688426653,
      "loss": 0.816,
      "num_input_tokens_seen": 53829208,
      "step": 92710
    },
    {
      "epoch": 13.809204647006256,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.007937002636267792,
      "loss": 0.8277,
      "num_input_tokens_seen": 53831864,
      "step": 92715
    },
    {
      "epoch": 13.809949359547215,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.007935282703431086,
      "loss": 0.7954,
      "num_input_tokens_seen": 53835096,
      "step": 92720
    },
    {
      "epoch": 13.810694072088173,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.007933562889945607,
      "loss": 0.7852,
      "num_input_tokens_seen": 53838104,
      "step": 92725
    },
    {
      "epoch": 13.811438784629132,
      "grad_norm": 0.044921875,
      "learning_rate": 0.007931843195840392,
      "loss": 0.8038,
      "num_input_tokens_seen": 53840728,
      "step": 92730
    },
    {
      "epoch": 13.812183497170093,
      "grad_norm": 0.08984375,
      "learning_rate": 0.007930123621144508,
      "loss": 0.8094,
      "num_input_tokens_seen": 53843512,
      "step": 92735
    },
    {
      "epoch": 13.812928209711052,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.007928404165887,
      "loss": 0.7864,
      "num_input_tokens_seen": 53846648,
      "step": 92740
    },
    {
      "epoch": 13.81367292225201,
      "grad_norm": 0.044921875,
      "learning_rate": 0.007926684830096913,
      "loss": 0.7847,
      "num_input_tokens_seen": 53849496,
      "step": 92745
    },
    {
      "epoch": 13.81441763479297,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.007924965613803299,
      "loss": 0.7726,
      "num_input_tokens_seen": 53852216,
      "step": 92750
    },
    {
      "epoch": 13.81516234733393,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.007923246517035198,
      "loss": 0.8147,
      "num_input_tokens_seen": 53855160,
      "step": 92755
    },
    {
      "epoch": 13.815907059874888,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.007921527539821659,
      "loss": 0.7807,
      "num_input_tokens_seen": 53857784,
      "step": 92760
    },
    {
      "epoch": 13.816651772415847,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.007919808682191712,
      "loss": 0.8031,
      "num_input_tokens_seen": 53860760,
      "step": 92765
    },
    {
      "epoch": 13.817396484956806,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.007918089944174407,
      "loss": 0.7836,
      "num_input_tokens_seen": 53864088,
      "step": 92770
    },
    {
      "epoch": 13.818141197497766,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.007916371325798775,
      "loss": 0.811,
      "num_input_tokens_seen": 53867096,
      "step": 92775
    },
    {
      "epoch": 13.818885910038725,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.007914652827093851,
      "loss": 0.7944,
      "num_input_tokens_seen": 53869752,
      "step": 92780
    },
    {
      "epoch": 13.819630622579684,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.007912934448088665,
      "loss": 0.7967,
      "num_input_tokens_seen": 53872760,
      "step": 92785
    },
    {
      "epoch": 13.820375335120643,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.007911216188812242,
      "loss": 0.8281,
      "num_input_tokens_seen": 53875544,
      "step": 92790
    },
    {
      "epoch": 13.821120047661603,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.007909498049293624,
      "loss": 0.7798,
      "num_input_tokens_seen": 53878424,
      "step": 92795
    },
    {
      "epoch": 13.821864760202562,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.007907780029561823,
      "loss": 0.8099,
      "num_input_tokens_seen": 53881240,
      "step": 92800
    },
    {
      "epoch": 13.82260947274352,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.007906062129645876,
      "loss": 0.799,
      "num_input_tokens_seen": 53884216,
      "step": 92805
    },
    {
      "epoch": 13.82335418528448,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.007904344349574795,
      "loss": 0.7987,
      "num_input_tokens_seen": 53887160,
      "step": 92810
    },
    {
      "epoch": 13.82409889782544,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.007902626689377608,
      "loss": 0.8381,
      "num_input_tokens_seen": 53890680,
      "step": 92815
    },
    {
      "epoch": 13.824843610366399,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.007900909149083332,
      "loss": 0.7872,
      "num_input_tokens_seen": 53893688,
      "step": 92820
    },
    {
      "epoch": 13.825588322907358,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.007899191728720974,
      "loss": 0.7848,
      "num_input_tokens_seen": 53896920,
      "step": 92825
    },
    {
      "epoch": 13.826333035448316,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00789747442831956,
      "loss": 0.7881,
      "num_input_tokens_seen": 53899896,
      "step": 92830
    },
    {
      "epoch": 13.827077747989277,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.007895757247908098,
      "loss": 0.8071,
      "num_input_tokens_seen": 53902744,
      "step": 92835
    },
    {
      "epoch": 13.827822460530236,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.007894040187515597,
      "loss": 0.7847,
      "num_input_tokens_seen": 53905784,
      "step": 92840
    },
    {
      "epoch": 13.828567173071194,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.007892323247171056,
      "loss": 0.8079,
      "num_input_tokens_seen": 53908760,
      "step": 92845
    },
    {
      "epoch": 13.829311885612153,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0078906064269035,
      "loss": 0.79,
      "num_input_tokens_seen": 53911736,
      "step": 92850
    },
    {
      "epoch": 13.830056598153114,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00788888972674192,
      "loss": 0.7809,
      "num_input_tokens_seen": 53914552,
      "step": 92855
    },
    {
      "epoch": 13.830801310694072,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.007887173146715316,
      "loss": 0.8003,
      "num_input_tokens_seen": 53917400,
      "step": 92860
    },
    {
      "epoch": 13.831546023235031,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.007885456686852699,
      "loss": 0.8017,
      "num_input_tokens_seen": 53920440,
      "step": 92865
    },
    {
      "epoch": 13.83229073577599,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.007883740347183053,
      "loss": 0.8106,
      "num_input_tokens_seen": 53923640,
      "step": 92870
    },
    {
      "epoch": 13.83303544831695,
      "grad_norm": 0.0546875,
      "learning_rate": 0.007882024127735388,
      "loss": 0.7735,
      "num_input_tokens_seen": 53926584,
      "step": 92875
    },
    {
      "epoch": 13.83378016085791,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.007880308028538684,
      "loss": 0.7898,
      "num_input_tokens_seen": 53929432,
      "step": 92880
    },
    {
      "epoch": 13.834524873398868,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.007878592049621946,
      "loss": 0.7925,
      "num_input_tokens_seen": 53932248,
      "step": 92885
    },
    {
      "epoch": 13.835269585939827,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.007876876191014152,
      "loss": 0.8032,
      "num_input_tokens_seen": 53935064,
      "step": 92890
    },
    {
      "epoch": 13.836014298480787,
      "grad_norm": 0.064453125,
      "learning_rate": 0.007875160452744301,
      "loss": 0.8144,
      "num_input_tokens_seen": 53938232,
      "step": 92895
    },
    {
      "epoch": 13.836759011021746,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.007873444834841372,
      "loss": 0.7946,
      "num_input_tokens_seen": 53941176,
      "step": 92900
    },
    {
      "epoch": 13.837503723562705,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.00787172933733435,
      "loss": 0.7826,
      "num_input_tokens_seen": 53944120,
      "step": 92905
    },
    {
      "epoch": 13.838248436103664,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.007870013960252217,
      "loss": 0.7595,
      "num_input_tokens_seen": 53946968,
      "step": 92910
    },
    {
      "epoch": 13.838993148644622,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.007868298703623945,
      "loss": 0.7785,
      "num_input_tokens_seen": 53950008,
      "step": 92915
    },
    {
      "epoch": 13.839737861185583,
      "grad_norm": 0.03515625,
      "learning_rate": 0.007866583567478522,
      "loss": 0.7865,
      "num_input_tokens_seen": 53953144,
      "step": 92920
    },
    {
      "epoch": 13.840482573726542,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.007864868551844916,
      "loss": 0.8027,
      "num_input_tokens_seen": 53956312,
      "step": 92925
    },
    {
      "epoch": 13.8412272862675,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.007863153656752109,
      "loss": 0.7941,
      "num_input_tokens_seen": 53959448,
      "step": 92930
    },
    {
      "epoch": 13.84197199880846,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00786143888222907,
      "loss": 0.8314,
      "num_input_tokens_seen": 53962488,
      "step": 92935
    },
    {
      "epoch": 13.84271671134942,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.007859724228304756,
      "loss": 0.7973,
      "num_input_tokens_seen": 53965592,
      "step": 92940
    },
    {
      "epoch": 13.843461423890378,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.007858009695008149,
      "loss": 0.8185,
      "num_input_tokens_seen": 53968312,
      "step": 92945
    },
    {
      "epoch": 13.844206136431337,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.007856295282368208,
      "loss": 0.7838,
      "num_input_tokens_seen": 53971192,
      "step": 92950
    },
    {
      "epoch": 13.844950848972296,
      "grad_norm": 0.03515625,
      "learning_rate": 0.007854580990413902,
      "loss": 0.8153,
      "num_input_tokens_seen": 53974008,
      "step": 92955
    },
    {
      "epoch": 13.845695561513256,
      "grad_norm": 0.041015625,
      "learning_rate": 0.00785286681917418,
      "loss": 0.7705,
      "num_input_tokens_seen": 53976696,
      "step": 92960
    },
    {
      "epoch": 13.846440274054215,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.007851152768678017,
      "loss": 0.7708,
      "num_input_tokens_seen": 53979480,
      "step": 92965
    },
    {
      "epoch": 13.847184986595174,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.007849438838954363,
      "loss": 0.7937,
      "num_input_tokens_seen": 53982360,
      "step": 92970
    },
    {
      "epoch": 13.847929699136133,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.00784772503003217,
      "loss": 0.7997,
      "num_input_tokens_seen": 53984952,
      "step": 92975
    },
    {
      "epoch": 13.848674411677093,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.007846011341940394,
      "loss": 0.8259,
      "num_input_tokens_seen": 53987576,
      "step": 92980
    },
    {
      "epoch": 13.849419124218052,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.00784429777470798,
      "loss": 0.8197,
      "num_input_tokens_seen": 53990136,
      "step": 92985
    },
    {
      "epoch": 13.85016383675901,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00784258432836389,
      "loss": 0.7833,
      "num_input_tokens_seen": 53993304,
      "step": 92990
    },
    {
      "epoch": 13.85090854929997,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.007840871002937054,
      "loss": 0.8123,
      "num_input_tokens_seen": 53996440,
      "step": 92995
    },
    {
      "epoch": 13.85165326184093,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.007839157798456434,
      "loss": 0.8368,
      "num_input_tokens_seen": 53999224,
      "step": 93000
    },
    {
      "epoch": 13.852397974381889,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.007837444714950959,
      "loss": 0.8604,
      "num_input_tokens_seen": 54002264,
      "step": 93005
    },
    {
      "epoch": 13.853142686922848,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.00783573175244958,
      "loss": 0.7977,
      "num_input_tokens_seen": 54004920,
      "step": 93010
    },
    {
      "epoch": 13.853887399463806,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.007834018910981228,
      "loss": 0.8276,
      "num_input_tokens_seen": 54007704,
      "step": 93015
    },
    {
      "epoch": 13.854632112004767,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.007832306190574848,
      "loss": 0.7816,
      "num_input_tokens_seen": 54010552,
      "step": 93020
    },
    {
      "epoch": 13.855376824545726,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.007830593591259372,
      "loss": 0.7988,
      "num_input_tokens_seen": 54013304,
      "step": 93025
    },
    {
      "epoch": 13.856121537086684,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.007828881113063721,
      "loss": 0.7986,
      "num_input_tokens_seen": 54016184,
      "step": 93030
    },
    {
      "epoch": 13.856866249627643,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.007827168756016844,
      "loss": 0.796,
      "num_input_tokens_seen": 54019128,
      "step": 93035
    },
    {
      "epoch": 13.857610962168604,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.00782545652014766,
      "loss": 0.7866,
      "num_input_tokens_seen": 54022072,
      "step": 93040
    },
    {
      "epoch": 13.858355674709562,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.007823744405485097,
      "loss": 0.7953,
      "num_input_tokens_seen": 54025048,
      "step": 93045
    },
    {
      "epoch": 13.859100387250521,
      "grad_norm": 0.033203125,
      "learning_rate": 0.007822032412058071,
      "loss": 0.7925,
      "num_input_tokens_seen": 54027928,
      "step": 93050
    },
    {
      "epoch": 13.85984509979148,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.00782032053989552,
      "loss": 0.778,
      "num_input_tokens_seen": 54030936,
      "step": 93055
    },
    {
      "epoch": 13.86058981233244,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.007818608789026356,
      "loss": 0.8169,
      "num_input_tokens_seen": 54034072,
      "step": 93060
    },
    {
      "epoch": 13.8613345248734,
      "grad_norm": 0.041015625,
      "learning_rate": 0.007816897159479488,
      "loss": 0.7868,
      "num_input_tokens_seen": 54037144,
      "step": 93065
    },
    {
      "epoch": 13.862079237414358,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.007815185651283852,
      "loss": 0.7811,
      "num_input_tokens_seen": 54040056,
      "step": 93070
    },
    {
      "epoch": 13.862823949955317,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.007813474264468346,
      "loss": 0.7746,
      "num_input_tokens_seen": 54042904,
      "step": 93075
    },
    {
      "epoch": 13.863568662496277,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.007811762999061894,
      "loss": 0.7963,
      "num_input_tokens_seen": 54045816,
      "step": 93080
    },
    {
      "epoch": 13.864313375037236,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.007810051855093391,
      "loss": 0.7834,
      "num_input_tokens_seen": 54048280,
      "step": 93085
    },
    {
      "epoch": 13.865058087578195,
      "grad_norm": 0.0625,
      "learning_rate": 0.007808340832591764,
      "loss": 0.8051,
      "num_input_tokens_seen": 54051224,
      "step": 93090
    },
    {
      "epoch": 13.865802800119154,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.007806629931585903,
      "loss": 0.7732,
      "num_input_tokens_seen": 54054200,
      "step": 93095
    },
    {
      "epoch": 13.866547512660112,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.007804919152104724,
      "loss": 0.8018,
      "num_input_tokens_seen": 54056984,
      "step": 93100
    },
    {
      "epoch": 13.867292225201073,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.007803208494177123,
      "loss": 0.8099,
      "num_input_tokens_seen": 54059640,
      "step": 93105
    },
    {
      "epoch": 13.868036937742032,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.007801497957831999,
      "loss": 0.7837,
      "num_input_tokens_seen": 54062264,
      "step": 93110
    },
    {
      "epoch": 13.86878165028299,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.007799787543098252,
      "loss": 0.7877,
      "num_input_tokens_seen": 54065016,
      "step": 93115
    },
    {
      "epoch": 13.86952636282395,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00779807725000477,
      "loss": 0.7754,
      "num_input_tokens_seen": 54067928,
      "step": 93120
    },
    {
      "epoch": 13.87027107536491,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.007796367078580459,
      "loss": 0.7972,
      "num_input_tokens_seen": 54070840,
      "step": 93125
    },
    {
      "epoch": 13.871015787905868,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0077946570288541985,
      "loss": 0.7998,
      "num_input_tokens_seen": 54073496,
      "step": 93130
    },
    {
      "epoch": 13.871760500446827,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00779294710085489,
      "loss": 0.7693,
      "num_input_tokens_seen": 54076408,
      "step": 93135
    },
    {
      "epoch": 13.872505212987786,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.007791237294611416,
      "loss": 0.7911,
      "num_input_tokens_seen": 54079096,
      "step": 93140
    },
    {
      "epoch": 13.873249925528746,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.007789527610152655,
      "loss": 0.7764,
      "num_input_tokens_seen": 54081848,
      "step": 93145
    },
    {
      "epoch": 13.873994638069705,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.007787818047507503,
      "loss": 0.8083,
      "num_input_tokens_seen": 54084984,
      "step": 93150
    },
    {
      "epoch": 13.874739350610664,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0077861086067048295,
      "loss": 0.7801,
      "num_input_tokens_seen": 54087736,
      "step": 93155
    },
    {
      "epoch": 13.875484063151623,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.007784399287773525,
      "loss": 0.8174,
      "num_input_tokens_seen": 54090456,
      "step": 93160
    },
    {
      "epoch": 13.876228775692583,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.0077826900907424605,
      "loss": 0.7853,
      "num_input_tokens_seen": 54093336,
      "step": 93165
    },
    {
      "epoch": 13.876973488233542,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.007780981015640511,
      "loss": 0.8424,
      "num_input_tokens_seen": 54096376,
      "step": 93170
    },
    {
      "epoch": 13.8777182007745,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.0077792720624965455,
      "loss": 0.8106,
      "num_input_tokens_seen": 54099128,
      "step": 93175
    },
    {
      "epoch": 13.87846291331546,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.007777563231339444,
      "loss": 0.7879,
      "num_input_tokens_seen": 54102200,
      "step": 93180
    },
    {
      "epoch": 13.87920762585642,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0077758545221980726,
      "loss": 0.7938,
      "num_input_tokens_seen": 54104984,
      "step": 93185
    },
    {
      "epoch": 13.879952338397379,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0077741459351012895,
      "loss": 0.7703,
      "num_input_tokens_seen": 54107704,
      "step": 93190
    },
    {
      "epoch": 13.880697050938338,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.007772437470077974,
      "loss": 0.7887,
      "num_input_tokens_seen": 54110680,
      "step": 93195
    },
    {
      "epoch": 13.881441763479296,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.007770729127156974,
      "loss": 0.8043,
      "num_input_tokens_seen": 54113656,
      "step": 93200
    },
    {
      "epoch": 13.882186476020257,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.007769020906367168,
      "loss": 0.8297,
      "num_input_tokens_seen": 54116760,
      "step": 93205
    },
    {
      "epoch": 13.882931188561216,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.007767312807737393,
      "loss": 0.8064,
      "num_input_tokens_seen": 54119768,
      "step": 93210
    },
    {
      "epoch": 13.883675901102174,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.007765604831296527,
      "loss": 0.8173,
      "num_input_tokens_seen": 54122712,
      "step": 93215
    },
    {
      "epoch": 13.884420613643133,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0077638969770734145,
      "loss": 0.8069,
      "num_input_tokens_seen": 54125464,
      "step": 93220
    },
    {
      "epoch": 13.885165326184094,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0077621892450969,
      "loss": 0.7999,
      "num_input_tokens_seen": 54128376,
      "step": 93225
    },
    {
      "epoch": 13.885910038725052,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.007760481635395848,
      "loss": 0.8099,
      "num_input_tokens_seen": 54131512,
      "step": 93230
    },
    {
      "epoch": 13.886654751266011,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.007758774147999103,
      "loss": 0.7938,
      "num_input_tokens_seen": 54134648,
      "step": 93235
    },
    {
      "epoch": 13.88739946380697,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.007757066782935508,
      "loss": 0.7799,
      "num_input_tokens_seen": 54137848,
      "step": 93240
    },
    {
      "epoch": 13.88814417634793,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0077553595402339035,
      "loss": 0.7832,
      "num_input_tokens_seen": 54140568,
      "step": 93245
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.007753652419923141,
      "loss": 0.7901,
      "num_input_tokens_seen": 54143736,
      "step": 93250
    },
    {
      "epoch": 13.889633601429848,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0077519454220320505,
      "loss": 0.7916,
      "num_input_tokens_seen": 54146488,
      "step": 93255
    },
    {
      "epoch": 13.890378313970807,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.007750238546589481,
      "loss": 0.8099,
      "num_input_tokens_seen": 54149368,
      "step": 93260
    },
    {
      "epoch": 13.891123026511767,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.007748531793624262,
      "loss": 0.8014,
      "num_input_tokens_seen": 54152312,
      "step": 93265
    },
    {
      "epoch": 13.891867739052726,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.007746825163165224,
      "loss": 0.8087,
      "num_input_tokens_seen": 54155032,
      "step": 93270
    },
    {
      "epoch": 13.892612451593685,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.007745118655241208,
      "loss": 0.7837,
      "num_input_tokens_seen": 54158104,
      "step": 93275
    },
    {
      "epoch": 13.893357164134644,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.007743412269881033,
      "loss": 0.8012,
      "num_input_tokens_seen": 54160728,
      "step": 93280
    },
    {
      "epoch": 13.894101876675602,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.007741706007113541,
      "loss": 0.7813,
      "num_input_tokens_seen": 54163672,
      "step": 93285
    },
    {
      "epoch": 13.894846589216563,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.00773999986696754,
      "loss": 0.7743,
      "num_input_tokens_seen": 54166520,
      "step": 93290
    },
    {
      "epoch": 13.895591301757522,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.007738293849471869,
      "loss": 0.8155,
      "num_input_tokens_seen": 54169432,
      "step": 93295
    },
    {
      "epoch": 13.89633601429848,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.007736587954655342,
      "loss": 0.7971,
      "num_input_tokens_seen": 54172472,
      "step": 93300
    },
    {
      "epoch": 13.89708072683944,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.007734882182546782,
      "loss": 0.7854,
      "num_input_tokens_seen": 54175736,
      "step": 93305
    },
    {
      "epoch": 13.8978254393804,
      "grad_norm": 0.134765625,
      "learning_rate": 0.007733176533175002,
      "loss": 0.8284,
      "num_input_tokens_seen": 54178680,
      "step": 93310
    },
    {
      "epoch": 13.898570151921358,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.007731471006568813,
      "loss": 0.7915,
      "num_input_tokens_seen": 54181784,
      "step": 93315
    },
    {
      "epoch": 13.899314864462317,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.007729765602757039,
      "loss": 0.7677,
      "num_input_tokens_seen": 54184664,
      "step": 93320
    },
    {
      "epoch": 13.900059577003276,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.007728060321768482,
      "loss": 0.8173,
      "num_input_tokens_seen": 54187480,
      "step": 93325
    },
    {
      "epoch": 13.900804289544237,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.007726355163631959,
      "loss": 0.8104,
      "num_input_tokens_seen": 54190520,
      "step": 93330
    },
    {
      "epoch": 13.901549002085195,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.007724650128376269,
      "loss": 0.7871,
      "num_input_tokens_seen": 54193304,
      "step": 93335
    },
    {
      "epoch": 13.902293714626154,
      "grad_norm": 0.056640625,
      "learning_rate": 0.007722945216030224,
      "loss": 0.7844,
      "num_input_tokens_seen": 54196088,
      "step": 93340
    },
    {
      "epoch": 13.903038427167113,
      "grad_norm": 0.03515625,
      "learning_rate": 0.007721240426622626,
      "loss": 0.7741,
      "num_input_tokens_seen": 54198904,
      "step": 93345
    },
    {
      "epoch": 13.903783139708073,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.007719535760182267,
      "loss": 0.7899,
      "num_input_tokens_seen": 54201752,
      "step": 93350
    },
    {
      "epoch": 13.904527852249032,
      "grad_norm": 0.0390625,
      "learning_rate": 0.007717831216737957,
      "loss": 0.7853,
      "num_input_tokens_seen": 54204696,
      "step": 93355
    },
    {
      "epoch": 13.90527256478999,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.007716126796318481,
      "loss": 0.7795,
      "num_input_tokens_seen": 54207640,
      "step": 93360
    },
    {
      "epoch": 13.90601727733095,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.007714422498952646,
      "loss": 0.8436,
      "num_input_tokens_seen": 54210584,
      "step": 93365
    },
    {
      "epoch": 13.90676198987191,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.00771271832466924,
      "loss": 0.7681,
      "num_input_tokens_seen": 54213432,
      "step": 93370
    },
    {
      "epoch": 13.907506702412869,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0077110142734970475,
      "loss": 0.78,
      "num_input_tokens_seen": 54215992,
      "step": 93375
    },
    {
      "epoch": 13.908251414953828,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.007709310345464864,
      "loss": 0.7969,
      "num_input_tokens_seen": 54218712,
      "step": 93380
    },
    {
      "epoch": 13.908996127494786,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.007707606540601464,
      "loss": 0.7907,
      "num_input_tokens_seen": 54221592,
      "step": 93385
    },
    {
      "epoch": 13.909740840035747,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.007705902858935648,
      "loss": 0.7744,
      "num_input_tokens_seen": 54224248,
      "step": 93390
    },
    {
      "epoch": 13.910485552576706,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.007704199300496184,
      "loss": 0.7991,
      "num_input_tokens_seen": 54227352,
      "step": 93395
    },
    {
      "epoch": 13.911230265117664,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.007702495865311862,
      "loss": 0.8059,
      "num_input_tokens_seen": 54230680,
      "step": 93400
    },
    {
      "epoch": 13.911974977658623,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.00770079255341145,
      "loss": 0.8198,
      "num_input_tokens_seen": 54233400,
      "step": 93405
    },
    {
      "epoch": 13.912719690199584,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0076990893648237356,
      "loss": 0.8177,
      "num_input_tokens_seen": 54236120,
      "step": 93410
    },
    {
      "epoch": 13.913464402740543,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0076973862995774805,
      "loss": 0.7935,
      "num_input_tokens_seen": 54239000,
      "step": 93415
    },
    {
      "epoch": 13.914209115281501,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.007695683357701467,
      "loss": 0.8069,
      "num_input_tokens_seen": 54241976,
      "step": 93420
    },
    {
      "epoch": 13.91495382782246,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.007693980539224459,
      "loss": 0.7988,
      "num_input_tokens_seen": 54245208,
      "step": 93425
    },
    {
      "epoch": 13.915698540363419,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.007692277844175219,
      "loss": 0.809,
      "num_input_tokens_seen": 54248056,
      "step": 93430
    },
    {
      "epoch": 13.91644325290438,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.007690575272582522,
      "loss": 0.8071,
      "num_input_tokens_seen": 54250840,
      "step": 93435
    },
    {
      "epoch": 13.917187965445338,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.007688872824475128,
      "loss": 0.8236,
      "num_input_tokens_seen": 54253912,
      "step": 93440
    },
    {
      "epoch": 13.917932677986297,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.007687170499881799,
      "loss": 0.7892,
      "num_input_tokens_seen": 54256728,
      "step": 93445
    },
    {
      "epoch": 13.918677390527257,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.007685468298831283,
      "loss": 0.8008,
      "num_input_tokens_seen": 54259416,
      "step": 93450
    },
    {
      "epoch": 13.919422103068216,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.007683766221352352,
      "loss": 0.8052,
      "num_input_tokens_seen": 54262232,
      "step": 93455
    },
    {
      "epoch": 13.920166815609175,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0076820642674737556,
      "loss": 0.7869,
      "num_input_tokens_seen": 54265272,
      "step": 93460
    },
    {
      "epoch": 13.920911528150134,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.007680362437224241,
      "loss": 0.8131,
      "num_input_tokens_seen": 54268408,
      "step": 93465
    },
    {
      "epoch": 13.921656240691092,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.007678660730632569,
      "loss": 0.7846,
      "num_input_tokens_seen": 54271288,
      "step": 93470
    },
    {
      "epoch": 13.922400953232053,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.007676959147727476,
      "loss": 0.802,
      "num_input_tokens_seen": 54274104,
      "step": 93475
    },
    {
      "epoch": 13.923145665773012,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.007675257688537722,
      "loss": 0.8053,
      "num_input_tokens_seen": 54276920,
      "step": 93480
    },
    {
      "epoch": 13.92389037831397,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00767355635309204,
      "loss": 0.7997,
      "num_input_tokens_seen": 54279704,
      "step": 93485
    },
    {
      "epoch": 13.92463509085493,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00767185514141918,
      "loss": 0.8086,
      "num_input_tokens_seen": 54282616,
      "step": 93490
    },
    {
      "epoch": 13.92537980339589,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0076701540535478825,
      "loss": 0.8241,
      "num_input_tokens_seen": 54285336,
      "step": 93495
    },
    {
      "epoch": 13.926124515936849,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0076684530895068815,
      "loss": 0.7877,
      "num_input_tokens_seen": 54288312,
      "step": 93500
    },
    {
      "epoch": 13.926869228477807,
      "grad_norm": 0.04296875,
      "learning_rate": 0.007666752249324908,
      "loss": 0.7944,
      "num_input_tokens_seen": 54291512,
      "step": 93505
    },
    {
      "epoch": 13.927613941018766,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.007665051533030707,
      "loss": 0.8017,
      "num_input_tokens_seen": 54294200,
      "step": 93510
    },
    {
      "epoch": 13.928358653559727,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.007663350940653006,
      "loss": 0.8005,
      "num_input_tokens_seen": 54296600,
      "step": 93515
    },
    {
      "epoch": 13.929103366100685,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.007661650472220529,
      "loss": 0.8076,
      "num_input_tokens_seen": 54299320,
      "step": 93520
    },
    {
      "epoch": 13.929848078641644,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.007659950127762016,
      "loss": 0.7706,
      "num_input_tokens_seen": 54302136,
      "step": 93525
    },
    {
      "epoch": 13.930592791182603,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.007658249907306178,
      "loss": 0.8039,
      "num_input_tokens_seen": 54305304,
      "step": 93530
    },
    {
      "epoch": 13.931337503723563,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.007656549810881752,
      "loss": 0.7724,
      "num_input_tokens_seen": 54308280,
      "step": 93535
    },
    {
      "epoch": 13.932082216264522,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0076548498385174525,
      "loss": 0.793,
      "num_input_tokens_seen": 54311160,
      "step": 93540
    },
    {
      "epoch": 13.932826928805481,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.007653149990241996,
      "loss": 0.7636,
      "num_input_tokens_seen": 54313944,
      "step": 93545
    },
    {
      "epoch": 13.93357164134644,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.007651450266084108,
      "loss": 0.8212,
      "num_input_tokens_seen": 54316792,
      "step": 93550
    },
    {
      "epoch": 13.9343163538874,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.007649750666072492,
      "loss": 0.8043,
      "num_input_tokens_seen": 54319992,
      "step": 93555
    },
    {
      "epoch": 13.935061066428359,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.007648051190235875,
      "loss": 0.7991,
      "num_input_tokens_seen": 54322840,
      "step": 93560
    },
    {
      "epoch": 13.935805778969318,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.007646351838602962,
      "loss": 0.8076,
      "num_input_tokens_seen": 54325848,
      "step": 93565
    },
    {
      "epoch": 13.936550491510276,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.007644652611202459,
      "loss": 0.7868,
      "num_input_tokens_seen": 54328600,
      "step": 93570
    },
    {
      "epoch": 13.937295204051237,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.00764295350806307,
      "loss": 0.7977,
      "num_input_tokens_seen": 54331224,
      "step": 93575
    },
    {
      "epoch": 13.938039916592196,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.00764125452921351,
      "loss": 0.7678,
      "num_input_tokens_seen": 54334200,
      "step": 93580
    },
    {
      "epoch": 13.938784629133155,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.007639555674682475,
      "loss": 0.7868,
      "num_input_tokens_seen": 54337016,
      "step": 93585
    },
    {
      "epoch": 13.939529341674113,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00763785694449866,
      "loss": 0.7832,
      "num_input_tokens_seen": 54339928,
      "step": 93590
    },
    {
      "epoch": 13.940274054215074,
      "grad_norm": 0.033203125,
      "learning_rate": 0.007636158338690776,
      "loss": 0.7919,
      "num_input_tokens_seen": 54342776,
      "step": 93595
    },
    {
      "epoch": 13.941018766756033,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0076344598572875046,
      "loss": 0.8139,
      "num_input_tokens_seen": 54345752,
      "step": 93600
    },
    {
      "epoch": 13.941763479296991,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.007632761500317555,
      "loss": 0.8167,
      "num_input_tokens_seen": 54348376,
      "step": 93605
    },
    {
      "epoch": 13.94250819183795,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0076310632678096065,
      "loss": 0.7979,
      "num_input_tokens_seen": 54351448,
      "step": 93610
    },
    {
      "epoch": 13.943252904378909,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.007629365159792362,
      "loss": 0.809,
      "num_input_tokens_seen": 54354456,
      "step": 93615
    },
    {
      "epoch": 13.94399761691987,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.007627667176294493,
      "loss": 0.8016,
      "num_input_tokens_seen": 54357432,
      "step": 93620
    },
    {
      "epoch": 13.944742329460828,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.007625969317344702,
      "loss": 0.8016,
      "num_input_tokens_seen": 54360536,
      "step": 93625
    },
    {
      "epoch": 13.945487042001787,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.007624271582971666,
      "loss": 0.8003,
      "num_input_tokens_seen": 54363352,
      "step": 93630
    },
    {
      "epoch": 13.946231754542747,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.007622573973204063,
      "loss": 0.8224,
      "num_input_tokens_seen": 54366200,
      "step": 93635
    },
    {
      "epoch": 13.946976467083706,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.007620876488070576,
      "loss": 0.7959,
      "num_input_tokens_seen": 54368792,
      "step": 93640
    },
    {
      "epoch": 13.947721179624665,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0076191791275998745,
      "loss": 0.7865,
      "num_input_tokens_seen": 54371480,
      "step": 93645
    },
    {
      "epoch": 13.948465892165624,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.007617481891820646,
      "loss": 0.8002,
      "num_input_tokens_seen": 54374488,
      "step": 93650
    },
    {
      "epoch": 13.949210604706582,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.007615784780761551,
      "loss": 0.83,
      "num_input_tokens_seen": 54377208,
      "step": 93655
    },
    {
      "epoch": 13.949955317247543,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.007614087794451274,
      "loss": 0.805,
      "num_input_tokens_seen": 54380088,
      "step": 93660
    },
    {
      "epoch": 13.950700029788502,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.007612390932918478,
      "loss": 0.7819,
      "num_input_tokens_seen": 54383064,
      "step": 93665
    },
    {
      "epoch": 13.95144474232946,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.007610694196191822,
      "loss": 0.7964,
      "num_input_tokens_seen": 54385912,
      "step": 93670
    },
    {
      "epoch": 13.95218945487042,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.007608997584299981,
      "loss": 0.8025,
      "num_input_tokens_seen": 54388728,
      "step": 93675
    },
    {
      "epoch": 13.95293416741138,
      "grad_norm": 0.025390625,
      "learning_rate": 0.00760730109727161,
      "loss": 0.8054,
      "num_input_tokens_seen": 54391640,
      "step": 93680
    },
    {
      "epoch": 13.953678879952339,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.007605604735135379,
      "loss": 0.7982,
      "num_input_tokens_seen": 54394680,
      "step": 93685
    },
    {
      "epoch": 13.954423592493297,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.007603908497919936,
      "loss": 0.8141,
      "num_input_tokens_seen": 54397560,
      "step": 93690
    },
    {
      "epoch": 13.955168305034256,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.007602212385653946,
      "loss": 0.802,
      "num_input_tokens_seen": 54400408,
      "step": 93695
    },
    {
      "epoch": 13.955913017575217,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0076005163983660605,
      "loss": 0.8128,
      "num_input_tokens_seen": 54403256,
      "step": 93700
    },
    {
      "epoch": 13.956657730116175,
      "grad_norm": 0.044921875,
      "learning_rate": 0.00759882053608493,
      "loss": 0.8097,
      "num_input_tokens_seen": 54406040,
      "step": 93705
    },
    {
      "epoch": 13.957402442657134,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.007597124798839202,
      "loss": 0.7792,
      "num_input_tokens_seen": 54408856,
      "step": 93710
    },
    {
      "epoch": 13.958147155198093,
      "grad_norm": 0.0546875,
      "learning_rate": 0.007595429186657523,
      "loss": 0.8134,
      "num_input_tokens_seen": 54411544,
      "step": 93715
    },
    {
      "epoch": 13.958891867739053,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.007593733699568546,
      "loss": 0.7927,
      "num_input_tokens_seen": 54414296,
      "step": 93720
    },
    {
      "epoch": 13.959636580280012,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.007592038337600907,
      "loss": 0.7944,
      "num_input_tokens_seen": 54416984,
      "step": 93725
    },
    {
      "epoch": 13.960381292820971,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.007590343100783258,
      "loss": 0.7975,
      "num_input_tokens_seen": 54419864,
      "step": 93730
    },
    {
      "epoch": 13.96112600536193,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.007588647989144223,
      "loss": 0.7996,
      "num_input_tokens_seen": 54422808,
      "step": 93735
    },
    {
      "epoch": 13.96187071790289,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.007586953002712453,
      "loss": 0.8115,
      "num_input_tokens_seen": 54426136,
      "step": 93740
    },
    {
      "epoch": 13.962615430443849,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.00758525814151658,
      "loss": 0.791,
      "num_input_tokens_seen": 54428984,
      "step": 93745
    },
    {
      "epoch": 13.963360142984808,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0075835634055852275,
      "loss": 0.8081,
      "num_input_tokens_seen": 54432088,
      "step": 93750
    },
    {
      "epoch": 13.964104855525767,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.00758186879494704,
      "loss": 0.805,
      "num_input_tokens_seen": 54434904,
      "step": 93755
    },
    {
      "epoch": 13.964849568066727,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.007580174309630634,
      "loss": 0.805,
      "num_input_tokens_seen": 54437720,
      "step": 93760
    },
    {
      "epoch": 13.965594280607686,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.007578479949664647,
      "loss": 0.7979,
      "num_input_tokens_seen": 54440792,
      "step": 93765
    },
    {
      "epoch": 13.966338993148645,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.007576785715077698,
      "loss": 0.7993,
      "num_input_tokens_seen": 54443704,
      "step": 93770
    },
    {
      "epoch": 13.967083705689603,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.007575091605898411,
      "loss": 0.7809,
      "num_input_tokens_seen": 54446392,
      "step": 93775
    },
    {
      "epoch": 13.967828418230564,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0075733976221554,
      "loss": 0.8077,
      "num_input_tokens_seen": 54449304,
      "step": 93780
    },
    {
      "epoch": 13.968573130771523,
      "grad_norm": 0.03125,
      "learning_rate": 0.007571703763877291,
      "loss": 0.8047,
      "num_input_tokens_seen": 54452056,
      "step": 93785
    },
    {
      "epoch": 13.969317843312481,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.007570010031092698,
      "loss": 0.8132,
      "num_input_tokens_seen": 54454968,
      "step": 93790
    },
    {
      "epoch": 13.97006255585344,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.00756831642383023,
      "loss": 0.8152,
      "num_input_tokens_seen": 54457752,
      "step": 93795
    },
    {
      "epoch": 13.970807268394399,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.007566622942118508,
      "loss": 0.7952,
      "num_input_tokens_seen": 54460440,
      "step": 93800
    },
    {
      "epoch": 13.97155198093536,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.007564929585986129,
      "loss": 0.7742,
      "num_input_tokens_seen": 54463256,
      "step": 93805
    },
    {
      "epoch": 13.972296693476318,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.007563236355461715,
      "loss": 0.7834,
      "num_input_tokens_seen": 54465912,
      "step": 93810
    },
    {
      "epoch": 13.973041406017277,
      "grad_norm": 0.037109375,
      "learning_rate": 0.007561543250573859,
      "loss": 0.8085,
      "num_input_tokens_seen": 54468568,
      "step": 93815
    },
    {
      "epoch": 13.973786118558236,
      "grad_norm": 0.037109375,
      "learning_rate": 0.007559850271351175,
      "loss": 0.8036,
      "num_input_tokens_seen": 54471192,
      "step": 93820
    },
    {
      "epoch": 13.974530831099196,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.007558157417822259,
      "loss": 0.7786,
      "num_input_tokens_seen": 54474072,
      "step": 93825
    },
    {
      "epoch": 13.975275543640155,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00755646469001571,
      "loss": 0.7821,
      "num_input_tokens_seen": 54477080,
      "step": 93830
    },
    {
      "epoch": 13.976020256181114,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.00755477208796012,
      "loss": 0.8173,
      "num_input_tokens_seen": 54480152,
      "step": 93835
    },
    {
      "epoch": 13.976764968722073,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.007553079611684092,
      "loss": 0.8095,
      "num_input_tokens_seen": 54483288,
      "step": 93840
    },
    {
      "epoch": 13.977509681263033,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.007551387261216216,
      "loss": 0.7941,
      "num_input_tokens_seen": 54486456,
      "step": 93845
    },
    {
      "epoch": 13.978254393803992,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0075496950365850765,
      "loss": 0.7801,
      "num_input_tokens_seen": 54489208,
      "step": 93850
    },
    {
      "epoch": 13.97899910634495,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.007548002937819273,
      "loss": 0.7962,
      "num_input_tokens_seen": 54492408,
      "step": 93855
    },
    {
      "epoch": 13.97974381888591,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.00754631096494738,
      "loss": 0.7734,
      "num_input_tokens_seen": 54495352,
      "step": 93860
    },
    {
      "epoch": 13.98048853142687,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.007544619117997993,
      "loss": 0.7951,
      "num_input_tokens_seen": 54498232,
      "step": 93865
    },
    {
      "epoch": 13.981233243967829,
      "grad_norm": 0.041015625,
      "learning_rate": 0.00754292739699969,
      "loss": 0.806,
      "num_input_tokens_seen": 54501304,
      "step": 93870
    },
    {
      "epoch": 13.981977956508787,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.007541235801981043,
      "loss": 0.819,
      "num_input_tokens_seen": 54503960,
      "step": 93875
    },
    {
      "epoch": 13.982722669049746,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.007539544332970642,
      "loss": 0.8019,
      "num_input_tokens_seen": 54506872,
      "step": 93880
    },
    {
      "epoch": 13.983467381590707,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.007537852989997051,
      "loss": 0.798,
      "num_input_tokens_seen": 54510008,
      "step": 93885
    },
    {
      "epoch": 13.984212094131665,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.0075361617730888555,
      "loss": 0.7816,
      "num_input_tokens_seen": 54513048,
      "step": 93890
    },
    {
      "epoch": 13.984956806672624,
      "grad_norm": 0.041015625,
      "learning_rate": 0.007534470682274622,
      "loss": 0.8184,
      "num_input_tokens_seen": 54515768,
      "step": 93895
    },
    {
      "epoch": 13.985701519213583,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.007532779717582917,
      "loss": 0.8043,
      "num_input_tokens_seen": 54518712,
      "step": 93900
    },
    {
      "epoch": 13.986446231754543,
      "grad_norm": 0.2021484375,
      "learning_rate": 0.00753108887904231,
      "loss": 0.8086,
      "num_input_tokens_seen": 54521944,
      "step": 93905
    },
    {
      "epoch": 13.987190944295502,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.00752939816668136,
      "loss": 0.7938,
      "num_input_tokens_seen": 54524696,
      "step": 93910
    },
    {
      "epoch": 13.987935656836461,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0075277075805286415,
      "loss": 0.8103,
      "num_input_tokens_seen": 54527448,
      "step": 93915
    },
    {
      "epoch": 13.98868036937742,
      "grad_norm": 0.023193359375,
      "learning_rate": 0.007526017120612701,
      "loss": 0.8001,
      "num_input_tokens_seen": 54530456,
      "step": 93920
    },
    {
      "epoch": 13.98942508191838,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.007524326786962114,
      "loss": 0.8003,
      "num_input_tokens_seen": 54533240,
      "step": 93925
    },
    {
      "epoch": 13.990169794459339,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.007522636579605419,
      "loss": 0.7835,
      "num_input_tokens_seen": 54535864,
      "step": 93930
    },
    {
      "epoch": 13.990914507000298,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.007520946498571187,
      "loss": 0.7874,
      "num_input_tokens_seen": 54538616,
      "step": 93935
    },
    {
      "epoch": 13.991659219541257,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.007519256543887956,
      "loss": 0.7907,
      "num_input_tokens_seen": 54541304,
      "step": 93940
    },
    {
      "epoch": 13.992403932082215,
      "grad_norm": 0.02734375,
      "learning_rate": 0.007517566715584288,
      "loss": 0.7839,
      "num_input_tokens_seen": 54544152,
      "step": 93945
    },
    {
      "epoch": 13.993148644623176,
      "grad_norm": 0.052734375,
      "learning_rate": 0.007515877013688726,
      "loss": 0.7868,
      "num_input_tokens_seen": 54547032,
      "step": 93950
    },
    {
      "epoch": 13.993893357164135,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.00751418743822981,
      "loss": 0.7984,
      "num_input_tokens_seen": 54549816,
      "step": 93955
    },
    {
      "epoch": 13.994638069705093,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.007512497989236097,
      "loss": 0.8068,
      "num_input_tokens_seen": 54552760,
      "step": 93960
    },
    {
      "epoch": 13.995382782246054,
      "grad_norm": 0.04296875,
      "learning_rate": 0.007510808666736116,
      "loss": 0.7956,
      "num_input_tokens_seen": 54555736,
      "step": 93965
    },
    {
      "epoch": 13.996127494787013,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.007509119470758413,
      "loss": 0.8256,
      "num_input_tokens_seen": 54558680,
      "step": 93970
    },
    {
      "epoch": 13.996872207327971,
      "grad_norm": 0.025390625,
      "learning_rate": 0.007507430401331517,
      "loss": 0.7962,
      "num_input_tokens_seen": 54561656,
      "step": 93975
    },
    {
      "epoch": 13.99761691986893,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.007505741458483976,
      "loss": 0.7933,
      "num_input_tokens_seen": 54564408,
      "step": 93980
    },
    {
      "epoch": 13.998361632409889,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0075040526422443165,
      "loss": 0.8037,
      "num_input_tokens_seen": 54567736,
      "step": 93985
    },
    {
      "epoch": 13.99910634495085,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.007502363952641061,
      "loss": 0.7968,
      "num_input_tokens_seen": 54570552,
      "step": 93990
    },
    {
      "epoch": 13.999851057491808,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0075006753897027545,
      "loss": 0.8153,
      "num_input_tokens_seen": 54573752,
      "step": 93995
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.7991399168968201,
      "eval_runtime": 70.5269,
      "eval_samples_per_second": 42.31,
      "eval_steps_per_second": 10.578,
      "num_input_tokens_seen": 54573896,
      "step": 93996
    },
    {
      "epoch": 14.000595770032767,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.007498986953457908,
      "loss": 0.7965,
      "num_input_tokens_seen": 54576136,
      "step": 94000
    },
    {
      "epoch": 14.001340482573726,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.007497298643935058,
      "loss": 0.7961,
      "num_input_tokens_seen": 54579176,
      "step": 94005
    },
    {
      "epoch": 14.002085195114686,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.007495610461162719,
      "loss": 0.7897,
      "num_input_tokens_seen": 54581896,
      "step": 94010
    },
    {
      "epoch": 14.002829907655645,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.007493922405169418,
      "loss": 0.8156,
      "num_input_tokens_seen": 54584648,
      "step": 94015
    },
    {
      "epoch": 14.003574620196604,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.007492234475983665,
      "loss": 0.7974,
      "num_input_tokens_seen": 54587400,
      "step": 94020
    },
    {
      "epoch": 14.004319332737563,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0074905466736339835,
      "loss": 0.8129,
      "num_input_tokens_seen": 54590728,
      "step": 94025
    },
    {
      "epoch": 14.005064045278523,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.007488858998148884,
      "loss": 0.7934,
      "num_input_tokens_seen": 54593704,
      "step": 94030
    },
    {
      "epoch": 14.005808757819482,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.0074871714495568795,
      "loss": 0.7936,
      "num_input_tokens_seen": 54596584,
      "step": 94035
    },
    {
      "epoch": 14.00655347036044,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.007485484027886476,
      "loss": 0.8029,
      "num_input_tokens_seen": 54599208,
      "step": 94040
    },
    {
      "epoch": 14.0072981829014,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.007483796733166178,
      "loss": 0.802,
      "num_input_tokens_seen": 54602152,
      "step": 94045
    },
    {
      "epoch": 14.00804289544236,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0074821095654244996,
      "loss": 0.8196,
      "num_input_tokens_seen": 54604776,
      "step": 94050
    },
    {
      "epoch": 14.008787607983319,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.007480422524689934,
      "loss": 0.8165,
      "num_input_tokens_seen": 54607528,
      "step": 94055
    },
    {
      "epoch": 14.009532320524277,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.007478735610990994,
      "loss": 0.8165,
      "num_input_tokens_seen": 54610664,
      "step": 94060
    },
    {
      "epoch": 14.010277033065236,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.007477048824356172,
      "loss": 0.7776,
      "num_input_tokens_seen": 54613672,
      "step": 94065
    },
    {
      "epoch": 14.011021745606197,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.007475362164813956,
      "loss": 0.7931,
      "num_input_tokens_seen": 54616520,
      "step": 94070
    },
    {
      "epoch": 14.011766458147155,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.007473675632392855,
      "loss": 0.812,
      "num_input_tokens_seen": 54619208,
      "step": 94075
    },
    {
      "epoch": 14.012511170688114,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0074719892271213504,
      "loss": 0.7731,
      "num_input_tokens_seen": 54622152,
      "step": 94080
    },
    {
      "epoch": 14.013255883229073,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.007470302949027942,
      "loss": 0.7773,
      "num_input_tokens_seen": 54625096,
      "step": 94085
    },
    {
      "epoch": 14.014000595770034,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.007468616798141105,
      "loss": 0.793,
      "num_input_tokens_seen": 54627880,
      "step": 94090
    },
    {
      "epoch": 14.014745308310992,
      "grad_norm": 0.03125,
      "learning_rate": 0.00746693077448934,
      "loss": 0.7814,
      "num_input_tokens_seen": 54630696,
      "step": 94095
    },
    {
      "epoch": 14.015490020851951,
      "grad_norm": 0.046875,
      "learning_rate": 0.007465244878101123,
      "loss": 0.7874,
      "num_input_tokens_seen": 54633768,
      "step": 94100
    },
    {
      "epoch": 14.01623473339291,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.007463559109004936,
      "loss": 0.7799,
      "num_input_tokens_seen": 54636616,
      "step": 94105
    },
    {
      "epoch": 14.01697944593387,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.007461873467229258,
      "loss": 0.765,
      "num_input_tokens_seen": 54639464,
      "step": 94110
    },
    {
      "epoch": 14.017724158474829,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.007460187952802559,
      "loss": 0.7889,
      "num_input_tokens_seen": 54642600,
      "step": 94115
    },
    {
      "epoch": 14.018468871015788,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.007458502565753329,
      "loss": 0.77,
      "num_input_tokens_seen": 54645480,
      "step": 94120
    },
    {
      "epoch": 14.019213583556747,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.007456817306110028,
      "loss": 0.7874,
      "num_input_tokens_seen": 54648776,
      "step": 94125
    },
    {
      "epoch": 14.019958296097707,
      "grad_norm": 0.046875,
      "learning_rate": 0.007455132173901137,
      "loss": 0.7556,
      "num_input_tokens_seen": 54651432,
      "step": 94130
    },
    {
      "epoch": 14.020703008638666,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.007453447169155113,
      "loss": 0.8188,
      "num_input_tokens_seen": 54654504,
      "step": 94135
    },
    {
      "epoch": 14.021447721179625,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.007451762291900436,
      "loss": 0.7931,
      "num_input_tokens_seen": 54657256,
      "step": 94140
    },
    {
      "epoch": 14.022192433720583,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.007450077542165564,
      "loss": 0.7905,
      "num_input_tokens_seen": 54660424,
      "step": 94145
    },
    {
      "epoch": 14.022937146261542,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0074483929199789505,
      "loss": 0.8206,
      "num_input_tokens_seen": 54663592,
      "step": 94150
    },
    {
      "epoch": 14.023681858802503,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.007446708425369071,
      "loss": 0.8091,
      "num_input_tokens_seen": 54666472,
      "step": 94155
    },
    {
      "epoch": 14.024426571343461,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.0074450240583643764,
      "loss": 0.7886,
      "num_input_tokens_seen": 54669704,
      "step": 94160
    },
    {
      "epoch": 14.02517128388442,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.007443339818993314,
      "loss": 0.8084,
      "num_input_tokens_seen": 54672904,
      "step": 94165
    },
    {
      "epoch": 14.025915996425379,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.007441655707284352,
      "loss": 0.7798,
      "num_input_tokens_seen": 54675976,
      "step": 94170
    },
    {
      "epoch": 14.02666070896634,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.007439971723265935,
      "loss": 0.8033,
      "num_input_tokens_seen": 54678952,
      "step": 94175
    },
    {
      "epoch": 14.027405421507298,
      "grad_norm": 0.037109375,
      "learning_rate": 0.007438287866966507,
      "loss": 0.8012,
      "num_input_tokens_seen": 54681672,
      "step": 94180
    },
    {
      "epoch": 14.028150134048257,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.007436604138414525,
      "loss": 0.7705,
      "num_input_tokens_seen": 54684520,
      "step": 94185
    },
    {
      "epoch": 14.028894846589216,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00743492053763843,
      "loss": 0.7949,
      "num_input_tokens_seen": 54687208,
      "step": 94190
    },
    {
      "epoch": 14.029639559130176,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.007433237064666656,
      "loss": 0.7804,
      "num_input_tokens_seen": 54689768,
      "step": 94195
    },
    {
      "epoch": 14.030384271671135,
      "grad_norm": 0.044921875,
      "learning_rate": 0.007431553719527658,
      "loss": 0.7831,
      "num_input_tokens_seen": 54692776,
      "step": 94200
    },
    {
      "epoch": 14.031128984212094,
      "grad_norm": 0.029296875,
      "learning_rate": 0.007429870502249861,
      "loss": 0.7673,
      "num_input_tokens_seen": 54695464,
      "step": 94205
    },
    {
      "epoch": 14.031873696753053,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0074281874128617155,
      "loss": 0.7595,
      "num_input_tokens_seen": 54698312,
      "step": 94210
    },
    {
      "epoch": 14.032618409294013,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0074265044513916415,
      "loss": 0.7753,
      "num_input_tokens_seen": 54701480,
      "step": 94215
    },
    {
      "epoch": 14.033363121834972,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.007424821617868082,
      "loss": 0.8011,
      "num_input_tokens_seen": 54704456,
      "step": 94220
    },
    {
      "epoch": 14.03410783437593,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.007423138912319464,
      "loss": 0.7846,
      "num_input_tokens_seen": 54707240,
      "step": 94225
    },
    {
      "epoch": 14.03485254691689,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0074214563347742134,
      "loss": 0.7721,
      "num_input_tokens_seen": 54710120,
      "step": 94230
    },
    {
      "epoch": 14.03559725945785,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.007419773885260755,
      "loss": 0.79,
      "num_input_tokens_seen": 54712808,
      "step": 94235
    },
    {
      "epoch": 14.036341971998809,
      "grad_norm": 0.046875,
      "learning_rate": 0.007418091563807505,
      "loss": 0.7879,
      "num_input_tokens_seen": 54715848,
      "step": 94240
    },
    {
      "epoch": 14.037086684539767,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0074164093704429,
      "loss": 0.7942,
      "num_input_tokens_seen": 54718920,
      "step": 94245
    },
    {
      "epoch": 14.037831397080726,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.007414727305195345,
      "loss": 0.7854,
      "num_input_tokens_seen": 54721928,
      "step": 94250
    },
    {
      "epoch": 14.038576109621687,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.007413045368093268,
      "loss": 0.8018,
      "num_input_tokens_seen": 54724712,
      "step": 94255
    },
    {
      "epoch": 14.039320822162646,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0074113635591650726,
      "loss": 0.782,
      "num_input_tokens_seen": 54727496,
      "step": 94260
    },
    {
      "epoch": 14.040065534703604,
      "grad_norm": 0.072265625,
      "learning_rate": 0.007409681878439184,
      "loss": 0.8522,
      "num_input_tokens_seen": 54730312,
      "step": 94265
    },
    {
      "epoch": 14.040810247244563,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.007408000325944005,
      "loss": 0.7897,
      "num_input_tokens_seen": 54733512,
      "step": 94270
    },
    {
      "epoch": 14.041554959785524,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0074063189017079395,
      "loss": 0.8376,
      "num_input_tokens_seen": 54736488,
      "step": 94275
    },
    {
      "epoch": 14.042299672326482,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.007404637605759402,
      "loss": 0.8029,
      "num_input_tokens_seen": 54739368,
      "step": 94280
    },
    {
      "epoch": 14.043044384867441,
      "grad_norm": 0.0625,
      "learning_rate": 0.007402956438126789,
      "loss": 0.7946,
      "num_input_tokens_seen": 54742152,
      "step": 94285
    },
    {
      "epoch": 14.0437890974084,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.00740127539883851,
      "loss": 0.798,
      "num_input_tokens_seen": 54745224,
      "step": 94290
    },
    {
      "epoch": 14.04453380994936,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.00739959448792296,
      "loss": 0.8069,
      "num_input_tokens_seen": 54748008,
      "step": 94295
    },
    {
      "epoch": 14.04527852249032,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.007397913705408537,
      "loss": 0.7838,
      "num_input_tokens_seen": 54750888,
      "step": 94300
    },
    {
      "epoch": 14.046023235031278,
      "grad_norm": 0.037109375,
      "learning_rate": 0.007396233051323629,
      "loss": 0.7802,
      "num_input_tokens_seen": 54753928,
      "step": 94305
    },
    {
      "epoch": 14.046767947572237,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.007394552525696642,
      "loss": 0.8282,
      "num_input_tokens_seen": 54757160,
      "step": 94310
    },
    {
      "epoch": 14.047512660113195,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.00739287212855596,
      "loss": 0.8006,
      "num_input_tokens_seen": 54760040,
      "step": 94315
    },
    {
      "epoch": 14.048257372654156,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.007391191859929967,
      "loss": 0.807,
      "num_input_tokens_seen": 54763304,
      "step": 94320
    },
    {
      "epoch": 14.049002085195115,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00738951171984706,
      "loss": 0.8084,
      "num_input_tokens_seen": 54766248,
      "step": 94325
    },
    {
      "epoch": 14.049746797736073,
      "grad_norm": 0.056640625,
      "learning_rate": 0.00738783170833561,
      "loss": 0.7982,
      "num_input_tokens_seen": 54769000,
      "step": 94330
    },
    {
      "epoch": 14.050491510277032,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0073861518254240135,
      "loss": 0.8144,
      "num_input_tokens_seen": 54771816,
      "step": 94335
    },
    {
      "epoch": 14.051236222817993,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.007384472071140637,
      "loss": 0.8175,
      "num_input_tokens_seen": 54774536,
      "step": 94340
    },
    {
      "epoch": 14.051980935358952,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0073827924455138715,
      "loss": 0.7961,
      "num_input_tokens_seen": 54777416,
      "step": 94345
    },
    {
      "epoch": 14.05272564789991,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.007381112948572087,
      "loss": 0.7782,
      "num_input_tokens_seen": 54780168,
      "step": 94350
    },
    {
      "epoch": 14.053470360440869,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0073794335803436485,
      "loss": 0.7915,
      "num_input_tokens_seen": 54783144,
      "step": 94355
    },
    {
      "epoch": 14.05421507298183,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.007377754340856938,
      "loss": 0.8069,
      "num_input_tokens_seen": 54786024,
      "step": 94360
    },
    {
      "epoch": 14.054959785522788,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.007376075230140325,
      "loss": 0.8074,
      "num_input_tokens_seen": 54789000,
      "step": 94365
    },
    {
      "epoch": 14.055704498063747,
      "grad_norm": 0.03125,
      "learning_rate": 0.007374396248222171,
      "loss": 0.8173,
      "num_input_tokens_seen": 54791912,
      "step": 94370
    },
    {
      "epoch": 14.056449210604706,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.007372717395130836,
      "loss": 0.7855,
      "num_input_tokens_seen": 54794856,
      "step": 94375
    },
    {
      "epoch": 14.057193923145666,
      "grad_norm": 0.0625,
      "learning_rate": 0.007371038670894694,
      "loss": 0.7882,
      "num_input_tokens_seen": 54797768,
      "step": 94380
    },
    {
      "epoch": 14.057938635686625,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0073693600755420936,
      "loss": 0.8133,
      "num_input_tokens_seen": 54800712,
      "step": 94385
    },
    {
      "epoch": 14.058683348227584,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0073676816091014065,
      "loss": 0.793,
      "num_input_tokens_seen": 54803400,
      "step": 94390
    },
    {
      "epoch": 14.059428060768543,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.007366003271600982,
      "loss": 0.7839,
      "num_input_tokens_seen": 54806312,
      "step": 94395
    },
    {
      "epoch": 14.060172773309503,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.007364325063069165,
      "loss": 0.8147,
      "num_input_tokens_seen": 54809480,
      "step": 94400
    },
    {
      "epoch": 14.060917485850462,
      "grad_norm": 0.044921875,
      "learning_rate": 0.007362646983534324,
      "loss": 0.7869,
      "num_input_tokens_seen": 54812392,
      "step": 94405
    },
    {
      "epoch": 14.06166219839142,
      "grad_norm": 0.052734375,
      "learning_rate": 0.007360969033024792,
      "loss": 0.7919,
      "num_input_tokens_seen": 54815304,
      "step": 94410
    },
    {
      "epoch": 14.06240691093238,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.007359291211568931,
      "loss": 0.7895,
      "num_input_tokens_seen": 54818216,
      "step": 94415
    },
    {
      "epoch": 14.06315162347334,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.007357613519195075,
      "loss": 0.7877,
      "num_input_tokens_seen": 54820872,
      "step": 94420
    },
    {
      "epoch": 14.063896336014299,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.007355935955931575,
      "loss": 0.8089,
      "num_input_tokens_seen": 54823720,
      "step": 94425
    },
    {
      "epoch": 14.064641048555258,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.007354258521806768,
      "loss": 0.7914,
      "num_input_tokens_seen": 54826536,
      "step": 94430
    },
    {
      "epoch": 14.065385761096216,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.007352581216848994,
      "loss": 0.7857,
      "num_input_tokens_seen": 54829352,
      "step": 94435
    },
    {
      "epoch": 14.066130473637177,
      "grad_norm": 0.025390625,
      "learning_rate": 0.007350904041086586,
      "loss": 0.7935,
      "num_input_tokens_seen": 54832200,
      "step": 94440
    },
    {
      "epoch": 14.066875186178136,
      "grad_norm": 0.0546875,
      "learning_rate": 0.007349226994547875,
      "loss": 0.7851,
      "num_input_tokens_seen": 54835144,
      "step": 94445
    },
    {
      "epoch": 14.067619898719094,
      "grad_norm": 0.037109375,
      "learning_rate": 0.007347550077261204,
      "loss": 0.7992,
      "num_input_tokens_seen": 54838120,
      "step": 94450
    },
    {
      "epoch": 14.068364611260053,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00734587328925489,
      "loss": 0.8079,
      "num_input_tokens_seen": 54840840,
      "step": 94455
    },
    {
      "epoch": 14.069109323801014,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.007344196630557275,
      "loss": 0.7959,
      "num_input_tokens_seen": 54843848,
      "step": 94460
    },
    {
      "epoch": 14.069854036341972,
      "grad_norm": 0.09765625,
      "learning_rate": 0.007342520101196671,
      "loss": 0.8178,
      "num_input_tokens_seen": 54846728,
      "step": 94465
    },
    {
      "epoch": 14.070598748882931,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.007340843701201412,
      "loss": 0.7881,
      "num_input_tokens_seen": 54849768,
      "step": 94470
    },
    {
      "epoch": 14.07134346142389,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.007339167430599813,
      "loss": 0.801,
      "num_input_tokens_seen": 54852680,
      "step": 94475
    },
    {
      "epoch": 14.07208817396485,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.007337491289420192,
      "loss": 0.8199,
      "num_input_tokens_seen": 54856040,
      "step": 94480
    },
    {
      "epoch": 14.07283288650581,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.00733581527769087,
      "loss": 0.7754,
      "num_input_tokens_seen": 54859208,
      "step": 94485
    },
    {
      "epoch": 14.073577599046768,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.007334139395440162,
      "loss": 0.7959,
      "num_input_tokens_seen": 54862056,
      "step": 94490
    },
    {
      "epoch": 14.074322311587727,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00733246364269637,
      "loss": 0.8156,
      "num_input_tokens_seen": 54865128,
      "step": 94495
    },
    {
      "epoch": 14.075067024128685,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.00733078801948782,
      "loss": 0.7776,
      "num_input_tokens_seen": 54867976,
      "step": 94500
    },
    {
      "epoch": 14.075811736669646,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.007329112525842809,
      "loss": 0.784,
      "num_input_tokens_seen": 54870824,
      "step": 94505
    },
    {
      "epoch": 14.076556449210605,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.007327437161789647,
      "loss": 0.7697,
      "num_input_tokens_seen": 54874056,
      "step": 94510
    },
    {
      "epoch": 14.077301161751564,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.00732576192735663,
      "loss": 0.8058,
      "num_input_tokens_seen": 54876744,
      "step": 94515
    },
    {
      "epoch": 14.078045874292522,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.007324086822572071,
      "loss": 0.7792,
      "num_input_tokens_seen": 54879656,
      "step": 94520
    },
    {
      "epoch": 14.078790586833483,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0073224118474642585,
      "loss": 0.7835,
      "num_input_tokens_seen": 54882600,
      "step": 94525
    },
    {
      "epoch": 14.079535299374442,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.007320737002061501,
      "loss": 0.874,
      "num_input_tokens_seen": 54885352,
      "step": 94530
    },
    {
      "epoch": 14.0802800119154,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.007319062286392079,
      "loss": 0.7968,
      "num_input_tokens_seen": 54888264,
      "step": 94535
    },
    {
      "epoch": 14.081024724456359,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0073173877004843,
      "loss": 0.7994,
      "num_input_tokens_seen": 54891336,
      "step": 94540
    },
    {
      "epoch": 14.08176943699732,
      "grad_norm": 0.03125,
      "learning_rate": 0.0073157132443664435,
      "loss": 0.7622,
      "num_input_tokens_seen": 54894280,
      "step": 94545
    },
    {
      "epoch": 14.082514149538278,
      "grad_norm": 0.03125,
      "learning_rate": 0.007314038918066804,
      "loss": 0.7924,
      "num_input_tokens_seen": 54897608,
      "step": 94550
    },
    {
      "epoch": 14.083258862079237,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.007312364721613669,
      "loss": 0.794,
      "num_input_tokens_seen": 54900296,
      "step": 94555
    },
    {
      "epoch": 14.084003574620196,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.007310690655035317,
      "loss": 0.8042,
      "num_input_tokens_seen": 54903112,
      "step": 94560
    },
    {
      "epoch": 14.084748287161156,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.00730901671836003,
      "loss": 0.767,
      "num_input_tokens_seen": 54906184,
      "step": 94565
    },
    {
      "epoch": 14.085492999702115,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.007307342911616085,
      "loss": 0.8189,
      "num_input_tokens_seen": 54908904,
      "step": 94570
    },
    {
      "epoch": 14.086237712243074,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.007305669234831768,
      "loss": 0.7889,
      "num_input_tokens_seen": 54911816,
      "step": 94575
    },
    {
      "epoch": 14.086982424784033,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.007303995688035345,
      "loss": 0.7916,
      "num_input_tokens_seen": 54914600,
      "step": 94580
    },
    {
      "epoch": 14.087727137324993,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.007302322271255097,
      "loss": 0.7946,
      "num_input_tokens_seen": 54917640,
      "step": 94585
    },
    {
      "epoch": 14.088471849865952,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00730064898451929,
      "loss": 0.8088,
      "num_input_tokens_seen": 54920392,
      "step": 94590
    },
    {
      "epoch": 14.08921656240691,
      "grad_norm": 0.072265625,
      "learning_rate": 0.007298975827856189,
      "loss": 0.7822,
      "num_input_tokens_seen": 54923080,
      "step": 94595
    },
    {
      "epoch": 14.08996127494787,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.007297302801294069,
      "loss": 0.7799,
      "num_input_tokens_seen": 54926024,
      "step": 94600
    },
    {
      "epoch": 14.09070598748883,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.007295629904861184,
      "loss": 0.7864,
      "num_input_tokens_seen": 54929064,
      "step": 94605
    },
    {
      "epoch": 14.091450700029789,
      "grad_norm": 0.05859375,
      "learning_rate": 0.007293957138585808,
      "loss": 0.826,
      "num_input_tokens_seen": 54932104,
      "step": 94610
    },
    {
      "epoch": 14.092195412570748,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0072922845024961875,
      "loss": 0.8036,
      "num_input_tokens_seen": 54934824,
      "step": 94615
    },
    {
      "epoch": 14.092940125111706,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.007290611996620592,
      "loss": 0.7901,
      "num_input_tokens_seen": 54937768,
      "step": 94620
    },
    {
      "epoch": 14.093684837652667,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0072889396209872715,
      "loss": 0.7831,
      "num_input_tokens_seen": 54940712,
      "step": 94625
    },
    {
      "epoch": 14.094429550193626,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00728726737562448,
      "loss": 0.7901,
      "num_input_tokens_seen": 54943464,
      "step": 94630
    },
    {
      "epoch": 14.095174262734584,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.007285595260560466,
      "loss": 0.8002,
      "num_input_tokens_seen": 54946184,
      "step": 94635
    },
    {
      "epoch": 14.095918975275543,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.007283923275823475,
      "loss": 0.8153,
      "num_input_tokens_seen": 54948968,
      "step": 94640
    },
    {
      "epoch": 14.096663687816504,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0072822514214417634,
      "loss": 0.7882,
      "num_input_tokens_seen": 54952040,
      "step": 94645
    },
    {
      "epoch": 14.097408400357462,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0072805796974435635,
      "loss": 0.7939,
      "num_input_tokens_seen": 54954984,
      "step": 94650
    },
    {
      "epoch": 14.098153112898421,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00727890810385713,
      "loss": 0.7821,
      "num_input_tokens_seen": 54958184,
      "step": 94655
    },
    {
      "epoch": 14.09889782543938,
      "grad_norm": 0.05078125,
      "learning_rate": 0.007277236640710691,
      "loss": 0.7871,
      "num_input_tokens_seen": 54961384,
      "step": 94660
    },
    {
      "epoch": 14.099642537980339,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0072755653080324965,
      "loss": 0.7908,
      "num_input_tokens_seen": 54964040,
      "step": 94665
    },
    {
      "epoch": 14.1003872505213,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.007273894105850774,
      "loss": 0.8013,
      "num_input_tokens_seen": 54966824,
      "step": 94670
    },
    {
      "epoch": 14.101131963062258,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.007272223034193753,
      "loss": 0.7728,
      "num_input_tokens_seen": 54969704,
      "step": 94675
    },
    {
      "epoch": 14.101876675603217,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.007270552093089675,
      "loss": 0.7905,
      "num_input_tokens_seen": 54972296,
      "step": 94680
    },
    {
      "epoch": 14.102621388144176,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0072688812825667595,
      "loss": 0.7991,
      "num_input_tokens_seen": 54975208,
      "step": 94685
    },
    {
      "epoch": 14.103366100685136,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.007267210602653242,
      "loss": 0.7761,
      "num_input_tokens_seen": 54977864,
      "step": 94690
    },
    {
      "epoch": 14.104110813226095,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.007265540053377341,
      "loss": 0.7876,
      "num_input_tokens_seen": 54980712,
      "step": 94695
    },
    {
      "epoch": 14.104855525767054,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0072638696347672805,
      "loss": 0.7932,
      "num_input_tokens_seen": 54983400,
      "step": 94700
    },
    {
      "epoch": 14.105600238308012,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.007262199346851274,
      "loss": 0.781,
      "num_input_tokens_seen": 54986248,
      "step": 94705
    },
    {
      "epoch": 14.106344950848973,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.007260529189657552,
      "loss": 0.8112,
      "num_input_tokens_seen": 54989704,
      "step": 94710
    },
    {
      "epoch": 14.107089663389932,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.007258859163214323,
      "loss": 0.8105,
      "num_input_tokens_seen": 54992776,
      "step": 94715
    },
    {
      "epoch": 14.10783437593089,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.007257189267549793,
      "loss": 0.8012,
      "num_input_tokens_seen": 54995880,
      "step": 94720
    },
    {
      "epoch": 14.10857908847185,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0072555195026921876,
      "loss": 0.8444,
      "num_input_tokens_seen": 54998472,
      "step": 94725
    },
    {
      "epoch": 14.10932380101281,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0072538498686697045,
      "loss": 0.8081,
      "num_input_tokens_seen": 55001384,
      "step": 94730
    },
    {
      "epoch": 14.110068513553768,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.007252180365510561,
      "loss": 0.812,
      "num_input_tokens_seen": 55004264,
      "step": 94735
    },
    {
      "epoch": 14.110813226094727,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.007250510993242948,
      "loss": 0.795,
      "num_input_tokens_seen": 55007368,
      "step": 94740
    },
    {
      "epoch": 14.111557938635686,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0072488417518950824,
      "loss": 0.8052,
      "num_input_tokens_seen": 55010024,
      "step": 94745
    },
    {
      "epoch": 14.112302651176647,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.007247172641495157,
      "loss": 0.7769,
      "num_input_tokens_seen": 55012872,
      "step": 94750
    },
    {
      "epoch": 14.113047363717605,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.007245503662071366,
      "loss": 0.8415,
      "num_input_tokens_seen": 55015784,
      "step": 94755
    },
    {
      "epoch": 14.113792076258564,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0072438348136519125,
      "loss": 0.7826,
      "num_input_tokens_seen": 55018600,
      "step": 94760
    },
    {
      "epoch": 14.114536788799523,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.007242166096264989,
      "loss": 0.8252,
      "num_input_tokens_seen": 55021576,
      "step": 94765
    },
    {
      "epoch": 14.115281501340483,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.007240497509938782,
      "loss": 0.7661,
      "num_input_tokens_seen": 55024136,
      "step": 94770
    },
    {
      "epoch": 14.116026213881442,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.00723882905470148,
      "loss": 0.7958,
      "num_input_tokens_seen": 55027112,
      "step": 94775
    },
    {
      "epoch": 14.1167709264224,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.007237160730581278,
      "loss": 0.7964,
      "num_input_tokens_seen": 55030088,
      "step": 94780
    },
    {
      "epoch": 14.11751563896336,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0072354925376063505,
      "loss": 0.8104,
      "num_input_tokens_seen": 55032872,
      "step": 94785
    },
    {
      "epoch": 14.11826035150432,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.007233824475804891,
      "loss": 0.7962,
      "num_input_tokens_seen": 55036168,
      "step": 94790
    },
    {
      "epoch": 14.119005064045279,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.007232156545205071,
      "loss": 0.8049,
      "num_input_tokens_seen": 55039112,
      "step": 94795
    },
    {
      "epoch": 14.119749776586238,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0072304887458350675,
      "loss": 0.7946,
      "num_input_tokens_seen": 55042024,
      "step": 94800
    },
    {
      "epoch": 14.120494489127196,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0072288210777230666,
      "loss": 0.8102,
      "num_input_tokens_seen": 55044968,
      "step": 94805
    },
    {
      "epoch": 14.121239201668157,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.00722715354089723,
      "loss": 0.7956,
      "num_input_tokens_seen": 55047912,
      "step": 94810
    },
    {
      "epoch": 14.121983914209116,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.007225486135385738,
      "loss": 0.7807,
      "num_input_tokens_seen": 55050984,
      "step": 94815
    },
    {
      "epoch": 14.122728626750074,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.0072238188612167575,
      "loss": 0.7922,
      "num_input_tokens_seen": 55053832,
      "step": 94820
    },
    {
      "epoch": 14.123473339291033,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.0072221517184184476,
      "loss": 0.7897,
      "num_input_tokens_seen": 55056712,
      "step": 94825
    },
    {
      "epoch": 14.124218051831992,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.007220484707018984,
      "loss": 0.784,
      "num_input_tokens_seen": 55059432,
      "step": 94830
    },
    {
      "epoch": 14.124962764372953,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0072188178270465265,
      "loss": 0.8103,
      "num_input_tokens_seen": 55062600,
      "step": 94835
    },
    {
      "epoch": 14.125707476913911,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.007217151078529233,
      "loss": 0.8016,
      "num_input_tokens_seen": 55065416,
      "step": 94840
    },
    {
      "epoch": 14.12645218945487,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.007215484461495255,
      "loss": 0.8048,
      "num_input_tokens_seen": 55068104,
      "step": 94845
    },
    {
      "epoch": 14.127196901995829,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.007213817975972762,
      "loss": 0.8115,
      "num_input_tokens_seen": 55070920,
      "step": 94850
    },
    {
      "epoch": 14.12794161453679,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0072121516219898945,
      "loss": 0.7757,
      "num_input_tokens_seen": 55073896,
      "step": 94855
    },
    {
      "epoch": 14.128686327077748,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.007210485399574816,
      "loss": 0.7884,
      "num_input_tokens_seen": 55076776,
      "step": 94860
    },
    {
      "epoch": 14.129431039618707,
      "grad_norm": 0.029296875,
      "learning_rate": 0.007208819308755665,
      "loss": 0.7769,
      "num_input_tokens_seen": 55079592,
      "step": 94865
    },
    {
      "epoch": 14.130175752159666,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.007207153349560597,
      "loss": 0.7896,
      "num_input_tokens_seen": 55082376,
      "step": 94870
    },
    {
      "epoch": 14.130920464700626,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0072054875220177545,
      "loss": 0.7855,
      "num_input_tokens_seen": 55084936,
      "step": 94875
    },
    {
      "epoch": 14.131665177241585,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.007203821826155273,
      "loss": 0.8034,
      "num_input_tokens_seen": 55087688,
      "step": 94880
    },
    {
      "epoch": 14.132409889782544,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.007202156262001301,
      "loss": 0.7961,
      "num_input_tokens_seen": 55090632,
      "step": 94885
    },
    {
      "epoch": 14.133154602323502,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.0072004908295839765,
      "loss": 0.7792,
      "num_input_tokens_seen": 55093928,
      "step": 94890
    },
    {
      "epoch": 14.133899314864463,
      "grad_norm": 0.046875,
      "learning_rate": 0.007198825528931431,
      "loss": 0.8033,
      "num_input_tokens_seen": 55096712,
      "step": 94895
    },
    {
      "epoch": 14.134644027405422,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.007197160360071795,
      "loss": 0.81,
      "num_input_tokens_seen": 55099624,
      "step": 94900
    },
    {
      "epoch": 14.13538873994638,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0071954953230332075,
      "loss": 0.804,
      "num_input_tokens_seen": 55102760,
      "step": 94905
    },
    {
      "epoch": 14.13613345248734,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.007193830417843791,
      "loss": 0.7928,
      "num_input_tokens_seen": 55105640,
      "step": 94910
    },
    {
      "epoch": 14.1368781650283,
      "grad_norm": 0.041015625,
      "learning_rate": 0.00719216564453168,
      "loss": 0.7875,
      "num_input_tokens_seen": 55108200,
      "step": 94915
    },
    {
      "epoch": 14.137622877569259,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.007190501003124995,
      "loss": 0.8007,
      "num_input_tokens_seen": 55111176,
      "step": 94920
    },
    {
      "epoch": 14.138367590110217,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.007188836493651853,
      "loss": 0.8048,
      "num_input_tokens_seen": 55114088,
      "step": 94925
    },
    {
      "epoch": 14.139112302651176,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.007187172116140386,
      "loss": 0.785,
      "num_input_tokens_seen": 55117128,
      "step": 94930
    },
    {
      "epoch": 14.139857015192137,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.007185507870618699,
      "loss": 0.781,
      "num_input_tokens_seen": 55120360,
      "step": 94935
    },
    {
      "epoch": 14.140601727733095,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.00718384375711492,
      "loss": 0.8006,
      "num_input_tokens_seen": 55123240,
      "step": 94940
    },
    {
      "epoch": 14.141346440274054,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.007182179775657151,
      "loss": 0.7873,
      "num_input_tokens_seen": 55126120,
      "step": 94945
    },
    {
      "epoch": 14.142091152815013,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.007180515926273515,
      "loss": 0.7806,
      "num_input_tokens_seen": 55129128,
      "step": 94950
    },
    {
      "epoch": 14.142835865355973,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.007178852208992115,
      "loss": 0.8014,
      "num_input_tokens_seen": 55132232,
      "step": 94955
    },
    {
      "epoch": 14.143580577896932,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.007177188623841057,
      "loss": 0.782,
      "num_input_tokens_seen": 55134920,
      "step": 94960
    },
    {
      "epoch": 14.14432529043789,
      "grad_norm": 0.033203125,
      "learning_rate": 0.007175525170848448,
      "loss": 0.8032,
      "num_input_tokens_seen": 55137832,
      "step": 94965
    },
    {
      "epoch": 14.14507000297885,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.007173861850042382,
      "loss": 0.7833,
      "num_input_tokens_seen": 55140616,
      "step": 94970
    },
    {
      "epoch": 14.14581471551981,
      "grad_norm": 0.048828125,
      "learning_rate": 0.007172198661450972,
      "loss": 0.8194,
      "num_input_tokens_seen": 55143528,
      "step": 94975
    },
    {
      "epoch": 14.146559428060769,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.007170535605102304,
      "loss": 0.7976,
      "num_input_tokens_seen": 55146568,
      "step": 94980
    },
    {
      "epoch": 14.147304140601728,
      "grad_norm": 0.072265625,
      "learning_rate": 0.007168872681024486,
      "loss": 0.7716,
      "num_input_tokens_seen": 55149256,
      "step": 94985
    },
    {
      "epoch": 14.148048853142686,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.007167209889245597,
      "loss": 0.7999,
      "num_input_tokens_seen": 55152392,
      "step": 94990
    },
    {
      "epoch": 14.148793565683647,
      "grad_norm": 0.154296875,
      "learning_rate": 0.007165547229793743,
      "loss": 0.8409,
      "num_input_tokens_seen": 55154952,
      "step": 94995
    },
    {
      "epoch": 14.149538278224606,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.007163884702697007,
      "loss": 0.803,
      "num_input_tokens_seen": 55157736,
      "step": 95000
    },
    {
      "epoch": 14.150282990765565,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.007162222307983467,
      "loss": 0.8022,
      "num_input_tokens_seen": 55160808,
      "step": 95005
    },
    {
      "epoch": 14.151027703306523,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0071605600456812225,
      "loss": 0.8123,
      "num_input_tokens_seen": 55163368,
      "step": 95010
    },
    {
      "epoch": 14.151772415847482,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.007158897915818341,
      "loss": 0.7981,
      "num_input_tokens_seen": 55166088,
      "step": 95015
    },
    {
      "epoch": 14.152517128388443,
      "grad_norm": 0.056640625,
      "learning_rate": 0.007157235918422916,
      "loss": 0.8082,
      "num_input_tokens_seen": 55168776,
      "step": 95020
    },
    {
      "epoch": 14.153261840929401,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.007155574053523022,
      "loss": 0.7785,
      "num_input_tokens_seen": 55171560,
      "step": 95025
    },
    {
      "epoch": 14.15400655347036,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.007153912321146728,
      "loss": 0.7971,
      "num_input_tokens_seen": 55174536,
      "step": 95030
    },
    {
      "epoch": 14.154751266011319,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.007152250721322112,
      "loss": 0.7965,
      "num_input_tokens_seen": 55177608,
      "step": 95035
    },
    {
      "epoch": 14.15549597855228,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.007150589254077239,
      "loss": 0.7893,
      "num_input_tokens_seen": 55180264,
      "step": 95040
    },
    {
      "epoch": 14.156240691093238,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.007148927919440187,
      "loss": 0.7916,
      "num_input_tokens_seen": 55183240,
      "step": 95045
    },
    {
      "epoch": 14.156985403634197,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.007147266717439015,
      "loss": 0.804,
      "num_input_tokens_seen": 55186056,
      "step": 95050
    },
    {
      "epoch": 14.157730116175156,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.007145605648101794,
      "loss": 0.7881,
      "num_input_tokens_seen": 55189064,
      "step": 95055
    },
    {
      "epoch": 14.158474828716116,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.007143944711456579,
      "loss": 0.7857,
      "num_input_tokens_seen": 55191624,
      "step": 95060
    },
    {
      "epoch": 14.159219541257075,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.007142283907531436,
      "loss": 0.8047,
      "num_input_tokens_seen": 55194696,
      "step": 95065
    },
    {
      "epoch": 14.159964253798034,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.007140623236354419,
      "loss": 0.7953,
      "num_input_tokens_seen": 55197800,
      "step": 95070
    },
    {
      "epoch": 14.160708966338992,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.007138962697953588,
      "loss": 0.7874,
      "num_input_tokens_seen": 55200616,
      "step": 95075
    },
    {
      "epoch": 14.161453678879953,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.007137302292356991,
      "loss": 0.7854,
      "num_input_tokens_seen": 55203688,
      "step": 95080
    },
    {
      "epoch": 14.162198391420912,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0071356420195926775,
      "loss": 0.8018,
      "num_input_tokens_seen": 55206504,
      "step": 95085
    },
    {
      "epoch": 14.16294310396187,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.007133981879688704,
      "loss": 0.7962,
      "num_input_tokens_seen": 55209384,
      "step": 95090
    },
    {
      "epoch": 14.16368781650283,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.00713232187267311,
      "loss": 0.795,
      "num_input_tokens_seen": 55212168,
      "step": 95095
    },
    {
      "epoch": 14.16443252904379,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.007130661998573944,
      "loss": 0.7927,
      "num_input_tokens_seen": 55215048,
      "step": 95100
    },
    {
      "epoch": 14.165177241584749,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0071290022574192384,
      "loss": 0.8023,
      "num_input_tokens_seen": 55217896,
      "step": 95105
    },
    {
      "epoch": 14.165921954125707,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.007127342649237046,
      "loss": 0.81,
      "num_input_tokens_seen": 55220904,
      "step": 95110
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.007125683174055399,
      "loss": 0.8029,
      "num_input_tokens_seen": 55223784,
      "step": 95115
    },
    {
      "epoch": 14.167411379207627,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.007124023831902326,
      "loss": 0.7874,
      "num_input_tokens_seen": 55226600,
      "step": 95120
    },
    {
      "epoch": 14.168156091748585,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00712236462280587,
      "loss": 0.7893,
      "num_input_tokens_seen": 55229640,
      "step": 95125
    },
    {
      "epoch": 14.168900804289544,
      "grad_norm": 0.080078125,
      "learning_rate": 0.007120705546794052,
      "loss": 0.7899,
      "num_input_tokens_seen": 55232680,
      "step": 95130
    },
    {
      "epoch": 14.169645516830503,
      "grad_norm": 0.05859375,
      "learning_rate": 0.007119046603894912,
      "loss": 0.814,
      "num_input_tokens_seen": 55235624,
      "step": 95135
    },
    {
      "epoch": 14.170390229371463,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.007117387794136463,
      "loss": 0.8025,
      "num_input_tokens_seen": 55238536,
      "step": 95140
    },
    {
      "epoch": 14.171134941912422,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.007115729117546742,
      "loss": 0.7967,
      "num_input_tokens_seen": 55242088,
      "step": 95145
    },
    {
      "epoch": 14.171879654453381,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.007114070574153765,
      "loss": 0.7978,
      "num_input_tokens_seen": 55245128,
      "step": 95150
    },
    {
      "epoch": 14.17262436699434,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.007112412163985544,
      "loss": 0.7955,
      "num_input_tokens_seen": 55247880,
      "step": 95155
    },
    {
      "epoch": 14.1733690795353,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.007110753887070108,
      "loss": 0.81,
      "num_input_tokens_seen": 55250952,
      "step": 95160
    },
    {
      "epoch": 14.174113792076259,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.007109095743435468,
      "loss": 0.7855,
      "num_input_tokens_seen": 55253800,
      "step": 95165
    },
    {
      "epoch": 14.174858504617218,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.007107437733109636,
      "loss": 0.7925,
      "num_input_tokens_seen": 55256712,
      "step": 95170
    },
    {
      "epoch": 14.175603217158177,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.007105779856120616,
      "loss": 0.8189,
      "num_input_tokens_seen": 55259528,
      "step": 95175
    },
    {
      "epoch": 14.176347929699135,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.007104122112496426,
      "loss": 0.7927,
      "num_input_tokens_seen": 55262440,
      "step": 95180
    },
    {
      "epoch": 14.177092642240096,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.007102464502265064,
      "loss": 0.7963,
      "num_input_tokens_seen": 55265128,
      "step": 95185
    },
    {
      "epoch": 14.177837354781055,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.007100807025454542,
      "loss": 0.7857,
      "num_input_tokens_seen": 55268040,
      "step": 95190
    },
    {
      "epoch": 14.178582067322013,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.007099149682092858,
      "loss": 0.7998,
      "num_input_tokens_seen": 55270888,
      "step": 95195
    },
    {
      "epoch": 14.179326779862972,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.007097492472208002,
      "loss": 0.8053,
      "num_input_tokens_seen": 55273608,
      "step": 95200
    },
    {
      "epoch": 14.180071492403933,
      "grad_norm": 0.0390625,
      "learning_rate": 0.007095835395827988,
      "loss": 0.8099,
      "num_input_tokens_seen": 55276840,
      "step": 95205
    },
    {
      "epoch": 14.180816204944891,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.007094178452980794,
      "loss": 0.7915,
      "num_input_tokens_seen": 55279688,
      "step": 95210
    },
    {
      "epoch": 14.18156091748585,
      "grad_norm": 0.04296875,
      "learning_rate": 0.007092521643694426,
      "loss": 0.7887,
      "num_input_tokens_seen": 55282664,
      "step": 95215
    },
    {
      "epoch": 14.182305630026809,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.007090864967996867,
      "loss": 0.7961,
      "num_input_tokens_seen": 55285640,
      "step": 95220
    },
    {
      "epoch": 14.18305034256777,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.0070892084259161076,
      "loss": 0.7979,
      "num_input_tokens_seen": 55288648,
      "step": 95225
    },
    {
      "epoch": 14.183795055108728,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.007087552017480125,
      "loss": 0.7964,
      "num_input_tokens_seen": 55291880,
      "step": 95230
    },
    {
      "epoch": 14.184539767649687,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.007085895742716915,
      "loss": 0.7789,
      "num_input_tokens_seen": 55294984,
      "step": 95235
    },
    {
      "epoch": 14.185284480190646,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.007084239601654453,
      "loss": 0.8312,
      "num_input_tokens_seen": 55297992,
      "step": 95240
    },
    {
      "epoch": 14.186029192731606,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0070825835943207115,
      "loss": 0.7823,
      "num_input_tokens_seen": 55301096,
      "step": 95245
    },
    {
      "epoch": 14.186773905272565,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.007080927720743679,
      "loss": 0.7939,
      "num_input_tokens_seen": 55304072,
      "step": 95250
    },
    {
      "epoch": 14.187518617813524,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.00707927198095132,
      "loss": 0.7875,
      "num_input_tokens_seen": 55307272,
      "step": 95255
    },
    {
      "epoch": 14.188263330354483,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.007077616374971614,
      "loss": 0.7986,
      "num_input_tokens_seen": 55310216,
      "step": 95260
    },
    {
      "epoch": 14.189008042895443,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.007075960902832523,
      "loss": 0.7911,
      "num_input_tokens_seen": 55313000,
      "step": 95265
    },
    {
      "epoch": 14.189752755436402,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.007074305564562025,
      "loss": 0.8468,
      "num_input_tokens_seen": 55316232,
      "step": 95270
    },
    {
      "epoch": 14.19049746797736,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.007072650360188079,
      "loss": 0.8125,
      "num_input_tokens_seen": 55319432,
      "step": 95275
    },
    {
      "epoch": 14.19124218051832,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.007070995289738641,
      "loss": 0.7887,
      "num_input_tokens_seen": 55322248,
      "step": 95280
    },
    {
      "epoch": 14.19198689305928,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.007069340353241685,
      "loss": 0.7941,
      "num_input_tokens_seen": 55325224,
      "step": 95285
    },
    {
      "epoch": 14.192731605600239,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.007067685550725166,
      "loss": 0.7999,
      "num_input_tokens_seen": 55328072,
      "step": 95290
    },
    {
      "epoch": 14.193476318141197,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.007066030882217035,
      "loss": 0.7677,
      "num_input_tokens_seen": 55330952,
      "step": 95295
    },
    {
      "epoch": 14.194221030682156,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.007064376347745244,
      "loss": 0.8003,
      "num_input_tokens_seen": 55333768,
      "step": 95300
    },
    {
      "epoch": 14.194965743223117,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0070627219473377535,
      "loss": 0.8118,
      "num_input_tokens_seen": 55336680,
      "step": 95305
    },
    {
      "epoch": 14.195710455764075,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.007061067681022505,
      "loss": 0.7736,
      "num_input_tokens_seen": 55339496,
      "step": 95310
    },
    {
      "epoch": 14.196455168305034,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.007059413548827453,
      "loss": 0.7879,
      "num_input_tokens_seen": 55342440,
      "step": 95315
    },
    {
      "epoch": 14.197199880845993,
      "grad_norm": 0.03125,
      "learning_rate": 0.0070577595507805395,
      "loss": 0.8096,
      "num_input_tokens_seen": 55345384,
      "step": 95320
    },
    {
      "epoch": 14.197944593386953,
      "grad_norm": 0.05859375,
      "learning_rate": 0.007056105686909701,
      "loss": 0.8349,
      "num_input_tokens_seen": 55348776,
      "step": 95325
    },
    {
      "epoch": 14.198689305927912,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.007054451957242888,
      "loss": 0.7751,
      "num_input_tokens_seen": 55351624,
      "step": 95330
    },
    {
      "epoch": 14.199434018468871,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.007052798361808027,
      "loss": 0.8001,
      "num_input_tokens_seen": 55354536,
      "step": 95335
    },
    {
      "epoch": 14.20017873100983,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.007051144900633068,
      "loss": 0.8121,
      "num_input_tokens_seen": 55357320,
      "step": 95340
    },
    {
      "epoch": 14.200923443550789,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00704949157374593,
      "loss": 0.7786,
      "num_input_tokens_seen": 55360136,
      "step": 95345
    },
    {
      "epoch": 14.201668156091749,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.007047838381174559,
      "loss": 0.7972,
      "num_input_tokens_seen": 55363240,
      "step": 95350
    },
    {
      "epoch": 14.202412868632708,
      "grad_norm": 0.044921875,
      "learning_rate": 0.007046185322946875,
      "loss": 0.7815,
      "num_input_tokens_seen": 55366312,
      "step": 95355
    },
    {
      "epoch": 14.203157581173667,
      "grad_norm": 0.046875,
      "learning_rate": 0.007044532399090805,
      "loss": 0.8105,
      "num_input_tokens_seen": 55369480,
      "step": 95360
    },
    {
      "epoch": 14.203902293714625,
      "grad_norm": 0.048828125,
      "learning_rate": 0.007042879609634277,
      "loss": 0.7885,
      "num_input_tokens_seen": 55372392,
      "step": 95365
    },
    {
      "epoch": 14.204647006255586,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.007041226954605202,
      "loss": 0.7967,
      "num_input_tokens_seen": 55375240,
      "step": 95370
    },
    {
      "epoch": 14.205391718796545,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.007039574434031517,
      "loss": 0.81,
      "num_input_tokens_seen": 55377928,
      "step": 95375
    },
    {
      "epoch": 14.206136431337503,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0070379220479411244,
      "loss": 0.8631,
      "num_input_tokens_seen": 55380840,
      "step": 95380
    },
    {
      "epoch": 14.206881143878462,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.007036269796361954,
      "loss": 0.7937,
      "num_input_tokens_seen": 55383560,
      "step": 95385
    },
    {
      "epoch": 14.207625856419423,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.007034617679321903,
      "loss": 0.7942,
      "num_input_tokens_seen": 55386216,
      "step": 95390
    },
    {
      "epoch": 14.208370568960381,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.007032965696848898,
      "loss": 0.8013,
      "num_input_tokens_seen": 55389000,
      "step": 95395
    },
    {
      "epoch": 14.20911528150134,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.00703131384897084,
      "loss": 0.8049,
      "num_input_tokens_seen": 55392136,
      "step": 95400
    },
    {
      "epoch": 14.209859994042299,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.00702966213571563,
      "loss": 0.781,
      "num_input_tokens_seen": 55395208,
      "step": 95405
    },
    {
      "epoch": 14.21060470658326,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.007028010557111182,
      "loss": 0.7992,
      "num_input_tokens_seen": 55398056,
      "step": 95410
    },
    {
      "epoch": 14.211349419124218,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.007026359113185389,
      "loss": 0.8018,
      "num_input_tokens_seen": 55401160,
      "step": 95415
    },
    {
      "epoch": 14.212094131665177,
      "grad_norm": 0.0625,
      "learning_rate": 0.00702470780396616,
      "loss": 0.7644,
      "num_input_tokens_seen": 55403880,
      "step": 95420
    },
    {
      "epoch": 14.212838844206136,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.007023056629481387,
      "loss": 0.8387,
      "num_input_tokens_seen": 55407112,
      "step": 95425
    },
    {
      "epoch": 14.213583556747096,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.007021405589758964,
      "loss": 0.8277,
      "num_input_tokens_seen": 55410024,
      "step": 95430
    },
    {
      "epoch": 14.214328269288055,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.007019754684826779,
      "loss": 0.7752,
      "num_input_tokens_seen": 55412744,
      "step": 95435
    },
    {
      "epoch": 14.215072981829014,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.007018103914712734,
      "loss": 0.7922,
      "num_input_tokens_seen": 55415976,
      "step": 95440
    },
    {
      "epoch": 14.215817694369973,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.007016453279444712,
      "loss": 0.7939,
      "num_input_tokens_seen": 55419144,
      "step": 95445
    },
    {
      "epoch": 14.216562406910933,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.007014802779050592,
      "loss": 0.7988,
      "num_input_tokens_seen": 55422248,
      "step": 95450
    },
    {
      "epoch": 14.217307119451892,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.007013152413558266,
      "loss": 0.7939,
      "num_input_tokens_seen": 55425064,
      "step": 95455
    },
    {
      "epoch": 14.21805183199285,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.007011502182995609,
      "loss": 0.7939,
      "num_input_tokens_seen": 55427848,
      "step": 95460
    },
    {
      "epoch": 14.21879654453381,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.00700985208739051,
      "loss": 0.7938,
      "num_input_tokens_seen": 55430536,
      "step": 95465
    },
    {
      "epoch": 14.21954125707477,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0070082021267708315,
      "loss": 0.7995,
      "num_input_tokens_seen": 55433416,
      "step": 95470
    },
    {
      "epoch": 14.220285969615729,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0070065523011644615,
      "loss": 0.7837,
      "num_input_tokens_seen": 55436264,
      "step": 95475
    },
    {
      "epoch": 14.221030682156687,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.007004902610599267,
      "loss": 0.8026,
      "num_input_tokens_seen": 55439144,
      "step": 95480
    },
    {
      "epoch": 14.221775394697646,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.007003253055103111,
      "loss": 0.8092,
      "num_input_tokens_seen": 55442248,
      "step": 95485
    },
    {
      "epoch": 14.222520107238607,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.007001603634703873,
      "loss": 0.7744,
      "num_input_tokens_seen": 55445000,
      "step": 95490
    },
    {
      "epoch": 14.223264819779565,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0069999543494294126,
      "loss": 0.7752,
      "num_input_tokens_seen": 55447944,
      "step": 95495
    },
    {
      "epoch": 14.224009532320524,
      "grad_norm": 0.03515625,
      "learning_rate": 0.006998305199307593,
      "loss": 0.8142,
      "num_input_tokens_seen": 55450664,
      "step": 95500
    },
    {
      "epoch": 14.224754244861483,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.006996656184366267,
      "loss": 0.8158,
      "num_input_tokens_seen": 55453544,
      "step": 95505
    },
    {
      "epoch": 14.225498957402444,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.006995007304633308,
      "loss": 0.7955,
      "num_input_tokens_seen": 55456360,
      "step": 95510
    },
    {
      "epoch": 14.226243669943402,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.006993358560136558,
      "loss": 0.7964,
      "num_input_tokens_seen": 55459272,
      "step": 95515
    },
    {
      "epoch": 14.226988382484361,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.006991709950903883,
      "loss": 0.7962,
      "num_input_tokens_seen": 55462024,
      "step": 95520
    },
    {
      "epoch": 14.22773309502532,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.006990061476963131,
      "loss": 0.7969,
      "num_input_tokens_seen": 55464936,
      "step": 95525
    },
    {
      "epoch": 14.228477807566279,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.006988413138342141,
      "loss": 0.8192,
      "num_input_tokens_seen": 55468200,
      "step": 95530
    },
    {
      "epoch": 14.229222520107239,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.006986764935068776,
      "loss": 0.802,
      "num_input_tokens_seen": 55470952,
      "step": 95535
    },
    {
      "epoch": 14.229967232648198,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.006985116867170867,
      "loss": 0.8101,
      "num_input_tokens_seen": 55473896,
      "step": 95540
    },
    {
      "epoch": 14.230711945189157,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.006983468934676268,
      "loss": 0.8158,
      "num_input_tokens_seen": 55476744,
      "step": 95545
    },
    {
      "epoch": 14.231456657730115,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.0069818211376128135,
      "loss": 0.7898,
      "num_input_tokens_seen": 55479752,
      "step": 95550
    },
    {
      "epoch": 14.232201370271076,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.006980173476008342,
      "loss": 0.7694,
      "num_input_tokens_seen": 55482536,
      "step": 95555
    },
    {
      "epoch": 14.232946082812035,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.006978525949890684,
      "loss": 0.789,
      "num_input_tokens_seen": 55485512,
      "step": 95560
    },
    {
      "epoch": 14.233690795352993,
      "grad_norm": 0.044921875,
      "learning_rate": 0.006976878559287681,
      "loss": 0.7844,
      "num_input_tokens_seen": 55488072,
      "step": 95565
    },
    {
      "epoch": 14.234435507893952,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.006975231304227163,
      "loss": 0.7913,
      "num_input_tokens_seen": 55491016,
      "step": 95570
    },
    {
      "epoch": 14.235180220434913,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.006973584184736948,
      "loss": 0.7805,
      "num_input_tokens_seen": 55494248,
      "step": 95575
    },
    {
      "epoch": 14.235924932975871,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.006971937200844877,
      "loss": 0.7955,
      "num_input_tokens_seen": 55497032,
      "step": 95580
    },
    {
      "epoch": 14.23666964551683,
      "grad_norm": 0.060546875,
      "learning_rate": 0.006970290352578762,
      "loss": 0.79,
      "num_input_tokens_seen": 55499752,
      "step": 95585
    },
    {
      "epoch": 14.237414358057789,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.006968643639966436,
      "loss": 0.8007,
      "num_input_tokens_seen": 55502760,
      "step": 95590
    },
    {
      "epoch": 14.23815907059875,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0069669970630357085,
      "loss": 0.7943,
      "num_input_tokens_seen": 55505672,
      "step": 95595
    },
    {
      "epoch": 14.238903783139708,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.006965350621814404,
      "loss": 0.7946,
      "num_input_tokens_seen": 55508520,
      "step": 95600
    },
    {
      "epoch": 14.239648495680667,
      "grad_norm": 0.0240478515625,
      "learning_rate": 0.006963704316330336,
      "loss": 0.8116,
      "num_input_tokens_seen": 55511112,
      "step": 95605
    },
    {
      "epoch": 14.240393208221626,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.006962058146611312,
      "loss": 0.8096,
      "num_input_tokens_seen": 55513896,
      "step": 95610
    },
    {
      "epoch": 14.241137920762586,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.00696041211268515,
      "loss": 0.8089,
      "num_input_tokens_seen": 55516744,
      "step": 95615
    },
    {
      "epoch": 14.241882633303545,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.006958766214579655,
      "loss": 0.7932,
      "num_input_tokens_seen": 55519624,
      "step": 95620
    },
    {
      "epoch": 14.242627345844504,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.006957120452322632,
      "loss": 0.7769,
      "num_input_tokens_seen": 55522600,
      "step": 95625
    },
    {
      "epoch": 14.243372058385463,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.00695547482594188,
      "loss": 0.8088,
      "num_input_tokens_seen": 55525544,
      "step": 95630
    },
    {
      "epoch": 14.244116770926423,
      "grad_norm": 0.03515625,
      "learning_rate": 0.006953829335465209,
      "loss": 0.7905,
      "num_input_tokens_seen": 55528520,
      "step": 95635
    },
    {
      "epoch": 14.244861483467382,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.006952183980920413,
      "loss": 0.8161,
      "num_input_tokens_seen": 55531624,
      "step": 95640
    },
    {
      "epoch": 14.24560619600834,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.006950538762335285,
      "loss": 0.7727,
      "num_input_tokens_seen": 55534472,
      "step": 95645
    },
    {
      "epoch": 14.2463509085493,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.006948893679737628,
      "loss": 0.8144,
      "num_input_tokens_seen": 55537608,
      "step": 95650
    },
    {
      "epoch": 14.24709562109026,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.0069472487331552245,
      "loss": 0.7834,
      "num_input_tokens_seen": 55540520,
      "step": 95655
    },
    {
      "epoch": 14.247840333631219,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.006945603922615875,
      "loss": 0.8076,
      "num_input_tokens_seen": 55543240,
      "step": 95660
    },
    {
      "epoch": 14.248585046172177,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0069439592481473555,
      "loss": 0.7872,
      "num_input_tokens_seen": 55546120,
      "step": 95665
    },
    {
      "epoch": 14.249329758713136,
      "grad_norm": 0.037109375,
      "learning_rate": 0.006942314709777462,
      "loss": 0.8069,
      "num_input_tokens_seen": 55549000,
      "step": 95670
    },
    {
      "epoch": 14.250074471254097,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0069406703075339665,
      "loss": 0.8011,
      "num_input_tokens_seen": 55551624,
      "step": 95675
    },
    {
      "epoch": 14.250819183795056,
      "grad_norm": 0.033203125,
      "learning_rate": 0.00693902604144466,
      "loss": 0.808,
      "num_input_tokens_seen": 55554280,
      "step": 95680
    },
    {
      "epoch": 14.251563896336014,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.006937381911537316,
      "loss": 0.8218,
      "num_input_tokens_seen": 55557288,
      "step": 95685
    },
    {
      "epoch": 14.252308608876973,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.006935737917839711,
      "loss": 0.7974,
      "num_input_tokens_seen": 55559944,
      "step": 95690
    },
    {
      "epoch": 14.253053321417934,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.006934094060379617,
      "loss": 0.8149,
      "num_input_tokens_seen": 55563208,
      "step": 95695
    },
    {
      "epoch": 14.253798033958892,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.006932450339184801,
      "loss": 0.8381,
      "num_input_tokens_seen": 55565896,
      "step": 95700
    },
    {
      "epoch": 14.254542746499851,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.006930806754283043,
      "loss": 0.7997,
      "num_input_tokens_seen": 55568776,
      "step": 95705
    },
    {
      "epoch": 14.25528745904081,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.006929163305702097,
      "loss": 0.793,
      "num_input_tokens_seen": 55571688,
      "step": 95710
    },
    {
      "epoch": 14.256032171581769,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.00692751999346974,
      "loss": 0.7746,
      "num_input_tokens_seen": 55574568,
      "step": 95715
    },
    {
      "epoch": 14.25677688412273,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0069258768176137285,
      "loss": 0.8098,
      "num_input_tokens_seen": 55577448,
      "step": 95720
    },
    {
      "epoch": 14.257521596663688,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.0069242337781618155,
      "loss": 0.7977,
      "num_input_tokens_seen": 55580136,
      "step": 95725
    },
    {
      "epoch": 14.258266309204647,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.006922590875141772,
      "loss": 0.7741,
      "num_input_tokens_seen": 55583048,
      "step": 95730
    },
    {
      "epoch": 14.259011021745605,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.006920948108581338,
      "loss": 0.8049,
      "num_input_tokens_seen": 55585928,
      "step": 95735
    },
    {
      "epoch": 14.259755734286566,
      "grad_norm": 0.0390625,
      "learning_rate": 0.006919305478508283,
      "loss": 0.8122,
      "num_input_tokens_seen": 55588872,
      "step": 95740
    },
    {
      "epoch": 14.260500446827525,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0069176629849503415,
      "loss": 0.8027,
      "num_input_tokens_seen": 55592104,
      "step": 95745
    },
    {
      "epoch": 14.261245159368483,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.006916020627935276,
      "loss": 0.8173,
      "num_input_tokens_seen": 55594600,
      "step": 95750
    },
    {
      "epoch": 14.261989871909442,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.006914378407490826,
      "loss": 0.8201,
      "num_input_tokens_seen": 55597704,
      "step": 95755
    },
    {
      "epoch": 14.262734584450403,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.006912736323644734,
      "loss": 0.7939,
      "num_input_tokens_seen": 55600424,
      "step": 95760
    },
    {
      "epoch": 14.263479296991362,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.006911094376424743,
      "loss": 0.7886,
      "num_input_tokens_seen": 55603336,
      "step": 95765
    },
    {
      "epoch": 14.26422400953232,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0069094525658585865,
      "loss": 0.7922,
      "num_input_tokens_seen": 55606376,
      "step": 95770
    },
    {
      "epoch": 14.264968722073279,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.00690781089197401,
      "loss": 0.7899,
      "num_input_tokens_seen": 55609256,
      "step": 95775
    },
    {
      "epoch": 14.26571343461424,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.00690616935479874,
      "loss": 0.7923,
      "num_input_tokens_seen": 55612104,
      "step": 95780
    },
    {
      "epoch": 14.266458147155198,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0069045279543605165,
      "loss": 0.8575,
      "num_input_tokens_seen": 55614984,
      "step": 95785
    },
    {
      "epoch": 14.267202859696157,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.00690288669068706,
      "loss": 0.7979,
      "num_input_tokens_seen": 55618376,
      "step": 95790
    },
    {
      "epoch": 14.267947572237116,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.006901245563806111,
      "loss": 0.7773,
      "num_input_tokens_seen": 55621256,
      "step": 95795
    },
    {
      "epoch": 14.268692284778076,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.006899604573745384,
      "loss": 0.7785,
      "num_input_tokens_seen": 55624136,
      "step": 95800
    },
    {
      "epoch": 14.269436997319035,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.006897963720532601,
      "loss": 0.8678,
      "num_input_tokens_seen": 55626664,
      "step": 95805
    },
    {
      "epoch": 14.270181709859994,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.006896323004195491,
      "loss": 0.7922,
      "num_input_tokens_seen": 55629576,
      "step": 95810
    },
    {
      "epoch": 14.270926422400953,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.006894682424761763,
      "loss": 0.8065,
      "num_input_tokens_seen": 55632264,
      "step": 95815
    },
    {
      "epoch": 14.271671134941913,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.006893041982259142,
      "loss": 0.8139,
      "num_input_tokens_seen": 55634952,
      "step": 95820
    },
    {
      "epoch": 14.272415847482872,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.006891401676715339,
      "loss": 0.8005,
      "num_input_tokens_seen": 55637736,
      "step": 95825
    },
    {
      "epoch": 14.27316056002383,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.006889761508158062,
      "loss": 0.8521,
      "num_input_tokens_seen": 55640488,
      "step": 95830
    },
    {
      "epoch": 14.27390527256479,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.006888121476615018,
      "loss": 0.8114,
      "num_input_tokens_seen": 55643176,
      "step": 95835
    },
    {
      "epoch": 14.27464998510575,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.006886481582113919,
      "loss": 0.8034,
      "num_input_tokens_seen": 55645960,
      "step": 95840
    },
    {
      "epoch": 14.275394697646709,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.00688484182468247,
      "loss": 0.8058,
      "num_input_tokens_seen": 55648712,
      "step": 95845
    },
    {
      "epoch": 14.276139410187668,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.006883202204348364,
      "loss": 0.7976,
      "num_input_tokens_seen": 55651464,
      "step": 95850
    },
    {
      "epoch": 14.276884122728626,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.006881562721139313,
      "loss": 0.7884,
      "num_input_tokens_seen": 55654344,
      "step": 95855
    },
    {
      "epoch": 14.277628835269585,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.006879923375083003,
      "loss": 0.781,
      "num_input_tokens_seen": 55657064,
      "step": 95860
    },
    {
      "epoch": 14.278373547810546,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.006878284166207142,
      "loss": 0.7963,
      "num_input_tokens_seen": 55660296,
      "step": 95865
    },
    {
      "epoch": 14.279118260351504,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.006876645094539408,
      "loss": 0.8024,
      "num_input_tokens_seen": 55663464,
      "step": 95870
    },
    {
      "epoch": 14.279862972892463,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.006875006160107504,
      "loss": 0.7848,
      "num_input_tokens_seen": 55666600,
      "step": 95875
    },
    {
      "epoch": 14.280607685433422,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0068733673629391145,
      "loss": 0.8047,
      "num_input_tokens_seen": 55669640,
      "step": 95880
    },
    {
      "epoch": 14.281352397974382,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.006871728703061923,
      "loss": 0.7961,
      "num_input_tokens_seen": 55672328,
      "step": 95885
    },
    {
      "epoch": 14.282097110515341,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.006870090180503609,
      "loss": 0.784,
      "num_input_tokens_seen": 55675112,
      "step": 95890
    },
    {
      "epoch": 14.2828418230563,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.006868451795291864,
      "loss": 0.8059,
      "num_input_tokens_seen": 55677992,
      "step": 95895
    },
    {
      "epoch": 14.283586535597259,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.00686681354745436,
      "loss": 0.8102,
      "num_input_tokens_seen": 55680904,
      "step": 95900
    },
    {
      "epoch": 14.28433124813822,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.00686517543701877,
      "loss": 0.7941,
      "num_input_tokens_seen": 55683784,
      "step": 95905
    },
    {
      "epoch": 14.285075960679178,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.00686353746401278,
      "loss": 0.814,
      "num_input_tokens_seen": 55686792,
      "step": 95910
    },
    {
      "epoch": 14.285820673220137,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.006861899628464049,
      "loss": 0.7874,
      "num_input_tokens_seen": 55689800,
      "step": 95915
    },
    {
      "epoch": 14.286565385761095,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.006860261930400256,
      "loss": 0.809,
      "num_input_tokens_seen": 55692616,
      "step": 95920
    },
    {
      "epoch": 14.287310098302056,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.006858624369849068,
      "loss": 0.79,
      "num_input_tokens_seen": 55695816,
      "step": 95925
    },
    {
      "epoch": 14.288054810843015,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.006856986946838139,
      "loss": 0.7992,
      "num_input_tokens_seen": 55699080,
      "step": 95930
    },
    {
      "epoch": 14.288799523383974,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.006855349661395145,
      "loss": 0.7869,
      "num_input_tokens_seen": 55701992,
      "step": 95935
    },
    {
      "epoch": 14.289544235924932,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.006853712513547737,
      "loss": 0.7886,
      "num_input_tokens_seen": 55705032,
      "step": 95940
    },
    {
      "epoch": 14.290288948465893,
      "grad_norm": 0.064453125,
      "learning_rate": 0.00685207550332358,
      "loss": 0.7985,
      "num_input_tokens_seen": 55708104,
      "step": 95945
    },
    {
      "epoch": 14.291033661006852,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.006850438630750327,
      "loss": 0.7989,
      "num_input_tokens_seen": 55711496,
      "step": 95950
    },
    {
      "epoch": 14.29177837354781,
      "grad_norm": 0.0390625,
      "learning_rate": 0.006848801895855631,
      "loss": 0.8041,
      "num_input_tokens_seen": 55714280,
      "step": 95955
    },
    {
      "epoch": 14.292523086088769,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.006847165298667136,
      "loss": 0.7888,
      "num_input_tokens_seen": 55717160,
      "step": 95960
    },
    {
      "epoch": 14.29326779862973,
      "grad_norm": 0.0390625,
      "learning_rate": 0.006845528839212503,
      "loss": 0.8099,
      "num_input_tokens_seen": 55719816,
      "step": 95965
    },
    {
      "epoch": 14.294012511170688,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.006843892517519373,
      "loss": 0.8085,
      "num_input_tokens_seen": 55722760,
      "step": 95970
    },
    {
      "epoch": 14.294757223711647,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.006842256333615382,
      "loss": 0.7964,
      "num_input_tokens_seen": 55725608,
      "step": 95975
    },
    {
      "epoch": 14.295501936252606,
      "grad_norm": 0.044921875,
      "learning_rate": 0.006840620287528187,
      "loss": 0.802,
      "num_input_tokens_seen": 55728584,
      "step": 95980
    },
    {
      "epoch": 14.296246648793566,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.006838984379285412,
      "loss": 0.7634,
      "num_input_tokens_seen": 55731432,
      "step": 95985
    },
    {
      "epoch": 14.296991361334525,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.006837348608914708,
      "loss": 0.805,
      "num_input_tokens_seen": 55734440,
      "step": 95990
    },
    {
      "epoch": 14.297736073875484,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0068357129764436975,
      "loss": 0.7791,
      "num_input_tokens_seen": 55737128,
      "step": 95995
    },
    {
      "epoch": 14.298480786416443,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.006834077481900024,
      "loss": 0.8004,
      "num_input_tokens_seen": 55739752,
      "step": 96000
    },
    {
      "epoch": 14.299225498957403,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.006832442125311312,
      "loss": 0.7985,
      "num_input_tokens_seen": 55742792,
      "step": 96005
    },
    {
      "epoch": 14.299970211498362,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.006830806906705183,
      "loss": 0.796,
      "num_input_tokens_seen": 55745864,
      "step": 96010
    },
    {
      "epoch": 14.30071492403932,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.006829171826109275,
      "loss": 0.7783,
      "num_input_tokens_seen": 55748648,
      "step": 96015
    },
    {
      "epoch": 14.30145963658028,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.006827536883551204,
      "loss": 0.8111,
      "num_input_tokens_seen": 55751656,
      "step": 96020
    },
    {
      "epoch": 14.30220434912124,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00682590207905859,
      "loss": 0.8049,
      "num_input_tokens_seen": 55754664,
      "step": 96025
    },
    {
      "epoch": 14.302949061662199,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.006824267412659049,
      "loss": 0.7878,
      "num_input_tokens_seen": 55757640,
      "step": 96030
    },
    {
      "epoch": 14.303693774203158,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.006822632884380206,
      "loss": 0.8049,
      "num_input_tokens_seen": 55760712,
      "step": 96035
    },
    {
      "epoch": 14.304438486744116,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.006820998494249664,
      "loss": 0.7962,
      "num_input_tokens_seen": 55763848,
      "step": 96040
    },
    {
      "epoch": 14.305183199285075,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.006819364242295044,
      "loss": 0.7981,
      "num_input_tokens_seen": 55766856,
      "step": 96045
    },
    {
      "epoch": 14.305927911826036,
      "grad_norm": 0.064453125,
      "learning_rate": 0.006817730128543951,
      "loss": 0.7969,
      "num_input_tokens_seen": 55769544,
      "step": 96050
    },
    {
      "epoch": 14.306672624366994,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.006816096153023985,
      "loss": 0.8445,
      "num_input_tokens_seen": 55772392,
      "step": 96055
    },
    {
      "epoch": 14.307417336907953,
      "grad_norm": 0.0390625,
      "learning_rate": 0.006814462315762764,
      "loss": 0.792,
      "num_input_tokens_seen": 55775144,
      "step": 96060
    },
    {
      "epoch": 14.308162049448912,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.006812828616787877,
      "loss": 0.7864,
      "num_input_tokens_seen": 55777992,
      "step": 96065
    },
    {
      "epoch": 14.308906761989872,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.006811195056126935,
      "loss": 0.7888,
      "num_input_tokens_seen": 55780552,
      "step": 96070
    },
    {
      "epoch": 14.309651474530831,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.006809561633807522,
      "loss": 0.7885,
      "num_input_tokens_seen": 55783656,
      "step": 96075
    },
    {
      "epoch": 14.31039618707179,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.006807928349857249,
      "loss": 0.7927,
      "num_input_tokens_seen": 55786344,
      "step": 96080
    },
    {
      "epoch": 14.311140899612749,
      "grad_norm": 0.025390625,
      "learning_rate": 0.0068062952043037,
      "loss": 0.8073,
      "num_input_tokens_seen": 55789288,
      "step": 96085
    },
    {
      "epoch": 14.31188561215371,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.006804662197174465,
      "loss": 0.8214,
      "num_input_tokens_seen": 55792168,
      "step": 96090
    },
    {
      "epoch": 14.312630324694668,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.006803029328497133,
      "loss": 0.7934,
      "num_input_tokens_seen": 55795304,
      "step": 96095
    },
    {
      "epoch": 14.313375037235627,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.006801396598299284,
      "loss": 0.7926,
      "num_input_tokens_seen": 55797960,
      "step": 96100
    },
    {
      "epoch": 14.314119749776586,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.006799764006608513,
      "loss": 0.7886,
      "num_input_tokens_seen": 55801032,
      "step": 96105
    },
    {
      "epoch": 14.314864462317546,
      "grad_norm": 0.138671875,
      "learning_rate": 0.006798131553452389,
      "loss": 0.8409,
      "num_input_tokens_seen": 55803816,
      "step": 96110
    },
    {
      "epoch": 14.315609174858505,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.006796499238858501,
      "loss": 0.7885,
      "num_input_tokens_seen": 55806952,
      "step": 96115
    },
    {
      "epoch": 14.316353887399464,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.006794867062854416,
      "loss": 0.7932,
      "num_input_tokens_seen": 55809800,
      "step": 96120
    },
    {
      "epoch": 14.317098599940422,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.006793235025467717,
      "loss": 0.7859,
      "num_input_tokens_seen": 55812552,
      "step": 96125
    },
    {
      "epoch": 14.317843312481383,
      "grad_norm": 0.04296875,
      "learning_rate": 0.006791603126725974,
      "loss": 0.7654,
      "num_input_tokens_seen": 55815464,
      "step": 96130
    },
    {
      "epoch": 14.318588025022342,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.006789971366656745,
      "loss": 0.7864,
      "num_input_tokens_seen": 55818312,
      "step": 96135
    },
    {
      "epoch": 14.3193327375633,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.006788339745287613,
      "loss": 0.7885,
      "num_input_tokens_seen": 55821064,
      "step": 96140
    },
    {
      "epoch": 14.32007745010426,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.006786708262646129,
      "loss": 0.796,
      "num_input_tokens_seen": 55824104,
      "step": 96145
    },
    {
      "epoch": 14.32082216264522,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.006785076918759865,
      "loss": 0.8056,
      "num_input_tokens_seen": 55826728,
      "step": 96150
    },
    {
      "epoch": 14.321566875186178,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.006783445713656379,
      "loss": 0.7929,
      "num_input_tokens_seen": 55829544,
      "step": 96155
    },
    {
      "epoch": 14.322311587727137,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.006781814647363227,
      "loss": 0.7784,
      "num_input_tokens_seen": 55832488,
      "step": 96160
    },
    {
      "epoch": 14.323056300268096,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.006780183719907963,
      "loss": 0.7883,
      "num_input_tokens_seen": 55835624,
      "step": 96165
    },
    {
      "epoch": 14.323801012809056,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0067785529313181365,
      "loss": 0.8106,
      "num_input_tokens_seen": 55838408,
      "step": 96170
    },
    {
      "epoch": 14.324545725350015,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0067769222816213065,
      "loss": 0.8019,
      "num_input_tokens_seen": 55841448,
      "step": 96175
    },
    {
      "epoch": 14.325290437890974,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.006775291770845013,
      "loss": 0.7912,
      "num_input_tokens_seen": 55844904,
      "step": 96180
    },
    {
      "epoch": 14.326035150431933,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0067736613990168094,
      "loss": 0.812,
      "num_input_tokens_seen": 55848072,
      "step": 96185
    },
    {
      "epoch": 14.326779862972893,
      "grad_norm": 0.064453125,
      "learning_rate": 0.006772031166164231,
      "loss": 0.7852,
      "num_input_tokens_seen": 55850856,
      "step": 96190
    },
    {
      "epoch": 14.327524575513852,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.006770401072314828,
      "loss": 0.8218,
      "num_input_tokens_seen": 55854120,
      "step": 96195
    },
    {
      "epoch": 14.32826928805481,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.006768771117496131,
      "loss": 0.7973,
      "num_input_tokens_seen": 55857064,
      "step": 96200
    },
    {
      "epoch": 14.32901400059577,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.006767141301735684,
      "loss": 0.786,
      "num_input_tokens_seen": 55860296,
      "step": 96205
    },
    {
      "epoch": 14.32975871313673,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0067655116250610195,
      "loss": 0.8286,
      "num_input_tokens_seen": 55863016,
      "step": 96210
    },
    {
      "epoch": 14.330503425677689,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0067638820874996635,
      "loss": 0.7779,
      "num_input_tokens_seen": 55866088,
      "step": 96215
    },
    {
      "epoch": 14.331248138218648,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.006762252689079147,
      "loss": 0.8087,
      "num_input_tokens_seen": 55868904,
      "step": 96220
    },
    {
      "epoch": 14.331992850759606,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0067606234298270024,
      "loss": 0.7888,
      "num_input_tokens_seen": 55872008,
      "step": 96225
    },
    {
      "epoch": 14.332737563300565,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.006758994309770751,
      "loss": 0.7836,
      "num_input_tokens_seen": 55874888,
      "step": 96230
    },
    {
      "epoch": 14.333482275841526,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.0067573653289379094,
      "loss": 0.8019,
      "num_input_tokens_seen": 55877736,
      "step": 96235
    },
    {
      "epoch": 14.334226988382484,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.006755736487356008,
      "loss": 0.7973,
      "num_input_tokens_seen": 55880648,
      "step": 96240
    },
    {
      "epoch": 14.334971700923443,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.00675410778505256,
      "loss": 0.7942,
      "num_input_tokens_seen": 55883400,
      "step": 96245
    },
    {
      "epoch": 14.335716413464402,
      "grad_norm": 0.08984375,
      "learning_rate": 0.006752479222055076,
      "loss": 0.8216,
      "num_input_tokens_seen": 55886504,
      "step": 96250
    },
    {
      "epoch": 14.336461126005362,
      "grad_norm": 0.033203125,
      "learning_rate": 0.006750850798391078,
      "loss": 0.8077,
      "num_input_tokens_seen": 55889224,
      "step": 96255
    },
    {
      "epoch": 14.337205838546321,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.006749222514088066,
      "loss": 0.8263,
      "num_input_tokens_seen": 55892072,
      "step": 96260
    },
    {
      "epoch": 14.33795055108728,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.006747594369173558,
      "loss": 0.7928,
      "num_input_tokens_seen": 55895016,
      "step": 96265
    },
    {
      "epoch": 14.338695263628239,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.006745966363675053,
      "loss": 0.8073,
      "num_input_tokens_seen": 55898024,
      "step": 96270
    },
    {
      "epoch": 14.3394399761692,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0067443384976200605,
      "loss": 0.8209,
      "num_input_tokens_seen": 55900968,
      "step": 96275
    },
    {
      "epoch": 14.340184688710158,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0067427107710360795,
      "loss": 0.8159,
      "num_input_tokens_seen": 55903912,
      "step": 96280
    },
    {
      "epoch": 14.340929401251117,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0067410831839506075,
      "loss": 0.8179,
      "num_input_tokens_seen": 55906952,
      "step": 96285
    },
    {
      "epoch": 14.341674113792076,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00673945573639114,
      "loss": 0.782,
      "num_input_tokens_seen": 55909768,
      "step": 96290
    },
    {
      "epoch": 14.342418826333036,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.006737828428385168,
      "loss": 0.7899,
      "num_input_tokens_seen": 55912744,
      "step": 96295
    },
    {
      "epoch": 14.343163538873995,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.006736201259960193,
      "loss": 0.7862,
      "num_input_tokens_seen": 55915656,
      "step": 96300
    },
    {
      "epoch": 14.343908251414954,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.006734574231143693,
      "loss": 0.7822,
      "num_input_tokens_seen": 55918440,
      "step": 96305
    },
    {
      "epoch": 14.344652963955912,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.006732947341963165,
      "loss": 0.7916,
      "num_input_tokens_seen": 55921192,
      "step": 96310
    },
    {
      "epoch": 14.345397676496873,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.006731320592446087,
      "loss": 0.79,
      "num_input_tokens_seen": 55923848,
      "step": 96315
    },
    {
      "epoch": 14.346142389037832,
      "grad_norm": 0.03515625,
      "learning_rate": 0.006729693982619945,
      "loss": 0.8448,
      "num_input_tokens_seen": 55926728,
      "step": 96320
    },
    {
      "epoch": 14.34688710157879,
      "grad_norm": 0.060546875,
      "learning_rate": 0.006728067512512221,
      "loss": 0.8061,
      "num_input_tokens_seen": 55929832,
      "step": 96325
    },
    {
      "epoch": 14.34763181411975,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.006726441182150382,
      "loss": 0.7947,
      "num_input_tokens_seen": 55933128,
      "step": 96330
    },
    {
      "epoch": 14.34837652666071,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.006724814991561916,
      "loss": 0.8094,
      "num_input_tokens_seen": 55935816,
      "step": 96335
    },
    {
      "epoch": 14.349121239201668,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.006723188940774285,
      "loss": 0.8173,
      "num_input_tokens_seen": 55938696,
      "step": 96340
    },
    {
      "epoch": 14.349865951742627,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0067215630298149715,
      "loss": 0.8026,
      "num_input_tokens_seen": 55941640,
      "step": 96345
    },
    {
      "epoch": 14.350610664283586,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.006719937258711435,
      "loss": 0.8142,
      "num_input_tokens_seen": 55944808,
      "step": 96350
    },
    {
      "epoch": 14.351355376824547,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.006718311627491143,
      "loss": 0.7869,
      "num_input_tokens_seen": 55947560,
      "step": 96355
    },
    {
      "epoch": 14.352100089365505,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.006716686136181554,
      "loss": 0.8069,
      "num_input_tokens_seen": 55950536,
      "step": 96360
    },
    {
      "epoch": 14.352844801906464,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.00671506078481014,
      "loss": 0.7786,
      "num_input_tokens_seen": 55953576,
      "step": 96365
    },
    {
      "epoch": 14.353589514447423,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.006713435573404353,
      "loss": 0.7794,
      "num_input_tokens_seen": 55956456,
      "step": 96370
    },
    {
      "epoch": 14.354334226988382,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.006711810501991643,
      "loss": 0.7973,
      "num_input_tokens_seen": 55958984,
      "step": 96375
    },
    {
      "epoch": 14.355078939529342,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.006710185570599479,
      "loss": 0.7915,
      "num_input_tokens_seen": 55962248,
      "step": 96380
    },
    {
      "epoch": 14.3558236520703,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.006708560779255297,
      "loss": 0.8184,
      "num_input_tokens_seen": 55965128,
      "step": 96385
    },
    {
      "epoch": 14.35656836461126,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.006706936127986559,
      "loss": 0.8124,
      "num_input_tokens_seen": 55968136,
      "step": 96390
    },
    {
      "epoch": 14.357313077152218,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0067053116168207015,
      "loss": 0.8055,
      "num_input_tokens_seen": 55971176,
      "step": 96395
    },
    {
      "epoch": 14.358057789693179,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.006703687245785179,
      "loss": 0.7831,
      "num_input_tokens_seen": 55973992,
      "step": 96400
    },
    {
      "epoch": 14.358802502234138,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.00670206301490743,
      "loss": 0.7857,
      "num_input_tokens_seen": 55976968,
      "step": 96405
    },
    {
      "epoch": 14.359547214775096,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0067004389242148864,
      "loss": 0.7992,
      "num_input_tokens_seen": 55979912,
      "step": 96410
    },
    {
      "epoch": 14.360291927316055,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.006698814973734997,
      "loss": 0.7934,
      "num_input_tokens_seen": 55982664,
      "step": 96415
    },
    {
      "epoch": 14.361036639857016,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.006697191163495192,
      "loss": 0.805,
      "num_input_tokens_seen": 55985704,
      "step": 96420
    },
    {
      "epoch": 14.361781352397974,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.006695567493522904,
      "loss": 0.7831,
      "num_input_tokens_seen": 55988648,
      "step": 96425
    },
    {
      "epoch": 14.362526064938933,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.006693943963845557,
      "loss": 0.7938,
      "num_input_tokens_seen": 55991240,
      "step": 96430
    },
    {
      "epoch": 14.363270777479892,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.00669232057449059,
      "loss": 0.7839,
      "num_input_tokens_seen": 55993992,
      "step": 96435
    },
    {
      "epoch": 14.364015490020853,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.006690697325485417,
      "loss": 0.8184,
      "num_input_tokens_seen": 55996904,
      "step": 96440
    },
    {
      "epoch": 14.364760202561811,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.006689074216857475,
      "loss": 0.7903,
      "num_input_tokens_seen": 56000104,
      "step": 96445
    },
    {
      "epoch": 14.36550491510277,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.006687451248634177,
      "loss": 0.8207,
      "num_input_tokens_seen": 56003272,
      "step": 96450
    },
    {
      "epoch": 14.366249627643729,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.006685828420842936,
      "loss": 0.7771,
      "num_input_tokens_seen": 56005928,
      "step": 96455
    },
    {
      "epoch": 14.36699434018469,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.006684205733511178,
      "loss": 0.7992,
      "num_input_tokens_seen": 56008648,
      "step": 96460
    },
    {
      "epoch": 14.367739052725648,
      "grad_norm": 0.033203125,
      "learning_rate": 0.006682583186666307,
      "loss": 0.7942,
      "num_input_tokens_seen": 56011816,
      "step": 96465
    },
    {
      "epoch": 14.368483765266607,
      "grad_norm": 0.04296875,
      "learning_rate": 0.006680960780335745,
      "loss": 0.8026,
      "num_input_tokens_seen": 56014856,
      "step": 96470
    },
    {
      "epoch": 14.369228477807566,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.006679338514546892,
      "loss": 0.8224,
      "num_input_tokens_seen": 56017448,
      "step": 96475
    },
    {
      "epoch": 14.369973190348526,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.006677716389327162,
      "loss": 0.8034,
      "num_input_tokens_seen": 56020680,
      "step": 96480
    },
    {
      "epoch": 14.370717902889485,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.006676094404703957,
      "loss": 0.7981,
      "num_input_tokens_seen": 56023528,
      "step": 96485
    },
    {
      "epoch": 14.371462615430444,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.006674472560704675,
      "loss": 0.8135,
      "num_input_tokens_seen": 56026888,
      "step": 96490
    },
    {
      "epoch": 14.372207327971402,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.006672850857356717,
      "loss": 0.7966,
      "num_input_tokens_seen": 56029672,
      "step": 96495
    },
    {
      "epoch": 14.372952040512363,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0066712292946874766,
      "loss": 0.781,
      "num_input_tokens_seen": 56032744,
      "step": 96500
    },
    {
      "epoch": 14.373696753053322,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.0066696078727243565,
      "loss": 0.7897,
      "num_input_tokens_seen": 56035432,
      "step": 96505
    },
    {
      "epoch": 14.37444146559428,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0066679865914947385,
      "loss": 0.7984,
      "num_input_tokens_seen": 56038056,
      "step": 96510
    },
    {
      "epoch": 14.37518617813524,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.006666365451026026,
      "loss": 0.804,
      "num_input_tokens_seen": 56041128,
      "step": 96515
    },
    {
      "epoch": 14.3759308906762,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.006664744451345594,
      "loss": 0.7953,
      "num_input_tokens_seen": 56044040,
      "step": 96520
    },
    {
      "epoch": 14.376675603217159,
      "grad_norm": 0.044921875,
      "learning_rate": 0.006663123592480839,
      "loss": 0.8024,
      "num_input_tokens_seen": 56046952,
      "step": 96525
    },
    {
      "epoch": 14.377420315758117,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.006661502874459134,
      "loss": 0.8159,
      "num_input_tokens_seen": 56049768,
      "step": 96530
    },
    {
      "epoch": 14.378165028299076,
      "grad_norm": 0.0546875,
      "learning_rate": 0.006659882297307861,
      "loss": 0.8163,
      "num_input_tokens_seen": 56052712,
      "step": 96535
    },
    {
      "epoch": 14.378909740840037,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.006658261861054404,
      "loss": 0.7898,
      "num_input_tokens_seen": 56055752,
      "step": 96540
    },
    {
      "epoch": 14.379654453380995,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.006656641565726136,
      "loss": 0.7949,
      "num_input_tokens_seen": 56058600,
      "step": 96545
    },
    {
      "epoch": 14.380399165921954,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.006655021411350422,
      "loss": 0.7805,
      "num_input_tokens_seen": 56061512,
      "step": 96550
    },
    {
      "epoch": 14.381143878462913,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.006653401397954646,
      "loss": 0.7947,
      "num_input_tokens_seen": 56064360,
      "step": 96555
    },
    {
      "epoch": 14.381888591003872,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.006651781525566171,
      "loss": 0.7925,
      "num_input_tokens_seen": 56067528,
      "step": 96560
    },
    {
      "epoch": 14.382633303544832,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.006650161794212356,
      "loss": 0.8013,
      "num_input_tokens_seen": 56070536,
      "step": 96565
    },
    {
      "epoch": 14.383378016085791,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.006648542203920577,
      "loss": 0.7834,
      "num_input_tokens_seen": 56073448,
      "step": 96570
    },
    {
      "epoch": 14.38412272862675,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.006646922754718188,
      "loss": 0.7863,
      "num_input_tokens_seen": 56075976,
      "step": 96575
    },
    {
      "epoch": 14.384867441167708,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.006645303446632545,
      "loss": 0.7896,
      "num_input_tokens_seen": 56078568,
      "step": 96580
    },
    {
      "epoch": 14.385612153708669,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.006643684279691015,
      "loss": 0.7918,
      "num_input_tokens_seen": 56081832,
      "step": 96585
    },
    {
      "epoch": 14.386356866249628,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00664206525392094,
      "loss": 0.7881,
      "num_input_tokens_seen": 56084488,
      "step": 96590
    },
    {
      "epoch": 14.387101578790586,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.006640446369349684,
      "loss": 0.7918,
      "num_input_tokens_seen": 56087560,
      "step": 96595
    },
    {
      "epoch": 14.387846291331545,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.006638827626004584,
      "loss": 0.7975,
      "num_input_tokens_seen": 56090472,
      "step": 96600
    },
    {
      "epoch": 14.388591003872506,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.006637209023912998,
      "loss": 0.8212,
      "num_input_tokens_seen": 56093416,
      "step": 96605
    },
    {
      "epoch": 14.389335716413465,
      "grad_norm": 0.044921875,
      "learning_rate": 0.006635590563102267,
      "loss": 0.8172,
      "num_input_tokens_seen": 56096296,
      "step": 96610
    },
    {
      "epoch": 14.390080428954423,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.006633972243599733,
      "loss": 0.8215,
      "num_input_tokens_seen": 56099176,
      "step": 96615
    },
    {
      "epoch": 14.390825141495382,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.006632354065432735,
      "loss": 0.8029,
      "num_input_tokens_seen": 56101768,
      "step": 96620
    },
    {
      "epoch": 14.391569854036343,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.006630736028628604,
      "loss": 0.7841,
      "num_input_tokens_seen": 56104648,
      "step": 96625
    },
    {
      "epoch": 14.392314566577301,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.006629118133214688,
      "loss": 0.8183,
      "num_input_tokens_seen": 56107656,
      "step": 96630
    },
    {
      "epoch": 14.39305927911826,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0066275003792183075,
      "loss": 0.7726,
      "num_input_tokens_seen": 56110376,
      "step": 96635
    },
    {
      "epoch": 14.393803991659219,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.006625882766666803,
      "loss": 0.7846,
      "num_input_tokens_seen": 56113352,
      "step": 96640
    },
    {
      "epoch": 14.39454870420018,
      "grad_norm": 0.048828125,
      "learning_rate": 0.006624265295587494,
      "loss": 0.7934,
      "num_input_tokens_seen": 56117672,
      "step": 96645
    },
    {
      "epoch": 14.395293416741138,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.006622647966007716,
      "loss": 0.8107,
      "num_input_tokens_seen": 56120776,
      "step": 96650
    },
    {
      "epoch": 14.396038129282097,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.006621030777954785,
      "loss": 0.8083,
      "num_input_tokens_seen": 56123464,
      "step": 96655
    },
    {
      "epoch": 14.396782841823056,
      "grad_norm": 0.0546875,
      "learning_rate": 0.006619413731456018,
      "loss": 0.7874,
      "num_input_tokens_seen": 56126376,
      "step": 96660
    },
    {
      "epoch": 14.397527554364016,
      "grad_norm": 0.03515625,
      "learning_rate": 0.006617796826538745,
      "loss": 0.8135,
      "num_input_tokens_seen": 56129320,
      "step": 96665
    },
    {
      "epoch": 14.398272266904975,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.006616180063230268,
      "loss": 0.7907,
      "num_input_tokens_seen": 56132072,
      "step": 96670
    },
    {
      "epoch": 14.399016979445934,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.006614563441557915,
      "loss": 0.7821,
      "num_input_tokens_seen": 56134952,
      "step": 96675
    },
    {
      "epoch": 14.399761691986892,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.006612946961548988,
      "loss": 0.8131,
      "num_input_tokens_seen": 56138152,
      "step": 96680
    },
    {
      "epoch": 14.400506404527853,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.006611330623230799,
      "loss": 0.8616,
      "num_input_tokens_seen": 56141352,
      "step": 96685
    },
    {
      "epoch": 14.401251117068812,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0066097144266306545,
      "loss": 0.7965,
      "num_input_tokens_seen": 56143976,
      "step": 96690
    },
    {
      "epoch": 14.40199582960977,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.00660809837177585,
      "loss": 0.7865,
      "num_input_tokens_seen": 56146792,
      "step": 96695
    },
    {
      "epoch": 14.40274054215073,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0066064824586937005,
      "loss": 0.7944,
      "num_input_tokens_seen": 56149928,
      "step": 96700
    },
    {
      "epoch": 14.40348525469169,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.006604866687411493,
      "loss": 0.7866,
      "num_input_tokens_seen": 56152904,
      "step": 96705
    },
    {
      "epoch": 14.404229967232649,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.006603251057956537,
      "loss": 0.814,
      "num_input_tokens_seen": 56155912,
      "step": 96710
    },
    {
      "epoch": 14.404974679773607,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.006601635570356115,
      "loss": 0.8159,
      "num_input_tokens_seen": 56158856,
      "step": 96715
    },
    {
      "epoch": 14.405719392314566,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.006600020224637529,
      "loss": 0.7817,
      "num_input_tokens_seen": 56161672,
      "step": 96720
    },
    {
      "epoch": 14.406464104855527,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.006598405020828059,
      "loss": 0.8181,
      "num_input_tokens_seen": 56164648,
      "step": 96725
    },
    {
      "epoch": 14.407208817396485,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.006596789958955001,
      "loss": 0.783,
      "num_input_tokens_seen": 56167880,
      "step": 96730
    },
    {
      "epoch": 14.407953529937444,
      "grad_norm": 0.041015625,
      "learning_rate": 0.006595175039045637,
      "loss": 0.7991,
      "num_input_tokens_seen": 56170728,
      "step": 96735
    },
    {
      "epoch": 14.408698242478403,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.006593560261127245,
      "loss": 0.8276,
      "num_input_tokens_seen": 56173736,
      "step": 96740
    },
    {
      "epoch": 14.409442955019362,
      "grad_norm": 0.07421875,
      "learning_rate": 0.006591945625227111,
      "loss": 0.7966,
      "num_input_tokens_seen": 56176776,
      "step": 96745
    },
    {
      "epoch": 14.410187667560322,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.006590331131372512,
      "loss": 0.7931,
      "num_input_tokens_seen": 56179624,
      "step": 96750
    },
    {
      "epoch": 14.410932380101281,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.00658871677959072,
      "loss": 0.8131,
      "num_input_tokens_seen": 56182280,
      "step": 96755
    },
    {
      "epoch": 14.41167709264224,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.006587102569909004,
      "loss": 0.7807,
      "num_input_tokens_seen": 56184936,
      "step": 96760
    },
    {
      "epoch": 14.412421805183198,
      "grad_norm": 0.0625,
      "learning_rate": 0.0065854885023546444,
      "loss": 0.777,
      "num_input_tokens_seen": 56187784,
      "step": 96765
    },
    {
      "epoch": 14.413166517724159,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.006583874576954906,
      "loss": 0.7844,
      "num_input_tokens_seen": 56190472,
      "step": 96770
    },
    {
      "epoch": 14.413911230265118,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.006582260793737047,
      "loss": 0.8069,
      "num_input_tokens_seen": 56193576,
      "step": 96775
    },
    {
      "epoch": 14.414655942806077,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.006580647152728342,
      "loss": 0.7929,
      "num_input_tokens_seen": 56196584,
      "step": 96780
    },
    {
      "epoch": 14.415400655347035,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00657903365395604,
      "loss": 0.7645,
      "num_input_tokens_seen": 56199880,
      "step": 96785
    },
    {
      "epoch": 14.416145367887996,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.006577420297447412,
      "loss": 0.8167,
      "num_input_tokens_seen": 56203144,
      "step": 96790
    },
    {
      "epoch": 14.416890080428955,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.006575807083229701,
      "loss": 0.7905,
      "num_input_tokens_seen": 56206088,
      "step": 96795
    },
    {
      "epoch": 14.417634792969913,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.006574194011330175,
      "loss": 0.8207,
      "num_input_tokens_seen": 56208968,
      "step": 96800
    },
    {
      "epoch": 14.418379505510872,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00657258108177607,
      "loss": 0.7975,
      "num_input_tokens_seen": 56211624,
      "step": 96805
    },
    {
      "epoch": 14.419124218051833,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.006570968294594649,
      "loss": 0.7855,
      "num_input_tokens_seen": 56214344,
      "step": 96810
    },
    {
      "epoch": 14.419868930592791,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.006569355649813152,
      "loss": 0.7894,
      "num_input_tokens_seen": 56217064,
      "step": 96815
    },
    {
      "epoch": 14.42061364313375,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.006567743147458821,
      "loss": 0.802,
      "num_input_tokens_seen": 56220168,
      "step": 96820
    },
    {
      "epoch": 14.421358355674709,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0065661307875589025,
      "loss": 0.7897,
      "num_input_tokens_seen": 56222792,
      "step": 96825
    },
    {
      "epoch": 14.42210306821567,
      "grad_norm": 0.046875,
      "learning_rate": 0.006564518570140625,
      "loss": 0.7881,
      "num_input_tokens_seen": 56225480,
      "step": 96830
    },
    {
      "epoch": 14.422847780756628,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.006562906495231238,
      "loss": 0.7982,
      "num_input_tokens_seen": 56228296,
      "step": 96835
    },
    {
      "epoch": 14.423592493297587,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.006561294562857966,
      "loss": 0.8431,
      "num_input_tokens_seen": 56231336,
      "step": 96840
    },
    {
      "epoch": 14.424337205838546,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00655968277304805,
      "loss": 0.7833,
      "num_input_tokens_seen": 56234248,
      "step": 96845
    },
    {
      "epoch": 14.425081918379506,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.006558071125828717,
      "loss": 0.7835,
      "num_input_tokens_seen": 56237224,
      "step": 96850
    },
    {
      "epoch": 14.425826630920465,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.006556459621227185,
      "loss": 0.8138,
      "num_input_tokens_seen": 56240552,
      "step": 96855
    },
    {
      "epoch": 14.426571343461424,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.006554848259270692,
      "loss": 0.793,
      "num_input_tokens_seen": 56243528,
      "step": 96860
    },
    {
      "epoch": 14.427316056002383,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.00655323703998645,
      "loss": 0.7604,
      "num_input_tokens_seen": 56246184,
      "step": 96865
    },
    {
      "epoch": 14.428060768543343,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.0065516259634016884,
      "loss": 0.8092,
      "num_input_tokens_seen": 56248968,
      "step": 96870
    },
    {
      "epoch": 14.428805481084302,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0065500150295436195,
      "loss": 0.7899,
      "num_input_tokens_seen": 56252296,
      "step": 96875
    },
    {
      "epoch": 14.42955019362526,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.006548404238439452,
      "loss": 0.8135,
      "num_input_tokens_seen": 56255048,
      "step": 96880
    },
    {
      "epoch": 14.43029490616622,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.006546793590116411,
      "loss": 0.7949,
      "num_input_tokens_seen": 56257960,
      "step": 96885
    },
    {
      "epoch": 14.43103961870718,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.006545183084601703,
      "loss": 0.7883,
      "num_input_tokens_seen": 56260840,
      "step": 96890
    },
    {
      "epoch": 14.431784331248139,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0065435727219225305,
      "loss": 0.7984,
      "num_input_tokens_seen": 56263592,
      "step": 96895
    },
    {
      "epoch": 14.432529043789097,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.006541962502106099,
      "loss": 0.8074,
      "num_input_tokens_seen": 56266440,
      "step": 96900
    },
    {
      "epoch": 14.433273756330056,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.006540352425179618,
      "loss": 0.7917,
      "num_input_tokens_seen": 56269448,
      "step": 96905
    },
    {
      "epoch": 14.434018468871017,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.006538742491170281,
      "loss": 0.7979,
      "num_input_tokens_seen": 56272392,
      "step": 96910
    },
    {
      "epoch": 14.434763181411975,
      "grad_norm": 0.056640625,
      "learning_rate": 0.006537132700105295,
      "loss": 0.7699,
      "num_input_tokens_seen": 56274920,
      "step": 96915
    },
    {
      "epoch": 14.435507893952934,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.006535523052011845,
      "loss": 0.784,
      "num_input_tokens_seen": 56277928,
      "step": 96920
    },
    {
      "epoch": 14.436252606493893,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.006533913546917137,
      "loss": 0.8384,
      "num_input_tokens_seen": 56280552,
      "step": 96925
    },
    {
      "epoch": 14.436997319034852,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.006532304184848353,
      "loss": 0.7833,
      "num_input_tokens_seen": 56283592,
      "step": 96930
    },
    {
      "epoch": 14.437742031575812,
      "grad_norm": 0.046875,
      "learning_rate": 0.0065306949658326795,
      "loss": 0.7937,
      "num_input_tokens_seen": 56286600,
      "step": 96935
    },
    {
      "epoch": 14.438486744116771,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.006529085889897311,
      "loss": 0.8101,
      "num_input_tokens_seen": 56289096,
      "step": 96940
    },
    {
      "epoch": 14.43923145665773,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.006527476957069428,
      "loss": 0.8068,
      "num_input_tokens_seen": 56292072,
      "step": 96945
    },
    {
      "epoch": 14.439976169198689,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.006525868167376212,
      "loss": 0.7924,
      "num_input_tokens_seen": 56294920,
      "step": 96950
    },
    {
      "epoch": 14.440720881739649,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0065242595208448335,
      "loss": 0.8197,
      "num_input_tokens_seen": 56298184,
      "step": 96955
    },
    {
      "epoch": 14.441465594280608,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.00652265101750248,
      "loss": 0.7857,
      "num_input_tokens_seen": 56301256,
      "step": 96960
    },
    {
      "epoch": 14.442210306821567,
      "grad_norm": 0.046875,
      "learning_rate": 0.006521042657376319,
      "loss": 0.779,
      "num_input_tokens_seen": 56303944,
      "step": 96965
    },
    {
      "epoch": 14.442955019362525,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.006519434440493527,
      "loss": 0.786,
      "num_input_tokens_seen": 56306920,
      "step": 96970
    },
    {
      "epoch": 14.443699731903486,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.006517826366881274,
      "loss": 0.8238,
      "num_input_tokens_seen": 56309736,
      "step": 96975
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.006516218436566716,
      "loss": 0.7729,
      "num_input_tokens_seen": 56312616,
      "step": 96980
    },
    {
      "epoch": 14.445189156985403,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0065146106495770325,
      "loss": 0.7837,
      "num_input_tokens_seen": 56315752,
      "step": 96985
    },
    {
      "epoch": 14.445933869526362,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.006513003005939371,
      "loss": 0.7921,
      "num_input_tokens_seen": 56318536,
      "step": 96990
    },
    {
      "epoch": 14.446678582067323,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.006511395505680903,
      "loss": 0.7858,
      "num_input_tokens_seen": 56321160,
      "step": 96995
    },
    {
      "epoch": 14.447423294608281,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.006509788148828776,
      "loss": 0.8089,
      "num_input_tokens_seen": 56324392,
      "step": 97000
    },
    {
      "epoch": 14.44816800714924,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0065081809354101535,
      "loss": 0.7993,
      "num_input_tokens_seen": 56327272,
      "step": 97005
    },
    {
      "epoch": 14.448912719690199,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0065065738654521855,
      "loss": 0.7995,
      "num_input_tokens_seen": 56330216,
      "step": 97010
    },
    {
      "epoch": 14.44965743223116,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.006504966938982018,
      "loss": 0.8077,
      "num_input_tokens_seen": 56333000,
      "step": 97015
    },
    {
      "epoch": 14.450402144772118,
      "grad_norm": 0.025390625,
      "learning_rate": 0.0065033601560268,
      "loss": 0.7886,
      "num_input_tokens_seen": 56335944,
      "step": 97020
    },
    {
      "epoch": 14.451146857313077,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.006501753516613672,
      "loss": 0.7959,
      "num_input_tokens_seen": 56339048,
      "step": 97025
    },
    {
      "epoch": 14.451891569854036,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0065001470207697865,
      "loss": 0.819,
      "num_input_tokens_seen": 56341800,
      "step": 97030
    },
    {
      "epoch": 14.452636282394996,
      "grad_norm": 0.03515625,
      "learning_rate": 0.006498540668522272,
      "loss": 0.805,
      "num_input_tokens_seen": 56344520,
      "step": 97035
    },
    {
      "epoch": 14.453380994935955,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.006496934459898278,
      "loss": 0.789,
      "num_input_tokens_seen": 56347816,
      "step": 97040
    },
    {
      "epoch": 14.454125707476914,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0064953283949249286,
      "loss": 0.7795,
      "num_input_tokens_seen": 56350888,
      "step": 97045
    },
    {
      "epoch": 14.454870420017873,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.00649372247362937,
      "loss": 0.8019,
      "num_input_tokens_seen": 56353800,
      "step": 97050
    },
    {
      "epoch": 14.455615132558833,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0064921166960387215,
      "loss": 0.8022,
      "num_input_tokens_seen": 56356680,
      "step": 97055
    },
    {
      "epoch": 14.456359845099792,
      "grad_norm": 0.046875,
      "learning_rate": 0.006490511062180111,
      "loss": 0.7791,
      "num_input_tokens_seen": 56359624,
      "step": 97060
    },
    {
      "epoch": 14.45710455764075,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0064889055720806724,
      "loss": 0.7863,
      "num_input_tokens_seen": 56362632,
      "step": 97065
    },
    {
      "epoch": 14.45784927018171,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.006487300225767518,
      "loss": 0.7832,
      "num_input_tokens_seen": 56365352,
      "step": 97070
    },
    {
      "epoch": 14.458593982722668,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.00648569502326778,
      "loss": 0.7999,
      "num_input_tokens_seen": 56368200,
      "step": 97075
    },
    {
      "epoch": 14.459338695263629,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.006484089964608571,
      "loss": 0.8169,
      "num_input_tokens_seen": 56370888,
      "step": 97080
    },
    {
      "epoch": 14.460083407804587,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.006482485049817007,
      "loss": 0.8083,
      "num_input_tokens_seen": 56373928,
      "step": 97085
    },
    {
      "epoch": 14.460828120345546,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.006480880278920201,
      "loss": 0.7858,
      "num_input_tokens_seen": 56376552,
      "step": 97090
    },
    {
      "epoch": 14.461572832886505,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.006479275651945259,
      "loss": 0.7973,
      "num_input_tokens_seen": 56379400,
      "step": 97095
    },
    {
      "epoch": 14.462317545427466,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.006477671168919301,
      "loss": 0.7952,
      "num_input_tokens_seen": 56382408,
      "step": 97100
    },
    {
      "epoch": 14.463062257968424,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.00647606682986942,
      "loss": 0.7914,
      "num_input_tokens_seen": 56385000,
      "step": 97105
    },
    {
      "epoch": 14.463806970509383,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.006474462634822732,
      "loss": 0.7894,
      "num_input_tokens_seen": 56388200,
      "step": 97110
    },
    {
      "epoch": 14.464551683050342,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.006472858583806329,
      "loss": 0.8083,
      "num_input_tokens_seen": 56390984,
      "step": 97115
    },
    {
      "epoch": 14.465296395591302,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.006471254676847318,
      "loss": 0.7984,
      "num_input_tokens_seen": 56393704,
      "step": 97120
    },
    {
      "epoch": 14.466041108132261,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.006469650913972785,
      "loss": 0.7753,
      "num_input_tokens_seen": 56396616,
      "step": 97125
    },
    {
      "epoch": 14.46678582067322,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.006468047295209837,
      "loss": 0.8006,
      "num_input_tokens_seen": 56399528,
      "step": 97130
    },
    {
      "epoch": 14.467530533214179,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.006466443820585557,
      "loss": 0.7911,
      "num_input_tokens_seen": 56402312,
      "step": 97135
    },
    {
      "epoch": 14.46827524575514,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.0064648404901270305,
      "loss": 0.8099,
      "num_input_tokens_seen": 56405032,
      "step": 97140
    },
    {
      "epoch": 14.469019958296098,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0064632373038613555,
      "loss": 0.7889,
      "num_input_tokens_seen": 56407624,
      "step": 97145
    },
    {
      "epoch": 14.469764670837057,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.006461634261815608,
      "loss": 0.794,
      "num_input_tokens_seen": 56410568,
      "step": 97150
    },
    {
      "epoch": 14.470509383378015,
      "grad_norm": 0.052734375,
      "learning_rate": 0.006460031364016873,
      "loss": 0.799,
      "num_input_tokens_seen": 56413736,
      "step": 97155
    },
    {
      "epoch": 14.471254095918976,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.006458428610492222,
      "loss": 0.8087,
      "num_input_tokens_seen": 56416936,
      "step": 97160
    },
    {
      "epoch": 14.471998808459935,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.006456826001268746,
      "loss": 0.789,
      "num_input_tokens_seen": 56420008,
      "step": 97165
    },
    {
      "epoch": 14.472743521000893,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.006455223536373505,
      "loss": 0.8062,
      "num_input_tokens_seen": 56422760,
      "step": 97170
    },
    {
      "epoch": 14.473488233541852,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0064536212158335824,
      "loss": 0.8089,
      "num_input_tokens_seen": 56425832,
      "step": 97175
    },
    {
      "epoch": 14.474232946082813,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.006452019039676044,
      "loss": 0.7956,
      "num_input_tokens_seen": 56428552,
      "step": 97180
    },
    {
      "epoch": 14.474977658623772,
      "grad_norm": 0.041015625,
      "learning_rate": 0.006450417007927951,
      "loss": 0.8083,
      "num_input_tokens_seen": 56431272,
      "step": 97185
    },
    {
      "epoch": 14.47572237116473,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.006448815120616378,
      "loss": 0.837,
      "num_input_tokens_seen": 56434056,
      "step": 97190
    },
    {
      "epoch": 14.476467083705689,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0064472133777683785,
      "loss": 0.8136,
      "num_input_tokens_seen": 56437064,
      "step": 97195
    },
    {
      "epoch": 14.47721179624665,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.00644561177941102,
      "loss": 0.7893,
      "num_input_tokens_seen": 56439944,
      "step": 97200
    },
    {
      "epoch": 14.477956508787608,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.006444010325571357,
      "loss": 0.7929,
      "num_input_tokens_seen": 56442856,
      "step": 97205
    },
    {
      "epoch": 14.478701221328567,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.006442409016276439,
      "loss": 0.7899,
      "num_input_tokens_seen": 56445864,
      "step": 97210
    },
    {
      "epoch": 14.479445933869526,
      "grad_norm": 0.0220947265625,
      "learning_rate": 0.006440807851553328,
      "loss": 0.7962,
      "num_input_tokens_seen": 56448648,
      "step": 97215
    },
    {
      "epoch": 14.480190646410486,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.006439206831429069,
      "loss": 0.7974,
      "num_input_tokens_seen": 56451304,
      "step": 97220
    },
    {
      "epoch": 14.480935358951445,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0064376059559307095,
      "loss": 0.8145,
      "num_input_tokens_seen": 56454312,
      "step": 97225
    },
    {
      "epoch": 14.481680071492404,
      "grad_norm": 0.0390625,
      "learning_rate": 0.00643600522508529,
      "loss": 0.7899,
      "num_input_tokens_seen": 56457608,
      "step": 97230
    },
    {
      "epoch": 14.482424784033363,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0064344046389198635,
      "loss": 0.7871,
      "num_input_tokens_seen": 56460584,
      "step": 97235
    },
    {
      "epoch": 14.483169496574323,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.006432804197461458,
      "loss": 0.797,
      "num_input_tokens_seen": 56463624,
      "step": 97240
    },
    {
      "epoch": 14.483914209115282,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.006431203900737125,
      "loss": 0.7851,
      "num_input_tokens_seen": 56466920,
      "step": 97245
    },
    {
      "epoch": 14.48465892165624,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.006429603748773888,
      "loss": 0.7912,
      "num_input_tokens_seen": 56470056,
      "step": 97250
    },
    {
      "epoch": 14.4854036341972,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.00642800374159879,
      "loss": 0.8119,
      "num_input_tokens_seen": 56473224,
      "step": 97255
    },
    {
      "epoch": 14.486148346738158,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.006426403879238858,
      "loss": 0.788,
      "num_input_tokens_seen": 56476104,
      "step": 97260
    },
    {
      "epoch": 14.486893059279119,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0064248041617211114,
      "loss": 0.7945,
      "num_input_tokens_seen": 56479272,
      "step": 97265
    },
    {
      "epoch": 14.487637771820078,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0064232045890725885,
      "loss": 0.8142,
      "num_input_tokens_seen": 56482280,
      "step": 97270
    },
    {
      "epoch": 14.488382484361036,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.006421605161320307,
      "loss": 0.787,
      "num_input_tokens_seen": 56485032,
      "step": 97275
    },
    {
      "epoch": 14.489127196901995,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.00642000587849129,
      "loss": 0.7781,
      "num_input_tokens_seen": 56488072,
      "step": 97280
    },
    {
      "epoch": 14.489871909442956,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.006418406740612545,
      "loss": 0.8123,
      "num_input_tokens_seen": 56490920,
      "step": 97285
    },
    {
      "epoch": 14.490616621983914,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.006416807747711104,
      "loss": 0.7913,
      "num_input_tokens_seen": 56493672,
      "step": 97290
    },
    {
      "epoch": 14.491361334524873,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.006415208899813969,
      "loss": 0.8077,
      "num_input_tokens_seen": 56496424,
      "step": 97295
    },
    {
      "epoch": 14.492106047065832,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.006413610196948152,
      "loss": 0.7862,
      "num_input_tokens_seen": 56499144,
      "step": 97300
    },
    {
      "epoch": 14.492850759606792,
      "grad_norm": 0.03515625,
      "learning_rate": 0.00641201163914067,
      "loss": 0.8003,
      "num_input_tokens_seen": 56502056,
      "step": 97305
    },
    {
      "epoch": 14.493595472147751,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.006410413226418516,
      "loss": 0.8022,
      "num_input_tokens_seen": 56504648,
      "step": 97310
    },
    {
      "epoch": 14.49434018468871,
      "grad_norm": 0.044921875,
      "learning_rate": 0.006408814958808706,
      "loss": 0.7873,
      "num_input_tokens_seen": 56507496,
      "step": 97315
    },
    {
      "epoch": 14.495084897229669,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00640721683633823,
      "loss": 0.8312,
      "num_input_tokens_seen": 56510312,
      "step": 97320
    },
    {
      "epoch": 14.49582960977063,
      "grad_norm": 0.037109375,
      "learning_rate": 0.006405618859034097,
      "loss": 0.8081,
      "num_input_tokens_seen": 56513288,
      "step": 97325
    },
    {
      "epoch": 14.496574322311588,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0064040210269232945,
      "loss": 0.7928,
      "num_input_tokens_seen": 56515944,
      "step": 97330
    },
    {
      "epoch": 14.497319034852547,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.006402423340032825,
      "loss": 0.8154,
      "num_input_tokens_seen": 56518888,
      "step": 97335
    },
    {
      "epoch": 14.498063747393505,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.006400825798389676,
      "loss": 0.7953,
      "num_input_tokens_seen": 56521736,
      "step": 97340
    },
    {
      "epoch": 14.498808459934466,
      "grad_norm": 0.033203125,
      "learning_rate": 0.006399228402020834,
      "loss": 0.7897,
      "num_input_tokens_seen": 56524328,
      "step": 97345
    },
    {
      "epoch": 14.499553172475425,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.006397631150953287,
      "loss": 0.8039,
      "num_input_tokens_seen": 56527336,
      "step": 97350
    },
    {
      "epoch": 14.500297885016384,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.006396034045214012,
      "loss": 0.7947,
      "num_input_tokens_seen": 56530088,
      "step": 97355
    },
    {
      "epoch": 14.501042597557342,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.006394437084830004,
      "loss": 0.7887,
      "num_input_tokens_seen": 56532904,
      "step": 97360
    },
    {
      "epoch": 14.501787310098303,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.00639284026982823,
      "loss": 0.8015,
      "num_input_tokens_seen": 56536040,
      "step": 97365
    },
    {
      "epoch": 14.502532022639262,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.006391243600235677,
      "loss": 0.793,
      "num_input_tokens_seen": 56538952,
      "step": 97370
    },
    {
      "epoch": 14.50327673518022,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.006389647076079314,
      "loss": 0.827,
      "num_input_tokens_seen": 56541928,
      "step": 97375
    },
    {
      "epoch": 14.504021447721179,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0063880506973861065,
      "loss": 0.8026,
      "num_input_tokens_seen": 56544808,
      "step": 97380
    },
    {
      "epoch": 14.50476616026214,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0063864544641830355,
      "loss": 0.8028,
      "num_input_tokens_seen": 56547400,
      "step": 97385
    },
    {
      "epoch": 14.505510872803098,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.006384858376497056,
      "loss": 0.8125,
      "num_input_tokens_seen": 56550056,
      "step": 97390
    },
    {
      "epoch": 14.506255585344057,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.006383262434355144,
      "loss": 0.8048,
      "num_input_tokens_seen": 56552872,
      "step": 97395
    },
    {
      "epoch": 14.507000297885016,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.006381666637784249,
      "loss": 0.7795,
      "num_input_tokens_seen": 56555848,
      "step": 97400
    },
    {
      "epoch": 14.507745010425975,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0063800709868113425,
      "loss": 0.7948,
      "num_input_tokens_seen": 56558952,
      "step": 97405
    },
    {
      "epoch": 14.508489722966935,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.006378475481463378,
      "loss": 0.7753,
      "num_input_tokens_seen": 56562088,
      "step": 97410
    },
    {
      "epoch": 14.509234435507894,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.006376880121767304,
      "loss": 0.7942,
      "num_input_tokens_seen": 56564872,
      "step": 97415
    },
    {
      "epoch": 14.509979148048853,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.006375284907750079,
      "loss": 0.7907,
      "num_input_tokens_seen": 56567848,
      "step": 97420
    },
    {
      "epoch": 14.510723860589813,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0063736898394386446,
      "loss": 0.7905,
      "num_input_tokens_seen": 56570696,
      "step": 97425
    },
    {
      "epoch": 14.511468573130772,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.006372094916859958,
      "loss": 0.8035,
      "num_input_tokens_seen": 56573448,
      "step": 97430
    },
    {
      "epoch": 14.51221328567173,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.006370500140040951,
      "loss": 0.7953,
      "num_input_tokens_seen": 56576424,
      "step": 97435
    },
    {
      "epoch": 14.51295799821269,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.006368905509008581,
      "loss": 0.7822,
      "num_input_tokens_seen": 56579432,
      "step": 97440
    },
    {
      "epoch": 14.513702710753648,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0063673110237897745,
      "loss": 0.7968,
      "num_input_tokens_seen": 56582728,
      "step": 97445
    },
    {
      "epoch": 14.514447423294609,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.00636571668441148,
      "loss": 0.8035,
      "num_input_tokens_seen": 56585672,
      "step": 97450
    },
    {
      "epoch": 14.515192135835568,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0063641224909006255,
      "loss": 0.788,
      "num_input_tokens_seen": 56588776,
      "step": 97455
    },
    {
      "epoch": 14.515936848376526,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.0063625284432841395,
      "loss": 0.8128,
      "num_input_tokens_seen": 56591656,
      "step": 97460
    },
    {
      "epoch": 14.516681560917485,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.006360934541588964,
      "loss": 0.8024,
      "num_input_tokens_seen": 56594600,
      "step": 97465
    },
    {
      "epoch": 14.517426273458446,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.006359340785842011,
      "loss": 0.7856,
      "num_input_tokens_seen": 56597448,
      "step": 97470
    },
    {
      "epoch": 14.518170985999404,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0063577471760702204,
      "loss": 0.8042,
      "num_input_tokens_seen": 56600424,
      "step": 97475
    },
    {
      "epoch": 14.518915698540363,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0063561537123005074,
      "loss": 0.796,
      "num_input_tokens_seen": 56603464,
      "step": 97480
    },
    {
      "epoch": 14.519660411081322,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.006354560394559792,
      "loss": 0.8002,
      "num_input_tokens_seen": 56606344,
      "step": 97485
    },
    {
      "epoch": 14.520405123622282,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.006352967222874987,
      "loss": 0.7733,
      "num_input_tokens_seen": 56609384,
      "step": 97490
    },
    {
      "epoch": 14.521149836163241,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.006351374197273017,
      "loss": 0.7709,
      "num_input_tokens_seen": 56612008,
      "step": 97495
    },
    {
      "epoch": 14.5218945487042,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00634978131778079,
      "loss": 0.7807,
      "num_input_tokens_seen": 56614824,
      "step": 97500
    },
    {
      "epoch": 14.522639261245159,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.006348188584425211,
      "loss": 0.8023,
      "num_input_tokens_seen": 56617576,
      "step": 97505
    },
    {
      "epoch": 14.52338397378612,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.006346595997233198,
      "loss": 0.7763,
      "num_input_tokens_seen": 56620584,
      "step": 97510
    },
    {
      "epoch": 14.524128686327078,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.006345003556231644,
      "loss": 0.7973,
      "num_input_tokens_seen": 56623272,
      "step": 97515
    },
    {
      "epoch": 14.524873398868037,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0063434112614474646,
      "loss": 0.7934,
      "num_input_tokens_seen": 56626088,
      "step": 97520
    },
    {
      "epoch": 14.525618111408996,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.006341819112907549,
      "loss": 0.8047,
      "num_input_tokens_seen": 56628904,
      "step": 97525
    },
    {
      "epoch": 14.526362823949956,
      "grad_norm": 0.056640625,
      "learning_rate": 0.006340227110638805,
      "loss": 0.7911,
      "num_input_tokens_seen": 56631624,
      "step": 97530
    },
    {
      "epoch": 14.527107536490915,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.006338635254668121,
      "loss": 0.8042,
      "num_input_tokens_seen": 56634536,
      "step": 97535
    },
    {
      "epoch": 14.527852249031874,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.006337043545022387,
      "loss": 0.802,
      "num_input_tokens_seen": 56637384,
      "step": 97540
    },
    {
      "epoch": 14.528596961572832,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.006335451981728504,
      "loss": 0.7851,
      "num_input_tokens_seen": 56640072,
      "step": 97545
    },
    {
      "epoch": 14.529341674113793,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.006333860564813353,
      "loss": 0.8102,
      "num_input_tokens_seen": 56643112,
      "step": 97550
    },
    {
      "epoch": 14.530086386654752,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.00633226929430382,
      "loss": 0.8047,
      "num_input_tokens_seen": 56646024,
      "step": 97555
    },
    {
      "epoch": 14.53083109919571,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.006330678170226781,
      "loss": 0.799,
      "num_input_tokens_seen": 56649032,
      "step": 97560
    },
    {
      "epoch": 14.53157581173667,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.00632908719260913,
      "loss": 0.8032,
      "num_input_tokens_seen": 56651752,
      "step": 97565
    },
    {
      "epoch": 14.53232052427763,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0063274963614777325,
      "loss": 0.8049,
      "num_input_tokens_seen": 56654536,
      "step": 97570
    },
    {
      "epoch": 14.533065236818588,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.0063259056768594744,
      "loss": 0.7839,
      "num_input_tokens_seen": 56657256,
      "step": 97575
    },
    {
      "epoch": 14.533809949359547,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.006324315138781225,
      "loss": 0.7981,
      "num_input_tokens_seen": 56659880,
      "step": 97580
    },
    {
      "epoch": 14.534554661900506,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.006322724747269848,
      "loss": 0.7844,
      "num_input_tokens_seen": 56662760,
      "step": 97585
    },
    {
      "epoch": 14.535299374441465,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.006321134502352222,
      "loss": 0.8,
      "num_input_tokens_seen": 56665672,
      "step": 97590
    },
    {
      "epoch": 14.536044086982425,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.006319544404055202,
      "loss": 0.7965,
      "num_input_tokens_seen": 56668712,
      "step": 97595
    },
    {
      "epoch": 14.536788799523384,
      "grad_norm": 0.041015625,
      "learning_rate": 0.006317954452405664,
      "loss": 0.7961,
      "num_input_tokens_seen": 56671496,
      "step": 97600
    },
    {
      "epoch": 14.537533512064343,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00631636464743046,
      "loss": 0.8392,
      "num_input_tokens_seen": 56674312,
      "step": 97605
    },
    {
      "epoch": 14.538278224605303,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00631477498915645,
      "loss": 0.7876,
      "num_input_tokens_seen": 56677192,
      "step": 97610
    },
    {
      "epoch": 14.539022937146262,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.006313185477610489,
      "loss": 0.808,
      "num_input_tokens_seen": 56680168,
      "step": 97615
    },
    {
      "epoch": 14.53976764968722,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.006311596112819426,
      "loss": 0.7979,
      "num_input_tokens_seen": 56682728,
      "step": 97620
    },
    {
      "epoch": 14.54051236222818,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00631000689481012,
      "loss": 0.8007,
      "num_input_tokens_seen": 56685608,
      "step": 97625
    },
    {
      "epoch": 14.541257074769138,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.006308417823609412,
      "loss": 0.799,
      "num_input_tokens_seen": 56688776,
      "step": 97630
    },
    {
      "epoch": 14.542001787310099,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.006306828899244156,
      "loss": 0.7827,
      "num_input_tokens_seen": 56691528,
      "step": 97635
    },
    {
      "epoch": 14.542746499851058,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.006305240121741184,
      "loss": 0.7998,
      "num_input_tokens_seen": 56694568,
      "step": 97640
    },
    {
      "epoch": 14.543491212392016,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.006303651491127349,
      "loss": 0.791,
      "num_input_tokens_seen": 56697768,
      "step": 97645
    },
    {
      "epoch": 14.544235924932975,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.006302063007429479,
      "loss": 0.8094,
      "num_input_tokens_seen": 56700936,
      "step": 97650
    },
    {
      "epoch": 14.544980637473936,
      "grad_norm": 0.05078125,
      "learning_rate": 0.006300474670674418,
      "loss": 0.7975,
      "num_input_tokens_seen": 56703560,
      "step": 97655
    },
    {
      "epoch": 14.545725350014894,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.006298886480888997,
      "loss": 0.7984,
      "num_input_tokens_seen": 56706568,
      "step": 97660
    },
    {
      "epoch": 14.546470062555853,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0062972984381000395,
      "loss": 0.8196,
      "num_input_tokens_seen": 56709512,
      "step": 97665
    },
    {
      "epoch": 14.547214775096812,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.006295710542334386,
      "loss": 0.8036,
      "num_input_tokens_seen": 56712616,
      "step": 97670
    },
    {
      "epoch": 14.547959487637772,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.006294122793618856,
      "loss": 0.7698,
      "num_input_tokens_seen": 56715272,
      "step": 97675
    },
    {
      "epoch": 14.548704200178731,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.006292535191980273,
      "loss": 0.8224,
      "num_input_tokens_seen": 56718248,
      "step": 97680
    },
    {
      "epoch": 14.54944891271969,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.006290947737445453,
      "loss": 0.797,
      "num_input_tokens_seen": 56720808,
      "step": 97685
    },
    {
      "epoch": 14.550193625260649,
      "grad_norm": 0.037109375,
      "learning_rate": 0.006289360430041225,
      "loss": 0.7828,
      "num_input_tokens_seen": 56723752,
      "step": 97690
    },
    {
      "epoch": 14.55093833780161,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.006287773269794398,
      "loss": 0.7734,
      "num_input_tokens_seen": 56726792,
      "step": 97695
    },
    {
      "epoch": 14.551683050342568,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.006286186256731782,
      "loss": 0.8083,
      "num_input_tokens_seen": 56729576,
      "step": 97700
    },
    {
      "epoch": 14.552427762883527,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.006284599390880198,
      "loss": 0.808,
      "num_input_tokens_seen": 56732424,
      "step": 97705
    },
    {
      "epoch": 14.553172475424486,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.006283012672266445,
      "loss": 0.8025,
      "num_input_tokens_seen": 56735208,
      "step": 97710
    },
    {
      "epoch": 14.553917187965446,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.006281426100917339,
      "loss": 0.7993,
      "num_input_tokens_seen": 56737992,
      "step": 97715
    },
    {
      "epoch": 14.554661900506405,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0062798396768596705,
      "loss": 0.825,
      "num_input_tokens_seen": 56740968,
      "step": 97720
    },
    {
      "epoch": 14.555406613047364,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.006278253400120255,
      "loss": 0.846,
      "num_input_tokens_seen": 56743880,
      "step": 97725
    },
    {
      "epoch": 14.556151325588322,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.006276667270725878,
      "loss": 0.7916,
      "num_input_tokens_seen": 56746568,
      "step": 97730
    },
    {
      "epoch": 14.556896038129283,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.006275081288703347,
      "loss": 0.8103,
      "num_input_tokens_seen": 56749448,
      "step": 97735
    },
    {
      "epoch": 14.557640750670242,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.00627349545407945,
      "loss": 0.7863,
      "num_input_tokens_seen": 56752392,
      "step": 97740
    },
    {
      "epoch": 14.5583854632112,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.006271909766880979,
      "loss": 0.7851,
      "num_input_tokens_seen": 56755624,
      "step": 97745
    },
    {
      "epoch": 14.55913017575216,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.006270324227134721,
      "loss": 0.7961,
      "num_input_tokens_seen": 56758792,
      "step": 97750
    },
    {
      "epoch": 14.55987488829312,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.006268738834867455,
      "loss": 0.8254,
      "num_input_tokens_seen": 56762184,
      "step": 97755
    },
    {
      "epoch": 14.560619600834078,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.006267153590105981,
      "loss": 0.7979,
      "num_input_tokens_seen": 56764872,
      "step": 97760
    },
    {
      "epoch": 14.561364313375037,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.006265568492877064,
      "loss": 0.7832,
      "num_input_tokens_seen": 56767848,
      "step": 97765
    },
    {
      "epoch": 14.562109025915996,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.006263983543207497,
      "loss": 0.7933,
      "num_input_tokens_seen": 56770888,
      "step": 97770
    },
    {
      "epoch": 14.562853738456955,
      "grad_norm": 0.033203125,
      "learning_rate": 0.006262398741124042,
      "loss": 0.8077,
      "num_input_tokens_seen": 56773608,
      "step": 97775
    },
    {
      "epoch": 14.563598450997915,
      "grad_norm": 0.041015625,
      "learning_rate": 0.006260814086653487,
      "loss": 0.7888,
      "num_input_tokens_seen": 56776680,
      "step": 97780
    },
    {
      "epoch": 14.564343163538874,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.006259229579822594,
      "loss": 0.7938,
      "num_input_tokens_seen": 56779432,
      "step": 97785
    },
    {
      "epoch": 14.565087876079833,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.006257645220658131,
      "loss": 0.8002,
      "num_input_tokens_seen": 56782120,
      "step": 97790
    },
    {
      "epoch": 14.565832588620792,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.006256061009186869,
      "loss": 0.7815,
      "num_input_tokens_seen": 56785064,
      "step": 97795
    },
    {
      "epoch": 14.566577301161752,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.006254476945435566,
      "loss": 0.8211,
      "num_input_tokens_seen": 56787848,
      "step": 97800
    },
    {
      "epoch": 14.56732201370271,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.006252893029430991,
      "loss": 0.8041,
      "num_input_tokens_seen": 56790536,
      "step": 97805
    },
    {
      "epoch": 14.56806672624367,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.006251309261199898,
      "loss": 0.7897,
      "num_input_tokens_seen": 56793416,
      "step": 97810
    },
    {
      "epoch": 14.568811438784628,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.006249725640769044,
      "loss": 0.8053,
      "num_input_tokens_seen": 56796136,
      "step": 97815
    },
    {
      "epoch": 14.569556151325589,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00624814216816518,
      "loss": 0.7899,
      "num_input_tokens_seen": 56799240,
      "step": 97820
    },
    {
      "epoch": 14.570300863866548,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.006246558843415055,
      "loss": 0.8,
      "num_input_tokens_seen": 56802184,
      "step": 97825
    },
    {
      "epoch": 14.571045576407506,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.006244975666545427,
      "loss": 0.7879,
      "num_input_tokens_seen": 56804840,
      "step": 97830
    },
    {
      "epoch": 14.571790288948465,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.006243392637583032,
      "loss": 0.7958,
      "num_input_tokens_seen": 56807624,
      "step": 97835
    },
    {
      "epoch": 14.572535001489426,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.006241809756554623,
      "loss": 0.8015,
      "num_input_tokens_seen": 56810600,
      "step": 97840
    },
    {
      "epoch": 14.573279714030384,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.006240227023486931,
      "loss": 0.7864,
      "num_input_tokens_seen": 56813416,
      "step": 97845
    },
    {
      "epoch": 14.574024426571343,
      "grad_norm": 0.044921875,
      "learning_rate": 0.006238644438406707,
      "loss": 0.8081,
      "num_input_tokens_seen": 56816200,
      "step": 97850
    },
    {
      "epoch": 14.574769139112302,
      "grad_norm": 0.033203125,
      "learning_rate": 0.006237062001340676,
      "loss": 0.8045,
      "num_input_tokens_seen": 56819368,
      "step": 97855
    },
    {
      "epoch": 14.575513851653263,
      "grad_norm": 0.0390625,
      "learning_rate": 0.00623547971231558,
      "loss": 0.7715,
      "num_input_tokens_seen": 56822216,
      "step": 97860
    },
    {
      "epoch": 14.576258564194221,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.006233897571358148,
      "loss": 0.8057,
      "num_input_tokens_seen": 56825064,
      "step": 97865
    },
    {
      "epoch": 14.57700327673518,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0062323155784951,
      "loss": 0.8012,
      "num_input_tokens_seen": 56828264,
      "step": 97870
    },
    {
      "epoch": 14.577747989276139,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.006230733733753175,
      "loss": 0.7672,
      "num_input_tokens_seen": 56831176,
      "step": 97875
    },
    {
      "epoch": 14.5784927018171,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.006229152037159091,
      "loss": 0.7782,
      "num_input_tokens_seen": 56834216,
      "step": 97880
    },
    {
      "epoch": 14.579237414358058,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.006227570488739569,
      "loss": 0.8002,
      "num_input_tokens_seen": 56837160,
      "step": 97885
    },
    {
      "epoch": 14.579982126899017,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0062259890885213226,
      "loss": 0.7962,
      "num_input_tokens_seen": 56840200,
      "step": 97890
    },
    {
      "epoch": 14.580726839439976,
      "grad_norm": 0.05078125,
      "learning_rate": 0.006224407836531079,
      "loss": 0.7992,
      "num_input_tokens_seen": 56843016,
      "step": 97895
    },
    {
      "epoch": 14.581471551980936,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.006222826732795545,
      "loss": 0.8099,
      "num_input_tokens_seen": 56845992,
      "step": 97900
    },
    {
      "epoch": 14.582216264521895,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.006221245777341426,
      "loss": 0.7822,
      "num_input_tokens_seen": 56849032,
      "step": 97905
    },
    {
      "epoch": 14.582960977062854,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.006219664970195445,
      "loss": 0.8065,
      "num_input_tokens_seen": 56851912,
      "step": 97910
    },
    {
      "epoch": 14.583705689603812,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.006218084311384294,
      "loss": 0.794,
      "num_input_tokens_seen": 56854728,
      "step": 97915
    },
    {
      "epoch": 14.584450402144771,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.006216503800934688,
      "loss": 0.8414,
      "num_input_tokens_seen": 56857352,
      "step": 97920
    },
    {
      "epoch": 14.585195114685732,
      "grad_norm": 0.025390625,
      "learning_rate": 0.006214923438873318,
      "loss": 0.8061,
      "num_input_tokens_seen": 56860200,
      "step": 97925
    },
    {
      "epoch": 14.58593982722669,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.006213343225226893,
      "loss": 0.8378,
      "num_input_tokens_seen": 56863144,
      "step": 97930
    },
    {
      "epoch": 14.58668453976765,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.006211763160022105,
      "loss": 0.7926,
      "num_input_tokens_seen": 56866152,
      "step": 97935
    },
    {
      "epoch": 14.58742925230861,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.006210183243285645,
      "loss": 0.7934,
      "num_input_tokens_seen": 56869064,
      "step": 97940
    },
    {
      "epoch": 14.588173964849569,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.006208603475044207,
      "loss": 0.7856,
      "num_input_tokens_seen": 56871656,
      "step": 97945
    },
    {
      "epoch": 14.588918677390527,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.006207023855324472,
      "loss": 0.818,
      "num_input_tokens_seen": 56874536,
      "step": 97950
    },
    {
      "epoch": 14.589663389931486,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0062054443841531364,
      "loss": 0.7824,
      "num_input_tokens_seen": 56877416,
      "step": 97955
    },
    {
      "epoch": 14.590408102472445,
      "grad_norm": 0.048828125,
      "learning_rate": 0.006203865061556875,
      "loss": 0.7848,
      "num_input_tokens_seen": 56880136,
      "step": 97960
    },
    {
      "epoch": 14.591152815013405,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.00620228588756238,
      "loss": 0.7897,
      "num_input_tokens_seen": 56883176,
      "step": 97965
    },
    {
      "epoch": 14.591897527554364,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.006200706862196317,
      "loss": 0.8333,
      "num_input_tokens_seen": 56886280,
      "step": 97970
    },
    {
      "epoch": 14.592642240095323,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.006199127985485373,
      "loss": 0.8059,
      "num_input_tokens_seen": 56888744,
      "step": 97975
    },
    {
      "epoch": 14.593386952636282,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.006197549257456217,
      "loss": 0.7851,
      "num_input_tokens_seen": 56891848,
      "step": 97980
    },
    {
      "epoch": 14.594131665177242,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.006195970678135515,
      "loss": 0.805,
      "num_input_tokens_seen": 56894632,
      "step": 97985
    },
    {
      "epoch": 14.594876377718201,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.006194392247549946,
      "loss": 0.798,
      "num_input_tokens_seen": 56897608,
      "step": 97990
    },
    {
      "epoch": 14.59562109025916,
      "grad_norm": 0.099609375,
      "learning_rate": 0.006192813965726165,
      "loss": 0.8115,
      "num_input_tokens_seen": 56900552,
      "step": 97995
    },
    {
      "epoch": 14.596365802800118,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.006191235832690847,
      "loss": 0.7959,
      "num_input_tokens_seen": 56903400,
      "step": 98000
    },
    {
      "epoch": 14.597110515341079,
      "grad_norm": 0.0703125,
      "learning_rate": 0.006189657848470646,
      "loss": 0.7867,
      "num_input_tokens_seen": 56906536,
      "step": 98005
    },
    {
      "epoch": 14.597855227882038,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0061880800130922195,
      "loss": 0.7931,
      "num_input_tokens_seen": 56909352,
      "step": 98010
    },
    {
      "epoch": 14.598599940422996,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.006186502326582221,
      "loss": 0.7946,
      "num_input_tokens_seen": 56912584,
      "step": 98015
    },
    {
      "epoch": 14.599344652963955,
      "grad_norm": 0.03515625,
      "learning_rate": 0.006184924788967315,
      "loss": 0.8031,
      "num_input_tokens_seen": 56915432,
      "step": 98020
    },
    {
      "epoch": 14.600089365504916,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.006183347400274144,
      "loss": 0.7829,
      "num_input_tokens_seen": 56918280,
      "step": 98025
    },
    {
      "epoch": 14.600834078045875,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0061817701605293525,
      "loss": 0.7955,
      "num_input_tokens_seen": 56921128,
      "step": 98030
    },
    {
      "epoch": 14.601578790586833,
      "grad_norm": 0.033203125,
      "learning_rate": 0.006180193069759597,
      "loss": 0.8437,
      "num_input_tokens_seen": 56923816,
      "step": 98035
    },
    {
      "epoch": 14.602323503127792,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00617861612799151,
      "loss": 0.8122,
      "num_input_tokens_seen": 56927176,
      "step": 98040
    },
    {
      "epoch": 14.603068215668753,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.006177039335251743,
      "loss": 0.8005,
      "num_input_tokens_seen": 56929992,
      "step": 98045
    },
    {
      "epoch": 14.603812928209711,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.006175462691566925,
      "loss": 0.7985,
      "num_input_tokens_seen": 56932840,
      "step": 98050
    },
    {
      "epoch": 14.60455764075067,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.006173886196963701,
      "loss": 0.8083,
      "num_input_tokens_seen": 56935848,
      "step": 98055
    },
    {
      "epoch": 14.605302353291629,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.006172309851468699,
      "loss": 0.8131,
      "num_input_tokens_seen": 56938664,
      "step": 98060
    },
    {
      "epoch": 14.60604706583259,
      "grad_norm": 0.060546875,
      "learning_rate": 0.006170733655108544,
      "loss": 0.7753,
      "num_input_tokens_seen": 56941384,
      "step": 98065
    },
    {
      "epoch": 14.606791778373548,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.006169157607909876,
      "loss": 0.7934,
      "num_input_tokens_seen": 56944072,
      "step": 98070
    },
    {
      "epoch": 14.607536490914507,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.006167581709899315,
      "loss": 0.7961,
      "num_input_tokens_seen": 56947080,
      "step": 98075
    },
    {
      "epoch": 14.608281203455466,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.006166005961103485,
      "loss": 0.7854,
      "num_input_tokens_seen": 56949992,
      "step": 98080
    },
    {
      "epoch": 14.609025915996426,
      "grad_norm": 0.021484375,
      "learning_rate": 0.006164430361548999,
      "loss": 0.8048,
      "num_input_tokens_seen": 56952488,
      "step": 98085
    },
    {
      "epoch": 14.609770628537385,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.006162854911262488,
      "loss": 0.7926,
      "num_input_tokens_seen": 56955272,
      "step": 98090
    },
    {
      "epoch": 14.610515341078344,
      "grad_norm": 0.0703125,
      "learning_rate": 0.006161279610270556,
      "loss": 0.7752,
      "num_input_tokens_seen": 56957928,
      "step": 98095
    },
    {
      "epoch": 14.611260053619302,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0061597044585998285,
      "loss": 0.7891,
      "num_input_tokens_seen": 56960904,
      "step": 98100
    },
    {
      "epoch": 14.612004766160261,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.006158129456276908,
      "loss": 0.7951,
      "num_input_tokens_seen": 56963880,
      "step": 98105
    },
    {
      "epoch": 14.612749478701222,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.006156554603328398,
      "loss": 0.8182,
      "num_input_tokens_seen": 56966888,
      "step": 98110
    },
    {
      "epoch": 14.61349419124218,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.006154979899780917,
      "loss": 0.8008,
      "num_input_tokens_seen": 56969544,
      "step": 98115
    },
    {
      "epoch": 14.61423890378314,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.006153405345661054,
      "loss": 0.7929,
      "num_input_tokens_seen": 56972456,
      "step": 98120
    },
    {
      "epoch": 14.6149836163241,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.006151830940995423,
      "loss": 0.806,
      "num_input_tokens_seen": 56975400,
      "step": 98125
    },
    {
      "epoch": 14.615728328865059,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.006150256685810609,
      "loss": 0.8017,
      "num_input_tokens_seen": 56978440,
      "step": 98130
    },
    {
      "epoch": 14.616473041406017,
      "grad_norm": 0.046875,
      "learning_rate": 0.006148682580133219,
      "loss": 0.7915,
      "num_input_tokens_seen": 56981448,
      "step": 98135
    },
    {
      "epoch": 14.617217753946976,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.00614710862398984,
      "loss": 0.7802,
      "num_input_tokens_seen": 56984360,
      "step": 98140
    },
    {
      "epoch": 14.617962466487935,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.006145534817407065,
      "loss": 0.8065,
      "num_input_tokens_seen": 56987208,
      "step": 98145
    },
    {
      "epoch": 14.618707179028895,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.006143961160411479,
      "loss": 0.8142,
      "num_input_tokens_seen": 56990056,
      "step": 98150
    },
    {
      "epoch": 14.619451891569854,
      "grad_norm": 0.05078125,
      "learning_rate": 0.006142387653029664,
      "loss": 0.8036,
      "num_input_tokens_seen": 56992968,
      "step": 98155
    },
    {
      "epoch": 14.620196604110813,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.006140814295288211,
      "loss": 0.7863,
      "num_input_tokens_seen": 56996168,
      "step": 98160
    },
    {
      "epoch": 14.620941316651772,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.006139241087213692,
      "loss": 0.789,
      "num_input_tokens_seen": 56998856,
      "step": 98165
    },
    {
      "epoch": 14.621686029192732,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.006137668028832695,
      "loss": 0.7955,
      "num_input_tokens_seen": 57001608,
      "step": 98170
    },
    {
      "epoch": 14.622430741733691,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0061360951201717825,
      "loss": 0.7936,
      "num_input_tokens_seen": 57004552,
      "step": 98175
    },
    {
      "epoch": 14.62317545427465,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.006134522361257541,
      "loss": 0.7706,
      "num_input_tokens_seen": 57007496,
      "step": 98180
    },
    {
      "epoch": 14.623920166815608,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.006132949752116533,
      "loss": 0.8562,
      "num_input_tokens_seen": 57010376,
      "step": 98185
    },
    {
      "epoch": 14.624664879356569,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.006131377292775321,
      "loss": 0.7904,
      "num_input_tokens_seen": 57013256,
      "step": 98190
    },
    {
      "epoch": 14.625409591897528,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.006129804983260482,
      "loss": 0.8089,
      "num_input_tokens_seen": 57016104,
      "step": 98195
    },
    {
      "epoch": 14.626154304438487,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.006128232823598566,
      "loss": 0.7852,
      "num_input_tokens_seen": 57018792,
      "step": 98200
    },
    {
      "epoch": 14.626899016979445,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.006126660813816146,
      "loss": 0.7926,
      "num_input_tokens_seen": 57021864,
      "step": 98205
    },
    {
      "epoch": 14.627643729520406,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.006125088953939773,
      "loss": 0.8013,
      "num_input_tokens_seen": 57024488,
      "step": 98210
    },
    {
      "epoch": 14.628388442061365,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.006123517243996002,
      "loss": 0.8195,
      "num_input_tokens_seen": 57027272,
      "step": 98215
    },
    {
      "epoch": 14.629133154602323,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.006121945684011387,
      "loss": 0.8014,
      "num_input_tokens_seen": 57030248,
      "step": 98220
    },
    {
      "epoch": 14.629877867143282,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0061203742740124694,
      "loss": 0.7855,
      "num_input_tokens_seen": 57033160,
      "step": 98225
    },
    {
      "epoch": 14.630622579684243,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.006118803014025808,
      "loss": 0.7983,
      "num_input_tokens_seen": 57035880,
      "step": 98230
    },
    {
      "epoch": 14.631367292225201,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00611723190407794,
      "loss": 0.7801,
      "num_input_tokens_seen": 57038632,
      "step": 98235
    },
    {
      "epoch": 14.63211200476616,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.006115660944195415,
      "loss": 0.797,
      "num_input_tokens_seen": 57041384,
      "step": 98240
    },
    {
      "epoch": 14.632856717307119,
      "grad_norm": 0.03515625,
      "learning_rate": 0.006114090134404764,
      "loss": 0.8034,
      "num_input_tokens_seen": 57044360,
      "step": 98245
    },
    {
      "epoch": 14.63360142984808,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.006112519474732534,
      "loss": 0.809,
      "num_input_tokens_seen": 57047304,
      "step": 98250
    },
    {
      "epoch": 14.634346142389038,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00611094896520525,
      "loss": 0.8027,
      "num_input_tokens_seen": 57049896,
      "step": 98255
    },
    {
      "epoch": 14.635090854929997,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.006109378605849453,
      "loss": 0.7916,
      "num_input_tokens_seen": 57052648,
      "step": 98260
    },
    {
      "epoch": 14.635835567470956,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.006107808396691669,
      "loss": 0.797,
      "num_input_tokens_seen": 57055432,
      "step": 98265
    },
    {
      "epoch": 14.636580280011916,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0061062383377584245,
      "loss": 0.7972,
      "num_input_tokens_seen": 57058216,
      "step": 98270
    },
    {
      "epoch": 14.637324992552875,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.006104668429076243,
      "loss": 0.7914,
      "num_input_tokens_seen": 57061384,
      "step": 98275
    },
    {
      "epoch": 14.638069705093834,
      "grad_norm": 0.04296875,
      "learning_rate": 0.006103098670671644,
      "loss": 0.7864,
      "num_input_tokens_seen": 57064296,
      "step": 98280
    },
    {
      "epoch": 14.638814417634793,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.006101529062571155,
      "loss": 0.7867,
      "num_input_tokens_seen": 57066920,
      "step": 98285
    },
    {
      "epoch": 14.639559130175751,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.006099959604801283,
      "loss": 0.7968,
      "num_input_tokens_seen": 57069896,
      "step": 98290
    },
    {
      "epoch": 14.640303842716712,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0060983902973885526,
      "loss": 0.7846,
      "num_input_tokens_seen": 57072808,
      "step": 98295
    },
    {
      "epoch": 14.64104855525767,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.006096821140359473,
      "loss": 0.7891,
      "num_input_tokens_seen": 57075944,
      "step": 98300
    },
    {
      "epoch": 14.64179326779863,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.006095252133740544,
      "loss": 0.8046,
      "num_input_tokens_seen": 57078952,
      "step": 98305
    },
    {
      "epoch": 14.642537980339588,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.006093683277558286,
      "loss": 0.7962,
      "num_input_tokens_seen": 57082216,
      "step": 98310
    },
    {
      "epoch": 14.643282692880549,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.006092114571839192,
      "loss": 0.7807,
      "num_input_tokens_seen": 57084936,
      "step": 98315
    },
    {
      "epoch": 14.644027405421507,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.006090546016609775,
      "loss": 0.8021,
      "num_input_tokens_seen": 57087848,
      "step": 98320
    },
    {
      "epoch": 14.644772117962466,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.006088977611896521,
      "loss": 0.8046,
      "num_input_tokens_seen": 57090600,
      "step": 98325
    },
    {
      "epoch": 14.645516830503425,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.006087409357725939,
      "loss": 0.797,
      "num_input_tokens_seen": 57093384,
      "step": 98330
    },
    {
      "epoch": 14.646261543044385,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.006085841254124517,
      "loss": 0.803,
      "num_input_tokens_seen": 57096488,
      "step": 98335
    },
    {
      "epoch": 14.647006255585344,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.006084273301118748,
      "loss": 0.7967,
      "num_input_tokens_seen": 57099432,
      "step": 98340
    },
    {
      "epoch": 14.647750968126303,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.006082705498735122,
      "loss": 0.7987,
      "num_input_tokens_seen": 57102376,
      "step": 98345
    },
    {
      "epoch": 14.648495680667262,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.0060811378470001145,
      "loss": 0.7782,
      "num_input_tokens_seen": 57105544,
      "step": 98350
    },
    {
      "epoch": 14.649240393208222,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0060795703459402255,
      "loss": 0.7953,
      "num_input_tokens_seen": 57108296,
      "step": 98355
    },
    {
      "epoch": 14.649985105749181,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.006078002995581924,
      "loss": 0.7936,
      "num_input_tokens_seen": 57111400,
      "step": 98360
    },
    {
      "epoch": 14.65072981829014,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0060764357959516985,
      "loss": 0.8053,
      "num_input_tokens_seen": 57113992,
      "step": 98365
    },
    {
      "epoch": 14.651474530831099,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.006074868747076015,
      "loss": 0.809,
      "num_input_tokens_seen": 57116456,
      "step": 98370
    },
    {
      "epoch": 14.652219243372059,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.006073301848981359,
      "loss": 0.8176,
      "num_input_tokens_seen": 57119560,
      "step": 98375
    },
    {
      "epoch": 14.652963955913018,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0060717351016941895,
      "loss": 0.7971,
      "num_input_tokens_seen": 57122760,
      "step": 98380
    },
    {
      "epoch": 14.653708668453977,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.006070168505240988,
      "loss": 0.7883,
      "num_input_tokens_seen": 57125416,
      "step": 98385
    },
    {
      "epoch": 14.654453380994935,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.006068602059648214,
      "loss": 0.7922,
      "num_input_tokens_seen": 57128392,
      "step": 98390
    },
    {
      "epoch": 14.655198093535896,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.006067035764942324,
      "loss": 0.7951,
      "num_input_tokens_seen": 57131496,
      "step": 98395
    },
    {
      "epoch": 14.655942806076855,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.006065469621149791,
      "loss": 0.7959,
      "num_input_tokens_seen": 57134312,
      "step": 98400
    },
    {
      "epoch": 14.656687518617813,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.006063903628297069,
      "loss": 0.7811,
      "num_input_tokens_seen": 57137288,
      "step": 98405
    },
    {
      "epoch": 14.657432231158772,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.006062337786410612,
      "loss": 0.847,
      "num_input_tokens_seen": 57140136,
      "step": 98410
    },
    {
      "epoch": 14.658176943699733,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0060607720955168694,
      "loss": 0.8211,
      "num_input_tokens_seen": 57142984,
      "step": 98415
    },
    {
      "epoch": 14.658921656240691,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.006059206555642302,
      "loss": 0.7828,
      "num_input_tokens_seen": 57145896,
      "step": 98420
    },
    {
      "epoch": 14.65966636878165,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.006057641166813353,
      "loss": 0.7988,
      "num_input_tokens_seen": 57149000,
      "step": 98425
    },
    {
      "epoch": 14.660411081322609,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.006056075929056463,
      "loss": 0.8014,
      "num_input_tokens_seen": 57151784,
      "step": 98430
    },
    {
      "epoch": 14.66115579386357,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.006054510842398085,
      "loss": 0.7997,
      "num_input_tokens_seen": 57154856,
      "step": 98435
    },
    {
      "epoch": 14.661900506404528,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.006052945906864648,
      "loss": 0.8013,
      "num_input_tokens_seen": 57157928,
      "step": 98440
    },
    {
      "epoch": 14.662645218945487,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.006051381122482603,
      "loss": 0.7955,
      "num_input_tokens_seen": 57160776,
      "step": 98445
    },
    {
      "epoch": 14.663389931486446,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.006049816489278374,
      "loss": 0.7927,
      "num_input_tokens_seen": 57163720,
      "step": 98450
    },
    {
      "epoch": 14.664134644027406,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.006048252007278403,
      "loss": 0.8014,
      "num_input_tokens_seen": 57166952,
      "step": 98455
    },
    {
      "epoch": 14.664879356568365,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.006046687676509112,
      "loss": 0.7976,
      "num_input_tokens_seen": 57169704,
      "step": 98460
    },
    {
      "epoch": 14.665624069109324,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.006045123496996939,
      "loss": 0.7791,
      "num_input_tokens_seen": 57172360,
      "step": 98465
    },
    {
      "epoch": 14.666368781650283,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0060435594687683015,
      "loss": 0.8046,
      "num_input_tokens_seen": 57175368,
      "step": 98470
    },
    {
      "epoch": 14.667113494191241,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0060419955918496245,
      "loss": 0.8008,
      "num_input_tokens_seen": 57178504,
      "step": 98475
    },
    {
      "epoch": 14.667858206732202,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.006040431866267327,
      "loss": 0.777,
      "num_input_tokens_seen": 57181448,
      "step": 98480
    },
    {
      "epoch": 14.66860291927316,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.006038868292047821,
      "loss": 0.8091,
      "num_input_tokens_seen": 57184264,
      "step": 98485
    },
    {
      "epoch": 14.66934763181412,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.006037304869217536,
      "loss": 0.7966,
      "num_input_tokens_seen": 57187208,
      "step": 98490
    },
    {
      "epoch": 14.670092344355078,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.006035741597802868,
      "loss": 0.7797,
      "num_input_tokens_seen": 57189896,
      "step": 98495
    },
    {
      "epoch": 14.670837056896039,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.00603417847783024,
      "loss": 0.808,
      "num_input_tokens_seen": 57192936,
      "step": 98500
    },
    {
      "epoch": 14.671581769436997,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.006032615509326056,
      "loss": 0.7808,
      "num_input_tokens_seen": 57195656,
      "step": 98505
    },
    {
      "epoch": 14.672326481977956,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.006031052692316711,
      "loss": 0.7942,
      "num_input_tokens_seen": 57198696,
      "step": 98510
    },
    {
      "epoch": 14.673071194518915,
      "grad_norm": 0.04296875,
      "learning_rate": 0.006029490026828623,
      "loss": 0.7992,
      "num_input_tokens_seen": 57201576,
      "step": 98515
    },
    {
      "epoch": 14.673815907059875,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.006027927512888177,
      "loss": 0.7879,
      "num_input_tokens_seen": 57204680,
      "step": 98520
    },
    {
      "epoch": 14.674560619600834,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.006026365150521781,
      "loss": 0.8093,
      "num_input_tokens_seen": 57207656,
      "step": 98525
    },
    {
      "epoch": 14.675305332141793,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.006024802939755823,
      "loss": 0.8016,
      "num_input_tokens_seen": 57210600,
      "step": 98530
    },
    {
      "epoch": 14.676050044682752,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.006023240880616701,
      "loss": 0.7913,
      "num_input_tokens_seen": 57213448,
      "step": 98535
    },
    {
      "epoch": 14.676794757223712,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0060216789731308016,
      "loss": 0.7878,
      "num_input_tokens_seen": 57216360,
      "step": 98540
    },
    {
      "epoch": 14.677539469764671,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.006020117217324509,
      "loss": 0.8041,
      "num_input_tokens_seen": 57219368,
      "step": 98545
    },
    {
      "epoch": 14.67828418230563,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.006018555613224212,
      "loss": 0.794,
      "num_input_tokens_seen": 57222184,
      "step": 98550
    },
    {
      "epoch": 14.679028894846589,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.006016994160856283,
      "loss": 0.7825,
      "num_input_tokens_seen": 57224936,
      "step": 98555
    },
    {
      "epoch": 14.679773607387549,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0060154328602471125,
      "loss": 0.7873,
      "num_input_tokens_seen": 57227784,
      "step": 98560
    },
    {
      "epoch": 14.680518319928508,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.006013871711423067,
      "loss": 0.777,
      "num_input_tokens_seen": 57230856,
      "step": 98565
    },
    {
      "epoch": 14.681263032469467,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0060123107144105315,
      "loss": 0.7998,
      "num_input_tokens_seen": 57234024,
      "step": 98570
    },
    {
      "epoch": 14.682007745010425,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.006010749869235866,
      "loss": 0.8051,
      "num_input_tokens_seen": 57238088,
      "step": 98575
    },
    {
      "epoch": 14.682752457551386,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.006009189175925452,
      "loss": 0.7946,
      "num_input_tokens_seen": 57240680,
      "step": 98580
    },
    {
      "epoch": 14.683497170092345,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.006007628634505649,
      "loss": 0.7831,
      "num_input_tokens_seen": 57243560,
      "step": 98585
    },
    {
      "epoch": 14.684241882633303,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.006006068245002813,
      "loss": 0.7956,
      "num_input_tokens_seen": 57246312,
      "step": 98590
    },
    {
      "epoch": 14.684986595174262,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.00600450800744332,
      "loss": 0.7832,
      "num_input_tokens_seen": 57249384,
      "step": 98595
    },
    {
      "epoch": 14.685731307715223,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.006002947921853521,
      "loss": 0.771,
      "num_input_tokens_seen": 57252520,
      "step": 98600
    },
    {
      "epoch": 14.686476020256181,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.006001387988259774,
      "loss": 0.7945,
      "num_input_tokens_seen": 57255464,
      "step": 98605
    },
    {
      "epoch": 14.68722073279714,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.005999828206688424,
      "loss": 0.7927,
      "num_input_tokens_seen": 57258184,
      "step": 98610
    },
    {
      "epoch": 14.687965445338099,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.005998268577165836,
      "loss": 0.7837,
      "num_input_tokens_seen": 57260808,
      "step": 98615
    },
    {
      "epoch": 14.688710157879058,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.005996709099718344,
      "loss": 0.7907,
      "num_input_tokens_seen": 57263656,
      "step": 98620
    },
    {
      "epoch": 14.689454870420018,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.005995149774372308,
      "loss": 0.8293,
      "num_input_tokens_seen": 57266696,
      "step": 98625
    },
    {
      "epoch": 14.690199582960977,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.005993590601154065,
      "loss": 0.7965,
      "num_input_tokens_seen": 57269384,
      "step": 98630
    },
    {
      "epoch": 14.690944295501936,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.005992031580089949,
      "loss": 0.8188,
      "num_input_tokens_seen": 57272136,
      "step": 98635
    },
    {
      "epoch": 14.691689008042896,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.00599047271120631,
      "loss": 0.8057,
      "num_input_tokens_seen": 57275048,
      "step": 98640
    },
    {
      "epoch": 14.692433720583855,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.005988913994529471,
      "loss": 0.8067,
      "num_input_tokens_seen": 57277960,
      "step": 98645
    },
    {
      "epoch": 14.693178433124814,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.005987355430085778,
      "loss": 0.8042,
      "num_input_tokens_seen": 57281064,
      "step": 98650
    },
    {
      "epoch": 14.693923145665773,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.00598579701790155,
      "loss": 0.8016,
      "num_input_tokens_seen": 57283944,
      "step": 98655
    },
    {
      "epoch": 14.694667858206731,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.005984238758003125,
      "loss": 0.7899,
      "num_input_tokens_seen": 57286728,
      "step": 98660
    },
    {
      "epoch": 14.695412570747692,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.005982680650416823,
      "loss": 0.7979,
      "num_input_tokens_seen": 57289992,
      "step": 98665
    },
    {
      "epoch": 14.69615728328865,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.005981122695168965,
      "loss": 0.8105,
      "num_input_tokens_seen": 57293064,
      "step": 98670
    },
    {
      "epoch": 14.69690199582961,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.005979564892285874,
      "loss": 0.8075,
      "num_input_tokens_seen": 57295848,
      "step": 98675
    },
    {
      "epoch": 14.697646708370568,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.005978007241793861,
      "loss": 0.815,
      "num_input_tokens_seen": 57298696,
      "step": 98680
    },
    {
      "epoch": 14.698391420911529,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.005976449743719251,
      "loss": 0.8,
      "num_input_tokens_seen": 57301512,
      "step": 98685
    },
    {
      "epoch": 14.699136133452487,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.005974892398088347,
      "loss": 0.7995,
      "num_input_tokens_seen": 57304232,
      "step": 98690
    },
    {
      "epoch": 14.699880845993446,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.005973335204927468,
      "loss": 0.8081,
      "num_input_tokens_seen": 57307112,
      "step": 98695
    },
    {
      "epoch": 14.700625558534405,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.00597177816426291,
      "loss": 0.817,
      "num_input_tokens_seen": 57309832,
      "step": 98700
    },
    {
      "epoch": 14.701370271075366,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.005970221276120991,
      "loss": 0.7992,
      "num_input_tokens_seen": 57312584,
      "step": 98705
    },
    {
      "epoch": 14.702114983616324,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0059686645405280046,
      "loss": 0.7921,
      "num_input_tokens_seen": 57315592,
      "step": 98710
    },
    {
      "epoch": 14.702859696157283,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.005967107957510248,
      "loss": 0.7973,
      "num_input_tokens_seen": 57318376,
      "step": 98715
    },
    {
      "epoch": 14.703604408698242,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.005965551527094027,
      "loss": 0.7895,
      "num_input_tokens_seen": 57321576,
      "step": 98720
    },
    {
      "epoch": 14.704349121239202,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0059639952493056245,
      "loss": 0.7778,
      "num_input_tokens_seen": 57324552,
      "step": 98725
    },
    {
      "epoch": 14.705093833780161,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.005962439124171345,
      "loss": 0.7771,
      "num_input_tokens_seen": 57327624,
      "step": 98730
    },
    {
      "epoch": 14.70583854632112,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0059608831517174695,
      "loss": 0.8014,
      "num_input_tokens_seen": 57330312,
      "step": 98735
    },
    {
      "epoch": 14.706583258862079,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.005959327331970287,
      "loss": 0.8039,
      "num_input_tokens_seen": 57333192,
      "step": 98740
    },
    {
      "epoch": 14.70732797140304,
      "grad_norm": 0.064453125,
      "learning_rate": 0.00595777166495608,
      "loss": 0.7933,
      "num_input_tokens_seen": 57336328,
      "step": 98745
    },
    {
      "epoch": 14.708072683943998,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.005956216150701126,
      "loss": 0.8115,
      "num_input_tokens_seen": 57339272,
      "step": 98750
    },
    {
      "epoch": 14.708817396484957,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.005954660789231714,
      "loss": 0.8014,
      "num_input_tokens_seen": 57342120,
      "step": 98755
    },
    {
      "epoch": 14.709562109025915,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.005953105580574109,
      "loss": 0.8061,
      "num_input_tokens_seen": 57345224,
      "step": 98760
    },
    {
      "epoch": 14.710306821566876,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.005951550524754596,
      "loss": 0.7972,
      "num_input_tokens_seen": 57348040,
      "step": 98765
    },
    {
      "epoch": 14.711051534107835,
      "grad_norm": 0.076171875,
      "learning_rate": 0.005949995621799435,
      "loss": 0.7719,
      "num_input_tokens_seen": 57350888,
      "step": 98770
    },
    {
      "epoch": 14.711796246648793,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.005948440871734905,
      "loss": 0.8272,
      "num_input_tokens_seen": 57353896,
      "step": 98775
    },
    {
      "epoch": 14.712540959189752,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.005946886274587262,
      "loss": 0.8097,
      "num_input_tokens_seen": 57356680,
      "step": 98780
    },
    {
      "epoch": 14.713285671730713,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.00594533183038278,
      "loss": 0.8069,
      "num_input_tokens_seen": 57359592,
      "step": 98785
    },
    {
      "epoch": 14.714030384271672,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.005943777539147712,
      "loss": 0.8124,
      "num_input_tokens_seen": 57362280,
      "step": 98790
    },
    {
      "epoch": 14.71477509681263,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.005942223400908314,
      "loss": 0.7956,
      "num_input_tokens_seen": 57365032,
      "step": 98795
    },
    {
      "epoch": 14.715519809353589,
      "grad_norm": 0.060546875,
      "learning_rate": 0.005940669415690851,
      "loss": 0.8071,
      "num_input_tokens_seen": 57367752,
      "step": 98800
    },
    {
      "epoch": 14.716264521894548,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0059391155835215705,
      "loss": 0.8223,
      "num_input_tokens_seen": 57370568,
      "step": 98805
    },
    {
      "epoch": 14.717009234435508,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.005937561904426724,
      "loss": 0.7896,
      "num_input_tokens_seen": 57373320,
      "step": 98810
    },
    {
      "epoch": 14.717753946976467,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.005936008378432553,
      "loss": 0.797,
      "num_input_tokens_seen": 57376264,
      "step": 98815
    },
    {
      "epoch": 14.718498659517426,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.005934455005565313,
      "loss": 0.797,
      "num_input_tokens_seen": 57379272,
      "step": 98820
    },
    {
      "epoch": 14.719243372058386,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.005932901785851242,
      "loss": 0.7905,
      "num_input_tokens_seen": 57381832,
      "step": 98825
    },
    {
      "epoch": 14.719988084599345,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.005931348719316574,
      "loss": 0.7881,
      "num_input_tokens_seen": 57384936,
      "step": 98830
    },
    {
      "epoch": 14.720732797140304,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.0059297958059875595,
      "loss": 0.7749,
      "num_input_tokens_seen": 57387496,
      "step": 98835
    },
    {
      "epoch": 14.721477509681263,
      "grad_norm": 0.0216064453125,
      "learning_rate": 0.00592824304589042,
      "loss": 0.813,
      "num_input_tokens_seen": 57390728,
      "step": 98840
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 0.033203125,
      "learning_rate": 0.005926690439051399,
      "loss": 0.7966,
      "num_input_tokens_seen": 57393512,
      "step": 98845
    },
    {
      "epoch": 14.722966934763182,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.005925137985496715,
      "loss": 0.8027,
      "num_input_tokens_seen": 57396584,
      "step": 98850
    },
    {
      "epoch": 14.72371164730414,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0059235856852526084,
      "loss": 0.8015,
      "num_input_tokens_seen": 57399464,
      "step": 98855
    },
    {
      "epoch": 14.7244563598451,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.005922033538345291,
      "loss": 0.8063,
      "num_input_tokens_seen": 57402888,
      "step": 98860
    },
    {
      "epoch": 14.725201072386058,
      "grad_norm": 0.0546875,
      "learning_rate": 0.005920481544800997,
      "loss": 0.7689,
      "num_input_tokens_seen": 57405576,
      "step": 98865
    },
    {
      "epoch": 14.725945784927019,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.005918929704645938,
      "loss": 0.7854,
      "num_input_tokens_seen": 57408424,
      "step": 98870
    },
    {
      "epoch": 14.726690497467978,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.005917378017906334,
      "loss": 0.8193,
      "num_input_tokens_seen": 57411528,
      "step": 98875
    },
    {
      "epoch": 14.727435210008936,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.005915826484608394,
      "loss": 0.8096,
      "num_input_tokens_seen": 57414344,
      "step": 98880
    },
    {
      "epoch": 14.728179922549895,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.005914275104778329,
      "loss": 0.8005,
      "num_input_tokens_seen": 57417320,
      "step": 98885
    },
    {
      "epoch": 14.728924635090856,
      "grad_norm": 0.037109375,
      "learning_rate": 0.005912723878442358,
      "loss": 0.7978,
      "num_input_tokens_seen": 57419816,
      "step": 98890
    },
    {
      "epoch": 14.729669347631814,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.005911172805626675,
      "loss": 0.803,
      "num_input_tokens_seen": 57422888,
      "step": 98895
    },
    {
      "epoch": 14.730414060172773,
      "grad_norm": 0.0625,
      "learning_rate": 0.0059096218863574955,
      "loss": 0.7894,
      "num_input_tokens_seen": 57425896,
      "step": 98900
    },
    {
      "epoch": 14.731158772713732,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.005908071120661014,
      "loss": 0.7894,
      "num_input_tokens_seen": 57428776,
      "step": 98905
    },
    {
      "epoch": 14.731903485254692,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.005906520508563426,
      "loss": 0.7967,
      "num_input_tokens_seen": 57431624,
      "step": 98910
    },
    {
      "epoch": 14.732648197795651,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.005904970050090938,
      "loss": 0.7982,
      "num_input_tokens_seen": 57434440,
      "step": 98915
    },
    {
      "epoch": 14.73339291033661,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.005903419745269731,
      "loss": 0.7868,
      "num_input_tokens_seen": 57437352,
      "step": 98920
    },
    {
      "epoch": 14.734137622877569,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.005901869594126007,
      "loss": 0.7924,
      "num_input_tokens_seen": 57440296,
      "step": 98925
    },
    {
      "epoch": 14.73488233541853,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.00590031959668595,
      "loss": 0.7799,
      "num_input_tokens_seen": 57442984,
      "step": 98930
    },
    {
      "epoch": 14.735627047959488,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.005898769752975745,
      "loss": 0.8001,
      "num_input_tokens_seen": 57446088,
      "step": 98935
    },
    {
      "epoch": 14.736371760500447,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.005897220063021569,
      "loss": 0.8127,
      "num_input_tokens_seen": 57449064,
      "step": 98940
    },
    {
      "epoch": 14.737116473041405,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.005895670526849615,
      "loss": 0.79,
      "num_input_tokens_seen": 57451912,
      "step": 98945
    },
    {
      "epoch": 14.737861185582366,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.005894121144486053,
      "loss": 0.82,
      "num_input_tokens_seen": 57455048,
      "step": 98950
    },
    {
      "epoch": 14.738605898123325,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.005892571915957053,
      "loss": 0.7983,
      "num_input_tokens_seen": 57458248,
      "step": 98955
    },
    {
      "epoch": 14.739350610664284,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.005891022841288801,
      "loss": 0.7939,
      "num_input_tokens_seen": 57461256,
      "step": 98960
    },
    {
      "epoch": 14.740095323205242,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.005889473920507455,
      "loss": 0.7936,
      "num_input_tokens_seen": 57463784,
      "step": 98965
    },
    {
      "epoch": 14.740840035746203,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.005887925153639192,
      "loss": 0.8012,
      "num_input_tokens_seen": 57466504,
      "step": 98970
    },
    {
      "epoch": 14.741584748287162,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.005886376540710168,
      "loss": 0.7932,
      "num_input_tokens_seen": 57469288,
      "step": 98975
    },
    {
      "epoch": 14.74232946082812,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.005884828081746555,
      "loss": 0.7871,
      "num_input_tokens_seen": 57472264,
      "step": 98980
    },
    {
      "epoch": 14.743074173369079,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0058832797767745006,
      "loss": 0.7962,
      "num_input_tokens_seen": 57475144,
      "step": 98985
    },
    {
      "epoch": 14.743818885910038,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0058817316258201755,
      "loss": 0.798,
      "num_input_tokens_seen": 57478056,
      "step": 98990
    },
    {
      "epoch": 14.744563598450998,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.005880183628909726,
      "loss": 0.8023,
      "num_input_tokens_seen": 57480904,
      "step": 98995
    },
    {
      "epoch": 14.745308310991957,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.005878635786069305,
      "loss": 0.8059,
      "num_input_tokens_seen": 57483880,
      "step": 99000
    },
    {
      "epoch": 14.746053023532916,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.005877088097325064,
      "loss": 0.7944,
      "num_input_tokens_seen": 57487048,
      "step": 99005
    },
    {
      "epoch": 14.746797736073875,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0058755405627031404,
      "loss": 0.8062,
      "num_input_tokens_seen": 57489800,
      "step": 99010
    },
    {
      "epoch": 14.747542448614835,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.005873993182229691,
      "loss": 0.7757,
      "num_input_tokens_seen": 57492744,
      "step": 99015
    },
    {
      "epoch": 14.748287161155794,
      "grad_norm": 0.033203125,
      "learning_rate": 0.005872445955930846,
      "loss": 0.7841,
      "num_input_tokens_seen": 57495464,
      "step": 99020
    },
    {
      "epoch": 14.749031873696753,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.005870898883832755,
      "loss": 0.7916,
      "num_input_tokens_seen": 57498600,
      "step": 99025
    },
    {
      "epoch": 14.749776586237711,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.005869351965961549,
      "loss": 0.8002,
      "num_input_tokens_seen": 57501288,
      "step": 99030
    },
    {
      "epoch": 14.750521298778672,
      "grad_norm": 0.0390625,
      "learning_rate": 0.005867805202343356,
      "loss": 0.8016,
      "num_input_tokens_seen": 57504136,
      "step": 99035
    },
    {
      "epoch": 14.75126601131963,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.005866258593004318,
      "loss": 0.7871,
      "num_input_tokens_seen": 57507016,
      "step": 99040
    },
    {
      "epoch": 14.75201072386059,
      "grad_norm": 0.046875,
      "learning_rate": 0.005864712137970554,
      "loss": 0.79,
      "num_input_tokens_seen": 57509992,
      "step": 99045
    },
    {
      "epoch": 14.752755436401548,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.005863165837268198,
      "loss": 0.8037,
      "num_input_tokens_seen": 57513064,
      "step": 99050
    },
    {
      "epoch": 14.753500148942509,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.005861619690923363,
      "loss": 0.8077,
      "num_input_tokens_seen": 57515816,
      "step": 99055
    },
    {
      "epoch": 14.754244861483468,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.005860073698962182,
      "loss": 0.7837,
      "num_input_tokens_seen": 57518696,
      "step": 99060
    },
    {
      "epoch": 14.754989574024426,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.005858527861410764,
      "loss": 0.792,
      "num_input_tokens_seen": 57521544,
      "step": 99065
    },
    {
      "epoch": 14.755734286565385,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.005856982178295229,
      "loss": 0.8096,
      "num_input_tokens_seen": 57524264,
      "step": 99070
    },
    {
      "epoch": 14.756478999106346,
      "grad_norm": 0.052734375,
      "learning_rate": 0.005855436649641687,
      "loss": 0.8008,
      "num_input_tokens_seen": 57527272,
      "step": 99075
    },
    {
      "epoch": 14.757223711647304,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.005853891275476244,
      "loss": 0.7973,
      "num_input_tokens_seen": 57529960,
      "step": 99080
    },
    {
      "epoch": 14.757968424188263,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.005852346055825017,
      "loss": 0.788,
      "num_input_tokens_seen": 57532616,
      "step": 99085
    },
    {
      "epoch": 14.758713136729222,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.005850800990714101,
      "loss": 0.8155,
      "num_input_tokens_seen": 57535368,
      "step": 99090
    },
    {
      "epoch": 14.759457849270182,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.005849256080169608,
      "loss": 0.7982,
      "num_input_tokens_seen": 57538216,
      "step": 99095
    },
    {
      "epoch": 14.760202561811141,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.005847711324217629,
      "loss": 0.7838,
      "num_input_tokens_seen": 57541192,
      "step": 99100
    },
    {
      "epoch": 14.7609472743521,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0058461667228842705,
      "loss": 0.8112,
      "num_input_tokens_seen": 57544040,
      "step": 99105
    },
    {
      "epoch": 14.761691986893059,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.005844622276195621,
      "loss": 0.7869,
      "num_input_tokens_seen": 57546888,
      "step": 99110
    },
    {
      "epoch": 14.76243669943402,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.005843077984177767,
      "loss": 0.7978,
      "num_input_tokens_seen": 57549640,
      "step": 99115
    },
    {
      "epoch": 14.763181411974978,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.00584153384685681,
      "loss": 0.794,
      "num_input_tokens_seen": 57552552,
      "step": 99120
    },
    {
      "epoch": 14.763926124515937,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0058399898642588245,
      "loss": 0.7952,
      "num_input_tokens_seen": 57555272,
      "step": 99125
    },
    {
      "epoch": 14.764670837056896,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.005838446036409906,
      "loss": 0.778,
      "num_input_tokens_seen": 57558152,
      "step": 99130
    },
    {
      "epoch": 14.765415549597854,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.005836902363336131,
      "loss": 0.792,
      "num_input_tokens_seen": 57561224,
      "step": 99135
    },
    {
      "epoch": 14.766160262138815,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.005835358845063577,
      "loss": 0.7905,
      "num_input_tokens_seen": 57564040,
      "step": 99140
    },
    {
      "epoch": 14.766904974679774,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.005833815481618316,
      "loss": 0.7931,
      "num_input_tokens_seen": 57566792,
      "step": 99145
    },
    {
      "epoch": 14.767649687220732,
      "grad_norm": 0.0390625,
      "learning_rate": 0.005832272273026432,
      "loss": 0.7713,
      "num_input_tokens_seen": 57569704,
      "step": 99150
    },
    {
      "epoch": 14.768394399761693,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.005830729219313989,
      "loss": 0.8109,
      "num_input_tokens_seen": 57572488,
      "step": 99155
    },
    {
      "epoch": 14.769139112302652,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.0058291863205070515,
      "loss": 0.7787,
      "num_input_tokens_seen": 57575560,
      "step": 99160
    },
    {
      "epoch": 14.76988382484361,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.005827643576631696,
      "loss": 0.8174,
      "num_input_tokens_seen": 57578472,
      "step": 99165
    },
    {
      "epoch": 14.77062853738457,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.005826100987713976,
      "loss": 0.806,
      "num_input_tokens_seen": 57581224,
      "step": 99170
    },
    {
      "epoch": 14.771373249925528,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.005824558553779959,
      "loss": 0.7826,
      "num_input_tokens_seen": 57584008,
      "step": 99175
    },
    {
      "epoch": 14.772117962466488,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.005823016274855696,
      "loss": 0.8096,
      "num_input_tokens_seen": 57586760,
      "step": 99180
    },
    {
      "epoch": 14.772862675007447,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.005821474150967251,
      "loss": 0.7928,
      "num_input_tokens_seen": 57589736,
      "step": 99185
    },
    {
      "epoch": 14.773607387548406,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00581993218214067,
      "loss": 0.7883,
      "num_input_tokens_seen": 57592744,
      "step": 99190
    },
    {
      "epoch": 14.774352100089365,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.005818390368402,
      "loss": 0.7995,
      "num_input_tokens_seen": 57595688,
      "step": 99195
    },
    {
      "epoch": 14.775096812630325,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.005816848709777299,
      "loss": 0.8047,
      "num_input_tokens_seen": 57598632,
      "step": 99200
    },
    {
      "epoch": 14.775841525171284,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.005815307206292605,
      "loss": 0.797,
      "num_input_tokens_seen": 57601672,
      "step": 99205
    },
    {
      "epoch": 14.776586237712243,
      "grad_norm": 0.046875,
      "learning_rate": 0.005813765857973961,
      "loss": 0.7759,
      "num_input_tokens_seen": 57604520,
      "step": 99210
    },
    {
      "epoch": 14.777330950253202,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0058122246648474015,
      "loss": 0.7917,
      "num_input_tokens_seen": 57607432,
      "step": 99215
    },
    {
      "epoch": 14.778075662794162,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0058106836269389735,
      "loss": 0.8079,
      "num_input_tokens_seen": 57610472,
      "step": 99220
    },
    {
      "epoch": 14.77882037533512,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0058091427442747,
      "loss": 0.781,
      "num_input_tokens_seen": 57613064,
      "step": 99225
    },
    {
      "epoch": 14.77956508787608,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.005807602016880625,
      "loss": 0.7885,
      "num_input_tokens_seen": 57615784,
      "step": 99230
    },
    {
      "epoch": 14.780309800417038,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.00580606144478277,
      "loss": 0.7861,
      "num_input_tokens_seen": 57618504,
      "step": 99235
    },
    {
      "epoch": 14.781054512957999,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.005804521028007159,
      "loss": 0.8003,
      "num_input_tokens_seen": 57621448,
      "step": 99240
    },
    {
      "epoch": 14.781799225498958,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.005802980766579824,
      "loss": 0.7946,
      "num_input_tokens_seen": 57624488,
      "step": 99245
    },
    {
      "epoch": 14.782543938039916,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.005801440660526775,
      "loss": 0.7847,
      "num_input_tokens_seen": 57627560,
      "step": 99250
    },
    {
      "epoch": 14.783288650580875,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0057999007098740426,
      "loss": 0.789,
      "num_input_tokens_seen": 57630984,
      "step": 99255
    },
    {
      "epoch": 14.784033363121836,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.005798360914647639,
      "loss": 0.8036,
      "num_input_tokens_seen": 57633736,
      "step": 99260
    },
    {
      "epoch": 14.784778075662794,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.005796821274873574,
      "loss": 0.7988,
      "num_input_tokens_seen": 57636424,
      "step": 99265
    },
    {
      "epoch": 14.785522788203753,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.005795281790577854,
      "loss": 0.8265,
      "num_input_tokens_seen": 57639272,
      "step": 99270
    },
    {
      "epoch": 14.786267500744712,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0057937424617865,
      "loss": 0.7845,
      "num_input_tokens_seen": 57642280,
      "step": 99275
    },
    {
      "epoch": 14.787012213285673,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.005792203288525508,
      "loss": 0.7623,
      "num_input_tokens_seen": 57645352,
      "step": 99280
    },
    {
      "epoch": 14.787756925826631,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.005790664270820879,
      "loss": 0.7923,
      "num_input_tokens_seen": 57648296,
      "step": 99285
    },
    {
      "epoch": 14.78850163836759,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.005789125408698624,
      "loss": 0.8038,
      "num_input_tokens_seen": 57651080,
      "step": 99290
    },
    {
      "epoch": 14.789246350908549,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.005787586702184727,
      "loss": 0.7945,
      "num_input_tokens_seen": 57654312,
      "step": 99295
    },
    {
      "epoch": 14.78999106344951,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.005786048151305195,
      "loss": 0.7798,
      "num_input_tokens_seen": 57657064,
      "step": 99300
    },
    {
      "epoch": 14.790735775990468,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00578450975608601,
      "loss": 0.795,
      "num_input_tokens_seen": 57660136,
      "step": 99305
    },
    {
      "epoch": 14.791480488531427,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.00578297151655317,
      "loss": 0.8077,
      "num_input_tokens_seen": 57662920,
      "step": 99310
    },
    {
      "epoch": 14.792225201072386,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0057814334327326595,
      "loss": 0.7893,
      "num_input_tokens_seen": 57665992,
      "step": 99315
    },
    {
      "epoch": 14.792969913613344,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.005779895504650458,
      "loss": 0.8202,
      "num_input_tokens_seen": 57669064,
      "step": 99320
    },
    {
      "epoch": 14.793714626154305,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.005778357732332554,
      "loss": 0.7958,
      "num_input_tokens_seen": 57671880,
      "step": 99325
    },
    {
      "epoch": 14.794459338695264,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.005776820115804925,
      "loss": 0.7951,
      "num_input_tokens_seen": 57674600,
      "step": 99330
    },
    {
      "epoch": 14.795204051236222,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.005775282655093546,
      "loss": 0.785,
      "num_input_tokens_seen": 57677512,
      "step": 99335
    },
    {
      "epoch": 14.795948763777183,
      "grad_norm": 0.048828125,
      "learning_rate": 0.005773745350224385,
      "loss": 0.7967,
      "num_input_tokens_seen": 57680264,
      "step": 99340
    },
    {
      "epoch": 14.796693476318142,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.005772208201223425,
      "loss": 0.8047,
      "num_input_tokens_seen": 57683240,
      "step": 99345
    },
    {
      "epoch": 14.7974381888591,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.00577067120811663,
      "loss": 0.7988,
      "num_input_tokens_seen": 57685832,
      "step": 99350
    },
    {
      "epoch": 14.79818290140006,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0057691343709299564,
      "loss": 0.8092,
      "num_input_tokens_seen": 57688712,
      "step": 99355
    },
    {
      "epoch": 14.798927613941018,
      "grad_norm": 0.05859375,
      "learning_rate": 0.005767597689689384,
      "loss": 0.8072,
      "num_input_tokens_seen": 57691656,
      "step": 99360
    },
    {
      "epoch": 14.799672326481979,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.005766061164420857,
      "loss": 0.7673,
      "num_input_tokens_seen": 57694440,
      "step": 99365
    },
    {
      "epoch": 14.800417039022937,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.005764524795150349,
      "loss": 0.7927,
      "num_input_tokens_seen": 57697448,
      "step": 99370
    },
    {
      "epoch": 14.801161751563896,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.005762988581903803,
      "loss": 0.7791,
      "num_input_tokens_seen": 57700168,
      "step": 99375
    },
    {
      "epoch": 14.801906464104855,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.005761452524707181,
      "loss": 0.7777,
      "num_input_tokens_seen": 57702984,
      "step": 99380
    },
    {
      "epoch": 14.802651176645815,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.005759916623586425,
      "loss": 0.792,
      "num_input_tokens_seen": 57705960,
      "step": 99385
    },
    {
      "epoch": 14.803395889186774,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.00575838087856749,
      "loss": 0.7927,
      "num_input_tokens_seen": 57708936,
      "step": 99390
    },
    {
      "epoch": 14.804140601727733,
      "grad_norm": 0.046875,
      "learning_rate": 0.0057568452896763185,
      "loss": 0.7969,
      "num_input_tokens_seen": 57711816,
      "step": 99395
    },
    {
      "epoch": 14.804885314268692,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.005755309856938851,
      "loss": 0.7912,
      "num_input_tokens_seen": 57714600,
      "step": 99400
    },
    {
      "epoch": 14.805630026809652,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.005753774580381028,
      "loss": 0.7718,
      "num_input_tokens_seen": 57717704,
      "step": 99405
    },
    {
      "epoch": 14.80637473935061,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00575223946002878,
      "loss": 0.7964,
      "num_input_tokens_seen": 57720232,
      "step": 99410
    },
    {
      "epoch": 14.80711945189157,
      "grad_norm": 0.05859375,
      "learning_rate": 0.005750704495908053,
      "loss": 0.7782,
      "num_input_tokens_seen": 57723432,
      "step": 99415
    },
    {
      "epoch": 14.807864164432528,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.005749169688044768,
      "loss": 0.803,
      "num_input_tokens_seen": 57726120,
      "step": 99420
    },
    {
      "epoch": 14.808608876973489,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.005747635036464865,
      "loss": 0.8174,
      "num_input_tokens_seen": 57728872,
      "step": 99425
    },
    {
      "epoch": 14.809353589514448,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.0057461005411942635,
      "loss": 0.8113,
      "num_input_tokens_seen": 57732072,
      "step": 99430
    },
    {
      "epoch": 14.810098302055406,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.005744566202258883,
      "loss": 0.7792,
      "num_input_tokens_seen": 57734888,
      "step": 99435
    },
    {
      "epoch": 14.810843014596365,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0057430320196846536,
      "loss": 0.7875,
      "num_input_tokens_seen": 57737896,
      "step": 99440
    },
    {
      "epoch": 14.811587727137326,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0057414979934974865,
      "loss": 0.7769,
      "num_input_tokens_seen": 57741224,
      "step": 99445
    },
    {
      "epoch": 14.812332439678285,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.005739964123723306,
      "loss": 0.7766,
      "num_input_tokens_seen": 57744008,
      "step": 99450
    },
    {
      "epoch": 14.813077152219243,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0057384304103880145,
      "loss": 0.8034,
      "num_input_tokens_seen": 57746888,
      "step": 99455
    },
    {
      "epoch": 14.813821864760202,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.005736896853517532,
      "loss": 0.8035,
      "num_input_tokens_seen": 57749768,
      "step": 99460
    },
    {
      "epoch": 14.814566577301163,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.005735363453137764,
      "loss": 0.7939,
      "num_input_tokens_seen": 57752872,
      "step": 99465
    },
    {
      "epoch": 14.815311289842121,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.005733830209274615,
      "loss": 0.775,
      "num_input_tokens_seen": 57755560,
      "step": 99470
    },
    {
      "epoch": 14.81605600238308,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.005732297121953985,
      "loss": 0.8059,
      "num_input_tokens_seen": 57758472,
      "step": 99475
    },
    {
      "epoch": 14.816800714924039,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.005730764191201772,
      "loss": 0.7701,
      "num_input_tokens_seen": 57761192,
      "step": 99480
    },
    {
      "epoch": 14.817545427465,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0057292314170438825,
      "loss": 0.7949,
      "num_input_tokens_seen": 57764136,
      "step": 99485
    },
    {
      "epoch": 14.818290140005958,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.005727698799506201,
      "loss": 0.7808,
      "num_input_tokens_seen": 57767112,
      "step": 99490
    },
    {
      "epoch": 14.819034852546917,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.005726166338614627,
      "loss": 0.8012,
      "num_input_tokens_seen": 57770248,
      "step": 99495
    },
    {
      "epoch": 14.819779565087876,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.005724634034395043,
      "loss": 0.7812,
      "num_input_tokens_seen": 57773128,
      "step": 99500
    },
    {
      "epoch": 14.820524277628834,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.005723101886873347,
      "loss": 0.8153,
      "num_input_tokens_seen": 57776328,
      "step": 99505
    },
    {
      "epoch": 14.821268990169795,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.005721569896075415,
      "loss": 0.7752,
      "num_input_tokens_seen": 57779016,
      "step": 99510
    },
    {
      "epoch": 14.822013702710754,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.005720038062027124,
      "loss": 0.7856,
      "num_input_tokens_seen": 57781832,
      "step": 99515
    },
    {
      "epoch": 14.822758415251712,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.005718506384754363,
      "loss": 0.7759,
      "num_input_tokens_seen": 57784616,
      "step": 99520
    },
    {
      "epoch": 14.823503127792671,
      "grad_norm": 0.052734375,
      "learning_rate": 0.005716974864282998,
      "loss": 0.7777,
      "num_input_tokens_seen": 57787624,
      "step": 99525
    },
    {
      "epoch": 14.824247840333632,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0057154435006389155,
      "loss": 0.7733,
      "num_input_tokens_seen": 57790440,
      "step": 99530
    },
    {
      "epoch": 14.82499255287459,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.005713912293847975,
      "loss": 0.7926,
      "num_input_tokens_seen": 57793192,
      "step": 99535
    },
    {
      "epoch": 14.82573726541555,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.005712381243936051,
      "loss": 0.7894,
      "num_input_tokens_seen": 57795976,
      "step": 99540
    },
    {
      "epoch": 14.826481977956508,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.005710850350929,
      "loss": 0.8,
      "num_input_tokens_seen": 57798632,
      "step": 99545
    },
    {
      "epoch": 14.827226690497469,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0057093196148526966,
      "loss": 0.8424,
      "num_input_tokens_seen": 57801352,
      "step": 99550
    },
    {
      "epoch": 14.827971403038427,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.005707789035732997,
      "loss": 0.786,
      "num_input_tokens_seen": 57804296,
      "step": 99555
    },
    {
      "epoch": 14.828716115579386,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.005706258613595751,
      "loss": 0.8149,
      "num_input_tokens_seen": 57807432,
      "step": 99560
    },
    {
      "epoch": 14.829460828120345,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.005704728348466825,
      "loss": 0.7733,
      "num_input_tokens_seen": 57810024,
      "step": 99565
    },
    {
      "epoch": 14.830205540661305,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.00570319824037206,
      "loss": 0.8072,
      "num_input_tokens_seen": 57812808,
      "step": 99570
    },
    {
      "epoch": 14.830950253202264,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.005701668289337319,
      "loss": 0.8117,
      "num_input_tokens_seen": 57815688,
      "step": 99575
    },
    {
      "epoch": 14.831694965743223,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0057001384953884356,
      "loss": 0.8088,
      "num_input_tokens_seen": 57818664,
      "step": 99580
    },
    {
      "epoch": 14.832439678284182,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.0056986088585512665,
      "loss": 0.8057,
      "num_input_tokens_seen": 57821608,
      "step": 99585
    },
    {
      "epoch": 14.833184390825142,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.005697079378851647,
      "loss": 0.7985,
      "num_input_tokens_seen": 57824264,
      "step": 99590
    },
    {
      "epoch": 14.833929103366101,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0056955500563154165,
      "loss": 0.8016,
      "num_input_tokens_seen": 57826952,
      "step": 99595
    },
    {
      "epoch": 14.83467381590706,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.005694020890968406,
      "loss": 0.7827,
      "num_input_tokens_seen": 57829864,
      "step": 99600
    },
    {
      "epoch": 14.835418528448018,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.005692491882836458,
      "loss": 0.7768,
      "num_input_tokens_seen": 57832456,
      "step": 99605
    },
    {
      "epoch": 14.836163240988979,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0056909630319454015,
      "loss": 0.7977,
      "num_input_tokens_seen": 57835624,
      "step": 99610
    },
    {
      "epoch": 14.836907953529938,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.005689434338321057,
      "loss": 0.7913,
      "num_input_tokens_seen": 57838504,
      "step": 99615
    },
    {
      "epoch": 14.837652666070897,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.005687905801989263,
      "loss": 0.7987,
      "num_input_tokens_seen": 57841480,
      "step": 99620
    },
    {
      "epoch": 14.838397378611855,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.005686377422975831,
      "loss": 0.7897,
      "num_input_tokens_seen": 57844328,
      "step": 99625
    },
    {
      "epoch": 14.839142091152816,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.005684849201306589,
      "loss": 0.8049,
      "num_input_tokens_seen": 57847240,
      "step": 99630
    },
    {
      "epoch": 14.839886803693775,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.005683321137007355,
      "loss": 0.7917,
      "num_input_tokens_seen": 57850120,
      "step": 99635
    },
    {
      "epoch": 14.840631516234733,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.005681793230103935,
      "loss": 0.809,
      "num_input_tokens_seen": 57853160,
      "step": 99640
    },
    {
      "epoch": 14.841376228775692,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.005680265480622153,
      "loss": 0.801,
      "num_input_tokens_seen": 57856040,
      "step": 99645
    },
    {
      "epoch": 14.842120941316653,
      "grad_norm": 0.05078125,
      "learning_rate": 0.005678737888587809,
      "loss": 0.7823,
      "num_input_tokens_seen": 57859016,
      "step": 99650
    },
    {
      "epoch": 14.842865653857611,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.005677210454026719,
      "loss": 0.7741,
      "num_input_tokens_seen": 57862024,
      "step": 99655
    },
    {
      "epoch": 14.84361036639857,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.005675683176964684,
      "loss": 0.7886,
      "num_input_tokens_seen": 57864968,
      "step": 99660
    },
    {
      "epoch": 14.844355078939529,
      "grad_norm": 0.095703125,
      "learning_rate": 0.005674156057427505,
      "loss": 0.8143,
      "num_input_tokens_seen": 57867656,
      "step": 99665
    },
    {
      "epoch": 14.84509979148049,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.005672629095440975,
      "loss": 0.8171,
      "num_input_tokens_seen": 57870568,
      "step": 99670
    },
    {
      "epoch": 14.845844504021448,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.005671102291030902,
      "loss": 0.8263,
      "num_input_tokens_seen": 57873288,
      "step": 99675
    },
    {
      "epoch": 14.846589216562407,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0056695756442230755,
      "loss": 0.7668,
      "num_input_tokens_seen": 57876328,
      "step": 99680
    },
    {
      "epoch": 14.847333929103366,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.005668049155043279,
      "loss": 0.8378,
      "num_input_tokens_seen": 57879208,
      "step": 99685
    },
    {
      "epoch": 14.848078641644324,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.005666522823517313,
      "loss": 0.7939,
      "num_input_tokens_seen": 57882184,
      "step": 99690
    },
    {
      "epoch": 14.848823354185285,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.005664996649670951,
      "loss": 0.7918,
      "num_input_tokens_seen": 57884808,
      "step": 99695
    },
    {
      "epoch": 14.849568066726244,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.00566347063352999,
      "loss": 0.8084,
      "num_input_tokens_seen": 57888072,
      "step": 99700
    },
    {
      "epoch": 14.850312779267203,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.005661944775120198,
      "loss": 0.829,
      "num_input_tokens_seen": 57890856,
      "step": 99705
    },
    {
      "epoch": 14.851057491808161,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.005660419074467364,
      "loss": 0.7895,
      "num_input_tokens_seen": 57894088,
      "step": 99710
    },
    {
      "epoch": 14.851802204349122,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.005658893531597256,
      "loss": 0.8454,
      "num_input_tokens_seen": 57896776,
      "step": 99715
    },
    {
      "epoch": 14.85254691689008,
      "grad_norm": 0.03515625,
      "learning_rate": 0.005657368146535642,
      "loss": 0.7994,
      "num_input_tokens_seen": 57899976,
      "step": 99720
    },
    {
      "epoch": 14.85329162943104,
      "grad_norm": 0.056640625,
      "learning_rate": 0.005655842919308306,
      "loss": 0.808,
      "num_input_tokens_seen": 57902728,
      "step": 99725
    },
    {
      "epoch": 14.854036341971998,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0056543178499410054,
      "loss": 0.79,
      "num_input_tokens_seen": 57905800,
      "step": 99730
    },
    {
      "epoch": 14.854781054512959,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.005652792938459505,
      "loss": 0.7813,
      "num_input_tokens_seen": 57908808,
      "step": 99735
    },
    {
      "epoch": 14.855525767053917,
      "grad_norm": 0.041015625,
      "learning_rate": 0.005651268184889565,
      "loss": 0.8117,
      "num_input_tokens_seen": 57911784,
      "step": 99740
    },
    {
      "epoch": 14.856270479594876,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.005649743589256951,
      "loss": 0.8044,
      "num_input_tokens_seen": 57914632,
      "step": 99745
    },
    {
      "epoch": 14.857015192135835,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.00564821915158741,
      "loss": 0.7957,
      "num_input_tokens_seen": 57917448,
      "step": 99750
    },
    {
      "epoch": 14.857759904676795,
      "grad_norm": 0.025390625,
      "learning_rate": 0.005646694871906707,
      "loss": 0.8091,
      "num_input_tokens_seen": 57920104,
      "step": 99755
    },
    {
      "epoch": 14.858504617217754,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.005645170750240589,
      "loss": 0.8182,
      "num_input_tokens_seen": 57922728,
      "step": 99760
    },
    {
      "epoch": 14.859249329758713,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.005643646786614797,
      "loss": 0.8034,
      "num_input_tokens_seen": 57925896,
      "step": 99765
    },
    {
      "epoch": 14.859994042299672,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.005642122981055091,
      "loss": 0.81,
      "num_input_tokens_seen": 57929096,
      "step": 99770
    },
    {
      "epoch": 14.860738754840632,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.005640599333587199,
      "loss": 0.8061,
      "num_input_tokens_seen": 57932136,
      "step": 99775
    },
    {
      "epoch": 14.861483467381591,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.005639075844236874,
      "loss": 0.7886,
      "num_input_tokens_seen": 57935400,
      "step": 99780
    },
    {
      "epoch": 14.86222817992255,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.005637552513029843,
      "loss": 0.8186,
      "num_input_tokens_seen": 57938152,
      "step": 99785
    },
    {
      "epoch": 14.862972892463509,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.005636029339991853,
      "loss": 0.7964,
      "num_input_tokens_seen": 57940840,
      "step": 99790
    },
    {
      "epoch": 14.863717605004469,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00563450632514863,
      "loss": 0.815,
      "num_input_tokens_seen": 57943880,
      "step": 99795
    },
    {
      "epoch": 14.864462317545428,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.005632983468525904,
      "loss": 0.8096,
      "num_input_tokens_seen": 57946792,
      "step": 99800
    },
    {
      "epoch": 14.865207030086387,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.005631460770149401,
      "loss": 0.8133,
      "num_input_tokens_seen": 57949672,
      "step": 99805
    },
    {
      "epoch": 14.865951742627345,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0056299382300448435,
      "loss": 0.7986,
      "num_input_tokens_seen": 57952648,
      "step": 99810
    },
    {
      "epoch": 14.866696455168306,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.005628415848237959,
      "loss": 0.7829,
      "num_input_tokens_seen": 57955176,
      "step": 99815
    },
    {
      "epoch": 14.867441167709265,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.005626893624754461,
      "loss": 0.7986,
      "num_input_tokens_seen": 57958280,
      "step": 99820
    },
    {
      "epoch": 14.868185880250223,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.005625371559620073,
      "loss": 0.7988,
      "num_input_tokens_seen": 57961224,
      "step": 99825
    },
    {
      "epoch": 14.868930592791182,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0056238496528604985,
      "loss": 0.7953,
      "num_input_tokens_seen": 57964232,
      "step": 99830
    },
    {
      "epoch": 14.86967530533214,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.00562232790450146,
      "loss": 0.7908,
      "num_input_tokens_seen": 57967176,
      "step": 99835
    },
    {
      "epoch": 14.870420017873101,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.005620806314568662,
      "loss": 0.8005,
      "num_input_tokens_seen": 57970024,
      "step": 99840
    },
    {
      "epoch": 14.87116473041406,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.005619284883087802,
      "loss": 0.8012,
      "num_input_tokens_seen": 57973000,
      "step": 99845
    },
    {
      "epoch": 14.871909442955019,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.005617763610084595,
      "loss": 0.8194,
      "num_input_tokens_seen": 57976296,
      "step": 99850
    },
    {
      "epoch": 14.87265415549598,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.005616242495584732,
      "loss": 0.8089,
      "num_input_tokens_seen": 57979080,
      "step": 99855
    },
    {
      "epoch": 14.873398868036938,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.00561472153961392,
      "loss": 0.8131,
      "num_input_tokens_seen": 57981800,
      "step": 99860
    },
    {
      "epoch": 14.874143580577897,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.005613200742197847,
      "loss": 0.7774,
      "num_input_tokens_seen": 57984424,
      "step": 99865
    },
    {
      "epoch": 14.874888293118856,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.00561168010336221,
      "loss": 0.7857,
      "num_input_tokens_seen": 57987304,
      "step": 99870
    },
    {
      "epoch": 14.875633005659815,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.005610159623132693,
      "loss": 0.7948,
      "num_input_tokens_seen": 57990088,
      "step": 99875
    },
    {
      "epoch": 14.876377718200775,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.005608639301534982,
      "loss": 0.8058,
      "num_input_tokens_seen": 57992904,
      "step": 99880
    },
    {
      "epoch": 14.877122430741734,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00560711913859477,
      "loss": 0.8146,
      "num_input_tokens_seen": 57995688,
      "step": 99885
    },
    {
      "epoch": 14.877867143282693,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0056055991343377295,
      "loss": 0.7995,
      "num_input_tokens_seen": 57998632,
      "step": 99890
    },
    {
      "epoch": 14.878611855823651,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0056040792887895475,
      "loss": 0.8123,
      "num_input_tokens_seen": 58001320,
      "step": 99895
    },
    {
      "epoch": 14.879356568364612,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.005602559601975892,
      "loss": 0.7994,
      "num_input_tokens_seen": 58004520,
      "step": 99900
    },
    {
      "epoch": 14.88010128090557,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.005601040073922444,
      "loss": 0.8009,
      "num_input_tokens_seen": 58007464,
      "step": 99905
    },
    {
      "epoch": 14.88084599344653,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0055995207046548685,
      "loss": 0.8164,
      "num_input_tokens_seen": 58010312,
      "step": 99910
    },
    {
      "epoch": 14.881590705987488,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.005598001494198841,
      "loss": 0.7861,
      "num_input_tokens_seen": 58013160,
      "step": 99915
    },
    {
      "epoch": 14.882335418528449,
      "grad_norm": 0.041015625,
      "learning_rate": 0.005596482442580022,
      "loss": 0.7962,
      "num_input_tokens_seen": 58016104,
      "step": 99920
    },
    {
      "epoch": 14.883080131069407,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.005594963549824076,
      "loss": 0.7856,
      "num_input_tokens_seen": 58019016,
      "step": 99925
    },
    {
      "epoch": 14.883824843610366,
      "grad_norm": 0.046875,
      "learning_rate": 0.005593444815956655,
      "loss": 0.7821,
      "num_input_tokens_seen": 58021864,
      "step": 99930
    },
    {
      "epoch": 14.884569556151325,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0055919262410034296,
      "loss": 0.7734,
      "num_input_tokens_seen": 58024584,
      "step": 99935
    },
    {
      "epoch": 14.885314268692285,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.005590407824990049,
      "loss": 0.8027,
      "num_input_tokens_seen": 58027624,
      "step": 99940
    },
    {
      "epoch": 14.886058981233244,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.005588889567942157,
      "loss": 0.8047,
      "num_input_tokens_seen": 58030792,
      "step": 99945
    },
    {
      "epoch": 14.886803693774203,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.0055873714698854195,
      "loss": 0.7945,
      "num_input_tokens_seen": 58034056,
      "step": 99950
    },
    {
      "epoch": 14.887548406315162,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.005585853530845473,
      "loss": 0.7854,
      "num_input_tokens_seen": 58036712,
      "step": 99955
    },
    {
      "epoch": 14.888293118856122,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.005584335750847957,
      "loss": 0.8197,
      "num_input_tokens_seen": 58039496,
      "step": 99960
    },
    {
      "epoch": 14.889037831397081,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.0055828181299185245,
      "loss": 0.7973,
      "num_input_tokens_seen": 58042216,
      "step": 99965
    },
    {
      "epoch": 14.88978254393804,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.005581300668082803,
      "loss": 0.8002,
      "num_input_tokens_seen": 58045128,
      "step": 99970
    },
    {
      "epoch": 14.890527256478999,
      "grad_norm": 0.0625,
      "learning_rate": 0.00557978336536644,
      "loss": 0.8302,
      "num_input_tokens_seen": 58047816,
      "step": 99975
    },
    {
      "epoch": 14.891271969019959,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0055782662217950555,
      "loss": 0.8083,
      "num_input_tokens_seen": 58050376,
      "step": 99980
    },
    {
      "epoch": 14.892016681560918,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.005576749237394295,
      "loss": 0.8201,
      "num_input_tokens_seen": 58053416,
      "step": 99985
    },
    {
      "epoch": 14.892761394101877,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.005575232412189777,
      "loss": 0.8135,
      "num_input_tokens_seen": 58056328,
      "step": 99990
    },
    {
      "epoch": 14.893506106642835,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.00557371574620713,
      "loss": 0.7955,
      "num_input_tokens_seen": 58059368,
      "step": 99995
    },
    {
      "epoch": 14.894250819183796,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.005572199239471975,
      "loss": 0.7827,
      "num_input_tokens_seen": 58062504,
      "step": 100000
    },
    {
      "epoch": 14.894995531724755,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0055706828920099255,
      "loss": 0.8077,
      "num_input_tokens_seen": 58065608,
      "step": 100005
    },
    {
      "epoch": 14.895740244265713,
      "grad_norm": 0.060546875,
      "learning_rate": 0.005569166703846612,
      "loss": 0.7922,
      "num_input_tokens_seen": 58068616,
      "step": 100010
    },
    {
      "epoch": 14.896484956806672,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0055676506750076335,
      "loss": 0.784,
      "num_input_tokens_seen": 58071400,
      "step": 100015
    },
    {
      "epoch": 14.897229669347631,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0055661348055186174,
      "loss": 0.7949,
      "num_input_tokens_seen": 58074344,
      "step": 100020
    },
    {
      "epoch": 14.897974381888591,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.005564619095405161,
      "loss": 0.8069,
      "num_input_tokens_seen": 58077256,
      "step": 100025
    },
    {
      "epoch": 14.89871909442955,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.00556310354469288,
      "loss": 0.7908,
      "num_input_tokens_seen": 58080072,
      "step": 100030
    },
    {
      "epoch": 14.899463806970509,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.005561588153407374,
      "loss": 0.7779,
      "num_input_tokens_seen": 58082696,
      "step": 100035
    },
    {
      "epoch": 14.900208519511468,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.005560072921574238,
      "loss": 0.7766,
      "num_input_tokens_seen": 58085832,
      "step": 100040
    },
    {
      "epoch": 14.900953232052428,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.005558557849219079,
      "loss": 0.7882,
      "num_input_tokens_seen": 58088776,
      "step": 100045
    },
    {
      "epoch": 14.901697944593387,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.005557042936367485,
      "loss": 0.8007,
      "num_input_tokens_seen": 58091688,
      "step": 100050
    },
    {
      "epoch": 14.902442657134346,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0055555281830450606,
      "loss": 0.8241,
      "num_input_tokens_seen": 58094408,
      "step": 100055
    },
    {
      "epoch": 14.903187369675305,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.005554013589277386,
      "loss": 0.8028,
      "num_input_tokens_seen": 58097288,
      "step": 100060
    },
    {
      "epoch": 14.903932082216265,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.005552499155090053,
      "loss": 0.8077,
      "num_input_tokens_seen": 58100360,
      "step": 100065
    },
    {
      "epoch": 14.904676794757224,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.005550984880508639,
      "loss": 0.7896,
      "num_input_tokens_seen": 58103336,
      "step": 100070
    },
    {
      "epoch": 14.905421507298183,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.005549470765558737,
      "loss": 0.8151,
      "num_input_tokens_seen": 58106120,
      "step": 100075
    },
    {
      "epoch": 14.906166219839141,
      "grad_norm": 0.0390625,
      "learning_rate": 0.005547956810265921,
      "loss": 0.7795,
      "num_input_tokens_seen": 58109224,
      "step": 100080
    },
    {
      "epoch": 14.906910932380102,
      "grad_norm": 0.060546875,
      "learning_rate": 0.005546443014655762,
      "loss": 0.831,
      "num_input_tokens_seen": 58111880,
      "step": 100085
    },
    {
      "epoch": 14.90765564492106,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.005544929378753847,
      "loss": 0.7764,
      "num_input_tokens_seen": 58114728,
      "step": 100090
    },
    {
      "epoch": 14.90840035746202,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.005543415902585735,
      "loss": 0.7855,
      "num_input_tokens_seen": 58117640,
      "step": 100095
    },
    {
      "epoch": 14.909145070002978,
      "grad_norm": 0.0546875,
      "learning_rate": 0.005541902586177004,
      "loss": 0.7851,
      "num_input_tokens_seen": 58120936,
      "step": 100100
    },
    {
      "epoch": 14.909889782543939,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.005540389429553212,
      "loss": 0.7901,
      "num_input_tokens_seen": 58124264,
      "step": 100105
    },
    {
      "epoch": 14.910634495084897,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.005538876432739931,
      "loss": 0.805,
      "num_input_tokens_seen": 58127080,
      "step": 100110
    },
    {
      "epoch": 14.911379207625856,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.005537363595762718,
      "loss": 0.8138,
      "num_input_tokens_seen": 58129928,
      "step": 100115
    },
    {
      "epoch": 14.912123920166815,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.005535850918647124,
      "loss": 0.7987,
      "num_input_tokens_seen": 58132904,
      "step": 100120
    },
    {
      "epoch": 14.912868632707776,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.005534338401418713,
      "loss": 0.7997,
      "num_input_tokens_seen": 58136008,
      "step": 100125
    },
    {
      "epoch": 14.913613345248734,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0055328260441030366,
      "loss": 0.7834,
      "num_input_tokens_seen": 58138856,
      "step": 100130
    },
    {
      "epoch": 14.914358057789693,
      "grad_norm": 0.056640625,
      "learning_rate": 0.00553131384672564,
      "loss": 0.8047,
      "num_input_tokens_seen": 58141800,
      "step": 100135
    },
    {
      "epoch": 14.915102770330652,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.005529801809312069,
      "loss": 0.8063,
      "num_input_tokens_seen": 58144680,
      "step": 100140
    },
    {
      "epoch": 14.915847482871612,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.005528289931887876,
      "loss": 0.7886,
      "num_input_tokens_seen": 58147624,
      "step": 100145
    },
    {
      "epoch": 14.916592195412571,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.005526778214478592,
      "loss": 0.7847,
      "num_input_tokens_seen": 58150216,
      "step": 100150
    },
    {
      "epoch": 14.91733690795353,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0055252666571097685,
      "loss": 0.7907,
      "num_input_tokens_seen": 58153096,
      "step": 100155
    },
    {
      "epoch": 14.918081620494489,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.005523755259806935,
      "loss": 0.7515,
      "num_input_tokens_seen": 58155784,
      "step": 100160
    },
    {
      "epoch": 14.91882633303545,
      "grad_norm": 0.0625,
      "learning_rate": 0.00552224402259562,
      "loss": 0.7907,
      "num_input_tokens_seen": 58159016,
      "step": 100165
    },
    {
      "epoch": 14.919571045576408,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.005520732945501364,
      "loss": 0.789,
      "num_input_tokens_seen": 58161736,
      "step": 100170
    },
    {
      "epoch": 14.920315758117367,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.005519222028549686,
      "loss": 0.7912,
      "num_input_tokens_seen": 58164968,
      "step": 100175
    },
    {
      "epoch": 14.921060470658325,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00551771127176612,
      "loss": 0.8005,
      "num_input_tokens_seen": 58168008,
      "step": 100180
    },
    {
      "epoch": 14.921805183199286,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.005516200675176181,
      "loss": 0.8052,
      "num_input_tokens_seen": 58171112,
      "step": 100185
    },
    {
      "epoch": 14.922549895740245,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.005514690238805397,
      "loss": 0.7846,
      "num_input_tokens_seen": 58173960,
      "step": 100190
    },
    {
      "epoch": 14.923294608281203,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.005513179962679281,
      "loss": 0.8337,
      "num_input_tokens_seen": 58177032,
      "step": 100195
    },
    {
      "epoch": 14.924039320822162,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.005511669846823348,
      "loss": 0.787,
      "num_input_tokens_seen": 58180040,
      "step": 100200
    },
    {
      "epoch": 14.924784033363121,
      "grad_norm": 0.07421875,
      "learning_rate": 0.005510159891263109,
      "loss": 0.8288,
      "num_input_tokens_seen": 58182920,
      "step": 100205
    },
    {
      "epoch": 14.925528745904082,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.005508650096024069,
      "loss": 0.7891,
      "num_input_tokens_seen": 58185672,
      "step": 100210
    },
    {
      "epoch": 14.92627345844504,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.005507140461131744,
      "loss": 0.7932,
      "num_input_tokens_seen": 58188392,
      "step": 100215
    },
    {
      "epoch": 14.927018170985999,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.005505630986611627,
      "loss": 0.8059,
      "num_input_tokens_seen": 58191432,
      "step": 100220
    },
    {
      "epoch": 14.927762883526958,
      "grad_norm": 0.056640625,
      "learning_rate": 0.00550412167248923,
      "loss": 0.7801,
      "num_input_tokens_seen": 58194472,
      "step": 100225
    },
    {
      "epoch": 14.928507596067918,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.005502612518790042,
      "loss": 0.7864,
      "num_input_tokens_seen": 58197384,
      "step": 100230
    },
    {
      "epoch": 14.929252308608877,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.005501103525539568,
      "loss": 0.8112,
      "num_input_tokens_seen": 58200488,
      "step": 100235
    },
    {
      "epoch": 14.929997021149836,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.005499594692763294,
      "loss": 0.7893,
      "num_input_tokens_seen": 58203176,
      "step": 100240
    },
    {
      "epoch": 14.930741733690795,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.005498086020486707,
      "loss": 0.8107,
      "num_input_tokens_seen": 58206152,
      "step": 100245
    },
    {
      "epoch": 14.931486446231755,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0054965775087353055,
      "loss": 0.792,
      "num_input_tokens_seen": 58209384,
      "step": 100250
    },
    {
      "epoch": 14.932231158772714,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.005495069157534568,
      "loss": 0.773,
      "num_input_tokens_seen": 58212168,
      "step": 100255
    },
    {
      "epoch": 14.932975871313673,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.005493560966909971,
      "loss": 0.7841,
      "num_input_tokens_seen": 58215208,
      "step": 100260
    },
    {
      "epoch": 14.933720583854631,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.005492052936887007,
      "loss": 0.8351,
      "num_input_tokens_seen": 58218184,
      "step": 100265
    },
    {
      "epoch": 14.934465296395592,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.005490545067491142,
      "loss": 0.7943,
      "num_input_tokens_seen": 58221256,
      "step": 100270
    },
    {
      "epoch": 14.93521000893655,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.00548903735874785,
      "loss": 0.808,
      "num_input_tokens_seen": 58224008,
      "step": 100275
    },
    {
      "epoch": 14.93595472147751,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.005487529810682609,
      "loss": 0.8181,
      "num_input_tokens_seen": 58226888,
      "step": 100280
    },
    {
      "epoch": 14.936699434018468,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.005486022423320885,
      "loss": 0.7869,
      "num_input_tokens_seen": 58229800,
      "step": 100285
    },
    {
      "epoch": 14.937444146559429,
      "grad_norm": 0.052734375,
      "learning_rate": 0.005484515196688139,
      "loss": 0.7885,
      "num_input_tokens_seen": 58232744,
      "step": 100290
    },
    {
      "epoch": 14.938188859100388,
      "grad_norm": 0.03515625,
      "learning_rate": 0.005483008130809839,
      "loss": 0.7987,
      "num_input_tokens_seen": 58235496,
      "step": 100295
    },
    {
      "epoch": 14.938933571641346,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0054815012257114415,
      "loss": 0.7756,
      "num_input_tokens_seen": 58238152,
      "step": 100300
    },
    {
      "epoch": 14.939678284182305,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0054799944814184105,
      "loss": 0.8069,
      "num_input_tokens_seen": 58240872,
      "step": 100305
    },
    {
      "epoch": 14.940422996723266,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.005478487897956192,
      "loss": 0.828,
      "num_input_tokens_seen": 58243624,
      "step": 100310
    },
    {
      "epoch": 14.941167709264224,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0054769814753502484,
      "loss": 0.8396,
      "num_input_tokens_seen": 58246376,
      "step": 100315
    },
    {
      "epoch": 14.941912421805183,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0054754752136260245,
      "loss": 0.7903,
      "num_input_tokens_seen": 58249128,
      "step": 100320
    },
    {
      "epoch": 14.942657134346142,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0054739691128089666,
      "loss": 0.8128,
      "num_input_tokens_seen": 58251848,
      "step": 100325
    },
    {
      "epoch": 14.943401846887102,
      "grad_norm": 0.046875,
      "learning_rate": 0.005472463172924519,
      "loss": 0.8142,
      "num_input_tokens_seen": 58254888,
      "step": 100330
    },
    {
      "epoch": 14.944146559428061,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0054709573939981165,
      "loss": 0.7929,
      "num_input_tokens_seen": 58257864,
      "step": 100335
    },
    {
      "epoch": 14.94489127196902,
      "grad_norm": 0.03125,
      "learning_rate": 0.00546945177605521,
      "loss": 0.7803,
      "num_input_tokens_seen": 58260808,
      "step": 100340
    },
    {
      "epoch": 14.945635984509979,
      "grad_norm": 0.03515625,
      "learning_rate": 0.005467946319121223,
      "loss": 0.7877,
      "num_input_tokens_seen": 58263432,
      "step": 100345
    },
    {
      "epoch": 14.946380697050937,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0054664410232216016,
      "loss": 0.7835,
      "num_input_tokens_seen": 58266280,
      "step": 100350
    },
    {
      "epoch": 14.947125409591898,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.005464935888381765,
      "loss": 0.8086,
      "num_input_tokens_seen": 58269256,
      "step": 100355
    },
    {
      "epoch": 14.947870122132857,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.005463430914627148,
      "loss": 0.8013,
      "num_input_tokens_seen": 58272072,
      "step": 100360
    },
    {
      "epoch": 14.948614834673815,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.005461926101983175,
      "loss": 0.8042,
      "num_input_tokens_seen": 58275144,
      "step": 100365
    },
    {
      "epoch": 14.949359547214776,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.005460421450475261,
      "loss": 0.8296,
      "num_input_tokens_seen": 58277768,
      "step": 100370
    },
    {
      "epoch": 14.950104259755735,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.005458916960128837,
      "loss": 0.7923,
      "num_input_tokens_seen": 58280328,
      "step": 100375
    },
    {
      "epoch": 14.950848972296694,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.005457412630969308,
      "loss": 0.8019,
      "num_input_tokens_seen": 58283304,
      "step": 100380
    },
    {
      "epoch": 14.951593684837652,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0054559084630221,
      "loss": 0.7888,
      "num_input_tokens_seen": 58286216,
      "step": 100385
    },
    {
      "epoch": 14.952338397378611,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0054544044563126165,
      "loss": 0.7859,
      "num_input_tokens_seen": 58289064,
      "step": 100390
    },
    {
      "epoch": 14.953083109919572,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.005452900610866269,
      "loss": 0.7923,
      "num_input_tokens_seen": 58292072,
      "step": 100395
    },
    {
      "epoch": 14.95382782246053,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.005451396926708462,
      "loss": 0.7975,
      "num_input_tokens_seen": 58294632,
      "step": 100400
    },
    {
      "epoch": 14.954572535001489,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.005449893403864594,
      "loss": 0.7888,
      "num_input_tokens_seen": 58297576,
      "step": 100405
    },
    {
      "epoch": 14.955317247542448,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.005448390042360076,
      "loss": 0.8125,
      "num_input_tokens_seen": 58300392,
      "step": 100410
    },
    {
      "epoch": 14.956061960083408,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.005446886842220297,
      "loss": 0.8093,
      "num_input_tokens_seen": 58303272,
      "step": 100415
    },
    {
      "epoch": 14.956806672624367,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.005445383803470658,
      "loss": 0.82,
      "num_input_tokens_seen": 58306472,
      "step": 100420
    },
    {
      "epoch": 14.957551385165326,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.005443880926136544,
      "loss": 0.7998,
      "num_input_tokens_seen": 58309224,
      "step": 100425
    },
    {
      "epoch": 14.958296097706285,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.005442378210243355,
      "loss": 0.7934,
      "num_input_tokens_seen": 58311912,
      "step": 100430
    },
    {
      "epoch": 14.959040810247245,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.005440875655816467,
      "loss": 0.7962,
      "num_input_tokens_seen": 58315048,
      "step": 100435
    },
    {
      "epoch": 14.959785522788204,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.005439373262881276,
      "loss": 0.795,
      "num_input_tokens_seen": 58318024,
      "step": 100440
    },
    {
      "epoch": 14.960530235329163,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.005437871031463154,
      "loss": 0.7942,
      "num_input_tokens_seen": 58320712,
      "step": 100445
    },
    {
      "epoch": 14.961274947870121,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0054363689615874785,
      "loss": 0.8345,
      "num_input_tokens_seen": 58323272,
      "step": 100450
    },
    {
      "epoch": 14.962019660411082,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0054348670532796345,
      "loss": 0.7954,
      "num_input_tokens_seen": 58326056,
      "step": 100455
    },
    {
      "epoch": 14.96276437295204,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00543336530656499,
      "loss": 0.7953,
      "num_input_tokens_seen": 58328936,
      "step": 100460
    },
    {
      "epoch": 14.963509085493,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.005431863721468916,
      "loss": 0.8036,
      "num_input_tokens_seen": 58331944,
      "step": 100465
    },
    {
      "epoch": 14.964253798033958,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0054303622980167755,
      "loss": 0.7761,
      "num_input_tokens_seen": 58334952,
      "step": 100470
    },
    {
      "epoch": 14.964998510574919,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.00542886103623394,
      "loss": 0.7993,
      "num_input_tokens_seen": 58337864,
      "step": 100475
    },
    {
      "epoch": 14.965743223115878,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.005427359936145772,
      "loss": 0.7739,
      "num_input_tokens_seen": 58340744,
      "step": 100480
    },
    {
      "epoch": 14.966487935656836,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.005425858997777624,
      "loss": 0.7849,
      "num_input_tokens_seen": 58343624,
      "step": 100485
    },
    {
      "epoch": 14.967232648197795,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.005424358221154861,
      "loss": 0.8039,
      "num_input_tokens_seen": 58346376,
      "step": 100490
    },
    {
      "epoch": 14.967977360738756,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.005422857606302829,
      "loss": 0.8137,
      "num_input_tokens_seen": 58349032,
      "step": 100495
    },
    {
      "epoch": 14.968722073279714,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.005421357153246889,
      "loss": 0.7866,
      "num_input_tokens_seen": 58351816,
      "step": 100500
    },
    {
      "epoch": 14.969466785820673,
      "grad_norm": 0.03125,
      "learning_rate": 0.00541985686201238,
      "loss": 0.8129,
      "num_input_tokens_seen": 58354792,
      "step": 100505
    },
    {
      "epoch": 14.970211498361632,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.005418356732624658,
      "loss": 0.7855,
      "num_input_tokens_seen": 58357672,
      "step": 100510
    },
    {
      "epoch": 14.970956210902592,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0054168567651090545,
      "loss": 0.7976,
      "num_input_tokens_seen": 58360552,
      "step": 100515
    },
    {
      "epoch": 14.971700923443551,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.005415356959490922,
      "loss": 0.7997,
      "num_input_tokens_seen": 58363528,
      "step": 100520
    },
    {
      "epoch": 14.97244563598451,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.005413857315795591,
      "loss": 0.7922,
      "num_input_tokens_seen": 58366088,
      "step": 100525
    },
    {
      "epoch": 14.973190348525469,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0054123578340483985,
      "loss": 0.8072,
      "num_input_tokens_seen": 58368840,
      "step": 100530
    },
    {
      "epoch": 14.973935061066427,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.005410858514274676,
      "loss": 0.8012,
      "num_input_tokens_seen": 58371592,
      "step": 100535
    },
    {
      "epoch": 14.974679773607388,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.005409359356499747,
      "loss": 0.8019,
      "num_input_tokens_seen": 58374376,
      "step": 100540
    },
    {
      "epoch": 14.975424486148347,
      "grad_norm": 0.080078125,
      "learning_rate": 0.005407860360748951,
      "loss": 0.8234,
      "num_input_tokens_seen": 58377288,
      "step": 100545
    },
    {
      "epoch": 14.976169198689306,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.005406361527047599,
      "loss": 0.7813,
      "num_input_tokens_seen": 58380136,
      "step": 100550
    },
    {
      "epoch": 14.976913911230266,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.005404862855421024,
      "loss": 0.7912,
      "num_input_tokens_seen": 58383336,
      "step": 100555
    },
    {
      "epoch": 14.977658623771225,
      "grad_norm": 0.037109375,
      "learning_rate": 0.005403364345894539,
      "loss": 0.8025,
      "num_input_tokens_seen": 58386408,
      "step": 100560
    },
    {
      "epoch": 14.978403336312184,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.005401865998493455,
      "loss": 0.7756,
      "num_input_tokens_seen": 58389288,
      "step": 100565
    },
    {
      "epoch": 14.979148048853142,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.005400367813243094,
      "loss": 0.7696,
      "num_input_tokens_seen": 58392072,
      "step": 100570
    },
    {
      "epoch": 14.979892761394101,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.005398869790168757,
      "loss": 0.7996,
      "num_input_tokens_seen": 58395240,
      "step": 100575
    },
    {
      "epoch": 14.980637473935062,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.005397371929295763,
      "loss": 0.7911,
      "num_input_tokens_seen": 58398056,
      "step": 100580
    },
    {
      "epoch": 14.98138218647602,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0053958742306494095,
      "loss": 0.7734,
      "num_input_tokens_seen": 58400904,
      "step": 100585
    },
    {
      "epoch": 14.98212689901698,
      "grad_norm": 0.03125,
      "learning_rate": 0.005394376694254996,
      "loss": 0.8302,
      "num_input_tokens_seen": 58403880,
      "step": 100590
    },
    {
      "epoch": 14.982871611557938,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.005392879320137828,
      "loss": 0.7916,
      "num_input_tokens_seen": 58406728,
      "step": 100595
    },
    {
      "epoch": 14.983616324098898,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.005391382108323201,
      "loss": 0.7952,
      "num_input_tokens_seen": 58409352,
      "step": 100600
    },
    {
      "epoch": 14.984361036639857,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0053898850588364085,
      "loss": 0.7745,
      "num_input_tokens_seen": 58412072,
      "step": 100605
    },
    {
      "epoch": 14.985105749180816,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.005388388171702735,
      "loss": 0.8188,
      "num_input_tokens_seen": 58414536,
      "step": 100610
    },
    {
      "epoch": 14.985850461721775,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0053868914469474794,
      "loss": 0.7943,
      "num_input_tokens_seen": 58417576,
      "step": 100615
    },
    {
      "epoch": 14.986595174262735,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.005385394884595917,
      "loss": 0.8049,
      "num_input_tokens_seen": 58420744,
      "step": 100620
    },
    {
      "epoch": 14.987339886803694,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.005383898484673341,
      "loss": 0.8041,
      "num_input_tokens_seen": 58423560,
      "step": 100625
    },
    {
      "epoch": 14.988084599344653,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.005382402247205023,
      "loss": 0.7841,
      "num_input_tokens_seen": 58426408,
      "step": 100630
    },
    {
      "epoch": 14.988829311885612,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0053809061722162495,
      "loss": 0.8152,
      "num_input_tokens_seen": 58429544,
      "step": 100635
    },
    {
      "epoch": 14.989574024426572,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.005379410259732288,
      "loss": 0.7977,
      "num_input_tokens_seen": 58432520,
      "step": 100640
    },
    {
      "epoch": 14.99031873696753,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.005377914509778409,
      "loss": 0.8179,
      "num_input_tokens_seen": 58435528,
      "step": 100645
    },
    {
      "epoch": 14.99106344950849,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.005376418922379888,
      "loss": 0.8351,
      "num_input_tokens_seen": 58438536,
      "step": 100650
    },
    {
      "epoch": 14.991808162049448,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.005374923497561989,
      "loss": 0.8065,
      "num_input_tokens_seen": 58441512,
      "step": 100655
    },
    {
      "epoch": 14.992552874590409,
      "grad_norm": 0.0625,
      "learning_rate": 0.005373428235349977,
      "loss": 0.7941,
      "num_input_tokens_seen": 58444360,
      "step": 100660
    },
    {
      "epoch": 14.993297587131368,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.005371933135769104,
      "loss": 0.7948,
      "num_input_tokens_seen": 58447464,
      "step": 100665
    },
    {
      "epoch": 14.994042299672326,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0053704381988446395,
      "loss": 0.8017,
      "num_input_tokens_seen": 58450440,
      "step": 100670
    },
    {
      "epoch": 14.994787012213285,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.00536894342460183,
      "loss": 0.8004,
      "num_input_tokens_seen": 58453224,
      "step": 100675
    },
    {
      "epoch": 14.995531724754246,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.005367448813065939,
      "loss": 0.8096,
      "num_input_tokens_seen": 58455976,
      "step": 100680
    },
    {
      "epoch": 14.996276437295204,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.00536595436426221,
      "loss": 0.8289,
      "num_input_tokens_seen": 58459016,
      "step": 100685
    },
    {
      "epoch": 14.997021149836163,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.005364460078215885,
      "loss": 0.8089,
      "num_input_tokens_seen": 58461800,
      "step": 100690
    },
    {
      "epoch": 14.997765862377122,
      "grad_norm": 0.072265625,
      "learning_rate": 0.005362965954952218,
      "loss": 0.784,
      "num_input_tokens_seen": 58464648,
      "step": 100695
    },
    {
      "epoch": 14.998510574918082,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.005361471994496442,
      "loss": 0.8199,
      "num_input_tokens_seen": 58467688,
      "step": 100700
    },
    {
      "epoch": 14.999255287459041,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.005359978196873806,
      "loss": 0.7957,
      "num_input_tokens_seen": 58470376,
      "step": 100705
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.005358484562109536,
      "loss": 0.8029,
      "num_input_tokens_seen": 58472760,
      "step": 100710
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.8003412485122681,
      "eval_runtime": 70.6033,
      "eval_samples_per_second": 42.264,
      "eval_steps_per_second": 10.566,
      "num_input_tokens_seen": 58472760,
      "step": 100710
    },
    {
      "epoch": 15.000744712540959,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.005356991090228875,
      "loss": 0.7785,
      "num_input_tokens_seen": 58475768,
      "step": 100715
    },
    {
      "epoch": 15.001489425081918,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0053554977812570485,
      "loss": 0.8096,
      "num_input_tokens_seen": 58478936,
      "step": 100720
    },
    {
      "epoch": 15.002234137622878,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.005354004635219285,
      "loss": 0.8076,
      "num_input_tokens_seen": 58481688,
      "step": 100725
    },
    {
      "epoch": 15.002978850163837,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.005352511652140809,
      "loss": 0.8007,
      "num_input_tokens_seen": 58484536,
      "step": 100730
    },
    {
      "epoch": 15.003723562704796,
      "grad_norm": 0.052734375,
      "learning_rate": 0.005351018832046837,
      "loss": 0.7867,
      "num_input_tokens_seen": 58487544,
      "step": 100735
    },
    {
      "epoch": 15.004468275245754,
      "grad_norm": 0.037109375,
      "learning_rate": 0.005349526174962603,
      "loss": 0.8126,
      "num_input_tokens_seen": 58490200,
      "step": 100740
    },
    {
      "epoch": 15.005212987786715,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.005348033680913309,
      "loss": 0.8048,
      "num_input_tokens_seen": 58493336,
      "step": 100745
    },
    {
      "epoch": 15.005957700327674,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0053465413499241814,
      "loss": 0.7925,
      "num_input_tokens_seen": 58496184,
      "step": 100750
    },
    {
      "epoch": 15.006702412868632,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.005345049182020422,
      "loss": 0.7839,
      "num_input_tokens_seen": 58499128,
      "step": 100755
    },
    {
      "epoch": 15.007447125409591,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.005343557177227249,
      "loss": 0.7975,
      "num_input_tokens_seen": 58502200,
      "step": 100760
    },
    {
      "epoch": 15.008191837950552,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.005342065335569862,
      "loss": 0.7999,
      "num_input_tokens_seen": 58505048,
      "step": 100765
    },
    {
      "epoch": 15.00893655049151,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00534057365707346,
      "loss": 0.799,
      "num_input_tokens_seen": 58507576,
      "step": 100770
    },
    {
      "epoch": 15.00968126303247,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.005339082141763254,
      "loss": 0.7849,
      "num_input_tokens_seen": 58510840,
      "step": 100775
    },
    {
      "epoch": 15.010425975573428,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0053375907896644325,
      "loss": 0.7711,
      "num_input_tokens_seen": 58513688,
      "step": 100780
    },
    {
      "epoch": 15.011170688114388,
      "grad_norm": 0.328125,
      "learning_rate": 0.0053360996008021975,
      "loss": 0.8082,
      "num_input_tokens_seen": 58516632,
      "step": 100785
    },
    {
      "epoch": 15.011915400655347,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00533460857520174,
      "loss": 0.7896,
      "num_input_tokens_seen": 58519352,
      "step": 100790
    },
    {
      "epoch": 15.012660113196306,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.005333117712888245,
      "loss": 0.7957,
      "num_input_tokens_seen": 58522040,
      "step": 100795
    },
    {
      "epoch": 15.013404825737265,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.005331627013886897,
      "loss": 0.7873,
      "num_input_tokens_seen": 58524888,
      "step": 100800
    },
    {
      "epoch": 15.014149538278225,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.00533013647822289,
      "loss": 0.8019,
      "num_input_tokens_seen": 58527896,
      "step": 100805
    },
    {
      "epoch": 15.014894250819184,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.005328646105921399,
      "loss": 0.8116,
      "num_input_tokens_seen": 58530744,
      "step": 100810
    },
    {
      "epoch": 15.015638963360143,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.005327155897007596,
      "loss": 0.7883,
      "num_input_tokens_seen": 58533656,
      "step": 100815
    },
    {
      "epoch": 15.016383675901102,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.005325665851506669,
      "loss": 0.7976,
      "num_input_tokens_seen": 58536792,
      "step": 100820
    },
    {
      "epoch": 15.017128388442062,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.005324175969443779,
      "loss": 0.8137,
      "num_input_tokens_seen": 58539640,
      "step": 100825
    },
    {
      "epoch": 15.01787310098302,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.005322686250844109,
      "loss": 0.7979,
      "num_input_tokens_seen": 58542328,
      "step": 100830
    },
    {
      "epoch": 15.01861781352398,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.005321196695732813,
      "loss": 0.8003,
      "num_input_tokens_seen": 58545208,
      "step": 100835
    },
    {
      "epoch": 15.019362526064938,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.005319707304135067,
      "loss": 0.8059,
      "num_input_tokens_seen": 58548152,
      "step": 100840
    },
    {
      "epoch": 15.020107238605899,
      "grad_norm": 0.037109375,
      "learning_rate": 0.005318218076076028,
      "loss": 0.782,
      "num_input_tokens_seen": 58551160,
      "step": 100845
    },
    {
      "epoch": 15.020851951146858,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0053167290115808475,
      "loss": 0.7809,
      "num_input_tokens_seen": 58554040,
      "step": 100850
    },
    {
      "epoch": 15.021596663687816,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.005315240110674696,
      "loss": 0.8056,
      "num_input_tokens_seen": 58556952,
      "step": 100855
    },
    {
      "epoch": 15.022341376228775,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.005313751373382718,
      "loss": 0.7931,
      "num_input_tokens_seen": 58559960,
      "step": 100860
    },
    {
      "epoch": 15.023086088769736,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.005312262799730067,
      "loss": 0.7942,
      "num_input_tokens_seen": 58563160,
      "step": 100865
    },
    {
      "epoch": 15.023830801310694,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.005310774389741884,
      "loss": 0.7882,
      "num_input_tokens_seen": 58566136,
      "step": 100870
    },
    {
      "epoch": 15.024575513851653,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.005309286143443325,
      "loss": 0.7812,
      "num_input_tokens_seen": 58569080,
      "step": 100875
    },
    {
      "epoch": 15.025320226392612,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.005307798060859524,
      "loss": 0.7785,
      "num_input_tokens_seen": 58571736,
      "step": 100880
    },
    {
      "epoch": 15.02606493893357,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.005306310142015626,
      "loss": 0.8242,
      "num_input_tokens_seen": 58574520,
      "step": 100885
    },
    {
      "epoch": 15.026809651474531,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.005304822386936768,
      "loss": 0.7848,
      "num_input_tokens_seen": 58577176,
      "step": 100890
    },
    {
      "epoch": 15.02755436401549,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.005303334795648077,
      "loss": 0.7746,
      "num_input_tokens_seen": 58580056,
      "step": 100895
    },
    {
      "epoch": 15.028299076556449,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.005301847368174695,
      "loss": 0.8004,
      "num_input_tokens_seen": 58582776,
      "step": 100900
    },
    {
      "epoch": 15.029043789097408,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.005300360104541738,
      "loss": 0.7981,
      "num_input_tokens_seen": 58585816,
      "step": 100905
    },
    {
      "epoch": 15.029788501638368,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.005298873004774345,
      "loss": 0.7938,
      "num_input_tokens_seen": 58588632,
      "step": 100910
    },
    {
      "epoch": 15.030533214179327,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0052973860688976335,
      "loss": 0.7902,
      "num_input_tokens_seen": 58591416,
      "step": 100915
    },
    {
      "epoch": 15.031277926720286,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.005295899296936716,
      "loss": 0.7995,
      "num_input_tokens_seen": 58594328,
      "step": 100920
    },
    {
      "epoch": 15.032022639261244,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.005294412688916725,
      "loss": 0.8017,
      "num_input_tokens_seen": 58597400,
      "step": 100925
    },
    {
      "epoch": 15.032767351802205,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.005292926244862764,
      "loss": 0.8083,
      "num_input_tokens_seen": 58600120,
      "step": 100930
    },
    {
      "epoch": 15.033512064343164,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.00529143996479995,
      "loss": 0.8046,
      "num_input_tokens_seen": 58603032,
      "step": 100935
    },
    {
      "epoch": 15.034256776884122,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.005289953848753385,
      "loss": 0.7959,
      "num_input_tokens_seen": 58605880,
      "step": 100940
    },
    {
      "epoch": 15.035001489425081,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.005288467896748187,
      "loss": 0.7957,
      "num_input_tokens_seen": 58608664,
      "step": 100945
    },
    {
      "epoch": 15.035746201966042,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.005286982108809448,
      "loss": 0.8016,
      "num_input_tokens_seen": 58611288,
      "step": 100950
    },
    {
      "epoch": 15.036490914507,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.005285496484962278,
      "loss": 0.8398,
      "num_input_tokens_seen": 58613944,
      "step": 100955
    },
    {
      "epoch": 15.03723562704796,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.005284011025231768,
      "loss": 0.795,
      "num_input_tokens_seen": 58616696,
      "step": 100960
    },
    {
      "epoch": 15.037980339588918,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.005282525729643022,
      "loss": 0.7944,
      "num_input_tokens_seen": 58619704,
      "step": 100965
    },
    {
      "epoch": 15.038725052129879,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.005281040598221126,
      "loss": 0.7944,
      "num_input_tokens_seen": 58622936,
      "step": 100970
    },
    {
      "epoch": 15.039469764670837,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.005279555630991168,
      "loss": 0.7797,
      "num_input_tokens_seen": 58625912,
      "step": 100975
    },
    {
      "epoch": 15.040214477211796,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0052780708279782425,
      "loss": 0.8104,
      "num_input_tokens_seen": 58628888,
      "step": 100980
    },
    {
      "epoch": 15.040959189752755,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.005276586189207427,
      "loss": 0.8042,
      "num_input_tokens_seen": 58631992,
      "step": 100985
    },
    {
      "epoch": 15.041703902293715,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.00527510171470381,
      "loss": 0.8044,
      "num_input_tokens_seen": 58634968,
      "step": 100990
    },
    {
      "epoch": 15.042448614834674,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.005273617404492456,
      "loss": 0.7681,
      "num_input_tokens_seen": 58637816,
      "step": 100995
    },
    {
      "epoch": 15.043193327375633,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0052721332585984595,
      "loss": 0.8128,
      "num_input_tokens_seen": 58640536,
      "step": 101000
    },
    {
      "epoch": 15.043938039916592,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0052706492770468825,
      "loss": 0.7952,
      "num_input_tokens_seen": 58643448,
      "step": 101005
    },
    {
      "epoch": 15.044682752457552,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.005269165459862791,
      "loss": 0.7831,
      "num_input_tokens_seen": 58646808,
      "step": 101010
    },
    {
      "epoch": 15.045427464998511,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.005267681807071267,
      "loss": 0.8093,
      "num_input_tokens_seen": 58649656,
      "step": 101015
    },
    {
      "epoch": 15.04617217753947,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.005266198318697362,
      "loss": 0.7712,
      "num_input_tokens_seen": 58652632,
      "step": 101020
    },
    {
      "epoch": 15.046916890080428,
      "grad_norm": 0.068359375,
      "learning_rate": 0.005264714994766146,
      "loss": 0.781,
      "num_input_tokens_seen": 58655480,
      "step": 101025
    },
    {
      "epoch": 15.047661602621389,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0052632318353026735,
      "loss": 0.7967,
      "num_input_tokens_seen": 58658520,
      "step": 101030
    },
    {
      "epoch": 15.048406315162348,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.005261748840332006,
      "loss": 0.7964,
      "num_input_tokens_seen": 58661432,
      "step": 101035
    },
    {
      "epoch": 15.049151027703306,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.005260266009879191,
      "loss": 0.7907,
      "num_input_tokens_seen": 58664408,
      "step": 101040
    },
    {
      "epoch": 15.049895740244265,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0052587833439692875,
      "loss": 0.8015,
      "num_input_tokens_seen": 58667544,
      "step": 101045
    },
    {
      "epoch": 15.050640452785226,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.005257300842627339,
      "loss": 0.7966,
      "num_input_tokens_seen": 58670776,
      "step": 101050
    },
    {
      "epoch": 15.051385165326185,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.005255818505878388,
      "loss": 0.8019,
      "num_input_tokens_seen": 58673496,
      "step": 101055
    },
    {
      "epoch": 15.052129877867143,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.005254336333747484,
      "loss": 0.7852,
      "num_input_tokens_seen": 58676600,
      "step": 101060
    },
    {
      "epoch": 15.052874590408102,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.005252854326259653,
      "loss": 0.7836,
      "num_input_tokens_seen": 58679480,
      "step": 101065
    },
    {
      "epoch": 15.05361930294906,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.00525137248343995,
      "loss": 0.7831,
      "num_input_tokens_seen": 58682648,
      "step": 101070
    },
    {
      "epoch": 15.054364015490021,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.005249890805313395,
      "loss": 0.8128,
      "num_input_tokens_seen": 58685336,
      "step": 101075
    },
    {
      "epoch": 15.05510872803098,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.005248409291905031,
      "loss": 0.7889,
      "num_input_tokens_seen": 58688280,
      "step": 101080
    },
    {
      "epoch": 15.055853440571939,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00524692794323988,
      "loss": 0.7915,
      "num_input_tokens_seen": 58691032,
      "step": 101085
    },
    {
      "epoch": 15.056598153112898,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.0052454467593429635,
      "loss": 0.832,
      "num_input_tokens_seen": 58693720,
      "step": 101090
    },
    {
      "epoch": 15.057342865653858,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.005243965740239315,
      "loss": 0.8045,
      "num_input_tokens_seen": 58696440,
      "step": 101095
    },
    {
      "epoch": 15.058087578194817,
      "grad_norm": 0.05078125,
      "learning_rate": 0.005242484885953946,
      "loss": 0.7916,
      "num_input_tokens_seen": 58699480,
      "step": 101100
    },
    {
      "epoch": 15.058832290735776,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.005241004196511882,
      "loss": 0.7844,
      "num_input_tokens_seen": 58702488,
      "step": 101105
    },
    {
      "epoch": 15.059577003276734,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.005239523671938129,
      "loss": 0.7887,
      "num_input_tokens_seen": 58705368,
      "step": 101110
    },
    {
      "epoch": 15.060321715817695,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.005238043312257706,
      "loss": 0.7956,
      "num_input_tokens_seen": 58708088,
      "step": 101115
    },
    {
      "epoch": 15.061066428358654,
      "grad_norm": 0.0390625,
      "learning_rate": 0.005236563117495622,
      "loss": 0.7886,
      "num_input_tokens_seen": 58711128,
      "step": 101120
    },
    {
      "epoch": 15.061811140899612,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.005235083087676881,
      "loss": 0.7903,
      "num_input_tokens_seen": 58714008,
      "step": 101125
    },
    {
      "epoch": 15.062555853440571,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.005233603222826486,
      "loss": 0.7781,
      "num_input_tokens_seen": 58716984,
      "step": 101130
    },
    {
      "epoch": 15.063300565981532,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0052321235229694325,
      "loss": 0.7878,
      "num_input_tokens_seen": 58719832,
      "step": 101135
    },
    {
      "epoch": 15.06404527852249,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.005230643988130732,
      "loss": 0.7792,
      "num_input_tokens_seen": 58722712,
      "step": 101140
    },
    {
      "epoch": 15.06478999106345,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.005229164618335365,
      "loss": 0.7985,
      "num_input_tokens_seen": 58725816,
      "step": 101145
    },
    {
      "epoch": 15.065534703604408,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.005227685413608337,
      "loss": 0.8006,
      "num_input_tokens_seen": 58728568,
      "step": 101150
    },
    {
      "epoch": 15.066279416145369,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.005226206373974626,
      "loss": 0.8095,
      "num_input_tokens_seen": 58731384,
      "step": 101155
    },
    {
      "epoch": 15.067024128686327,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00522472749945923,
      "loss": 0.8,
      "num_input_tokens_seen": 58734392,
      "step": 101160
    },
    {
      "epoch": 15.067768841227286,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.005223248790087129,
      "loss": 0.7992,
      "num_input_tokens_seen": 58737144,
      "step": 101165
    },
    {
      "epoch": 15.068513553768245,
      "grad_norm": 0.0390625,
      "learning_rate": 0.005221770245883296,
      "loss": 0.792,
      "num_input_tokens_seen": 58739992,
      "step": 101170
    },
    {
      "epoch": 15.069258266309205,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.005220291866872723,
      "loss": 0.8125,
      "num_input_tokens_seen": 58742840,
      "step": 101175
    },
    {
      "epoch": 15.070002978850164,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.005218813653080375,
      "loss": 0.7799,
      "num_input_tokens_seen": 58747128,
      "step": 101180
    },
    {
      "epoch": 15.070747691391123,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0052173356045312315,
      "loss": 0.7888,
      "num_input_tokens_seen": 58750200,
      "step": 101185
    },
    {
      "epoch": 15.071492403932082,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0052158577212502635,
      "loss": 0.7838,
      "num_input_tokens_seen": 58753304,
      "step": 101190
    },
    {
      "epoch": 15.072237116473042,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.005214380003262432,
      "loss": 0.7949,
      "num_input_tokens_seen": 58756344,
      "step": 101195
    },
    {
      "epoch": 15.072981829014001,
      "grad_norm": 0.07421875,
      "learning_rate": 0.005212902450592701,
      "loss": 0.7865,
      "num_input_tokens_seen": 58759032,
      "step": 101200
    },
    {
      "epoch": 15.07372654155496,
      "grad_norm": 0.029296875,
      "learning_rate": 0.005211425063266042,
      "loss": 0.782,
      "num_input_tokens_seen": 58762072,
      "step": 101205
    },
    {
      "epoch": 15.074471254095918,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.005209947841307407,
      "loss": 0.8028,
      "num_input_tokens_seen": 58764472,
      "step": 101210
    },
    {
      "epoch": 15.075215966636879,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.005208470784741748,
      "loss": 0.8334,
      "num_input_tokens_seen": 58767128,
      "step": 101215
    },
    {
      "epoch": 15.075960679177838,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.005206993893594029,
      "loss": 0.7951,
      "num_input_tokens_seen": 58770040,
      "step": 101220
    },
    {
      "epoch": 15.076705391718797,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.005205517167889188,
      "loss": 0.7927,
      "num_input_tokens_seen": 58772600,
      "step": 101225
    },
    {
      "epoch": 15.077450104259755,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.005204040607652187,
      "loss": 0.8248,
      "num_input_tokens_seen": 58775608,
      "step": 101230
    },
    {
      "epoch": 15.078194816800714,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0052025642129079576,
      "loss": 0.764,
      "num_input_tokens_seen": 58778456,
      "step": 101235
    },
    {
      "epoch": 15.078939529341675,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.005201087983681454,
      "loss": 0.7794,
      "num_input_tokens_seen": 58781080,
      "step": 101240
    },
    {
      "epoch": 15.079684241882633,
      "grad_norm": 0.04296875,
      "learning_rate": 0.00519961191999761,
      "loss": 0.809,
      "num_input_tokens_seen": 58784120,
      "step": 101245
    },
    {
      "epoch": 15.080428954423592,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.005198136021881355,
      "loss": 0.7881,
      "num_input_tokens_seen": 58787032,
      "step": 101250
    },
    {
      "epoch": 15.08117366696455,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.005196660289357637,
      "loss": 0.7999,
      "num_input_tokens_seen": 58789880,
      "step": 101255
    },
    {
      "epoch": 15.081918379505511,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.005195184722451379,
      "loss": 0.7944,
      "num_input_tokens_seen": 58793016,
      "step": 101260
    },
    {
      "epoch": 15.08266309204647,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.005193709321187509,
      "loss": 0.8031,
      "num_input_tokens_seen": 58796024,
      "step": 101265
    },
    {
      "epoch": 15.083407804587429,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.005192234085590949,
      "loss": 0.7824,
      "num_input_tokens_seen": 58798840,
      "step": 101270
    },
    {
      "epoch": 15.084152517128388,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.005190759015686631,
      "loss": 0.8004,
      "num_input_tokens_seen": 58801592,
      "step": 101275
    },
    {
      "epoch": 15.084897229669348,
      "grad_norm": 0.080078125,
      "learning_rate": 0.005189284111499465,
      "loss": 0.8134,
      "num_input_tokens_seen": 58804440,
      "step": 101280
    },
    {
      "epoch": 15.085641942210307,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.005187809373054378,
      "loss": 0.7725,
      "num_input_tokens_seen": 58807544,
      "step": 101285
    },
    {
      "epoch": 15.086386654751266,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.005186334800376277,
      "loss": 0.7897,
      "num_input_tokens_seen": 58810424,
      "step": 101290
    },
    {
      "epoch": 15.087131367292224,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.0051848603934900725,
      "loss": 0.801,
      "num_input_tokens_seen": 58813496,
      "step": 101295
    },
    {
      "epoch": 15.087876079833185,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.005183386152420681,
      "loss": 0.7964,
      "num_input_tokens_seen": 58816536,
      "step": 101300
    },
    {
      "epoch": 15.088620792374144,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.005181912077192997,
      "loss": 0.7955,
      "num_input_tokens_seen": 58819672,
      "step": 101305
    },
    {
      "epoch": 15.089365504915103,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.005180438167831937,
      "loss": 0.7855,
      "num_input_tokens_seen": 58822904,
      "step": 101310
    },
    {
      "epoch": 15.090110217456061,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.005178964424362392,
      "loss": 0.7955,
      "num_input_tokens_seen": 58825688,
      "step": 101315
    },
    {
      "epoch": 15.090854929997022,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.005177490846809262,
      "loss": 0.8023,
      "num_input_tokens_seen": 58828568,
      "step": 101320
    },
    {
      "epoch": 15.09159964253798,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.005176017435197435,
      "loss": 0.8142,
      "num_input_tokens_seen": 58831448,
      "step": 101325
    },
    {
      "epoch": 15.09234435507894,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.005174544189551816,
      "loss": 0.779,
      "num_input_tokens_seen": 58834136,
      "step": 101330
    },
    {
      "epoch": 15.093089067619898,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0051730711098972836,
      "loss": 0.7855,
      "num_input_tokens_seen": 58837336,
      "step": 101335
    },
    {
      "epoch": 15.093833780160859,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.005171598196258724,
      "loss": 0.7959,
      "num_input_tokens_seen": 58840376,
      "step": 101340
    },
    {
      "epoch": 15.094578492701817,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.005170125448661029,
      "loss": 0.7884,
      "num_input_tokens_seen": 58843608,
      "step": 101345
    },
    {
      "epoch": 15.095323205242776,
      "grad_norm": 0.05859375,
      "learning_rate": 0.005168652867129068,
      "loss": 0.7961,
      "num_input_tokens_seen": 58846488,
      "step": 101350
    },
    {
      "epoch": 15.096067917783735,
      "grad_norm": 0.0390625,
      "learning_rate": 0.005167180451687731,
      "loss": 0.8193,
      "num_input_tokens_seen": 58849624,
      "step": 101355
    },
    {
      "epoch": 15.096812630324695,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.005165708202361881,
      "loss": 0.8067,
      "num_input_tokens_seen": 58852696,
      "step": 101360
    },
    {
      "epoch": 15.097557342865654,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.005164236119176401,
      "loss": 0.8154,
      "num_input_tokens_seen": 58855288,
      "step": 101365
    },
    {
      "epoch": 15.098302055406613,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.005162764202156155,
      "loss": 0.8044,
      "num_input_tokens_seen": 58858008,
      "step": 101370
    },
    {
      "epoch": 15.099046767947572,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.005161292451326004,
      "loss": 0.7881,
      "num_input_tokens_seen": 58860696,
      "step": 101375
    },
    {
      "epoch": 15.099791480488532,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.005159820866710824,
      "loss": 0.7967,
      "num_input_tokens_seen": 58863448,
      "step": 101380
    },
    {
      "epoch": 15.100536193029491,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.005158349448335468,
      "loss": 0.7892,
      "num_input_tokens_seen": 58866104,
      "step": 101385
    },
    {
      "epoch": 15.10128090557045,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.005156878196224795,
      "loss": 0.7826,
      "num_input_tokens_seen": 58868952,
      "step": 101390
    },
    {
      "epoch": 15.102025618111409,
      "grad_norm": 0.03515625,
      "learning_rate": 0.005155407110403657,
      "loss": 0.8004,
      "num_input_tokens_seen": 58871800,
      "step": 101395
    },
    {
      "epoch": 15.102770330652369,
      "grad_norm": 0.05859375,
      "learning_rate": 0.005153936190896915,
      "loss": 0.8031,
      "num_input_tokens_seen": 58874488,
      "step": 101400
    },
    {
      "epoch": 15.103515043193328,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0051524654377294084,
      "loss": 0.7992,
      "num_input_tokens_seen": 58877528,
      "step": 101405
    },
    {
      "epoch": 15.104259755734287,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0051509948509259945,
      "loss": 0.7856,
      "num_input_tokens_seen": 58880632,
      "step": 101410
    },
    {
      "epoch": 15.105004468275245,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.005149524430511514,
      "loss": 0.7942,
      "num_input_tokens_seen": 58883480,
      "step": 101415
    },
    {
      "epoch": 15.105749180816204,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0051480541765108,
      "loss": 0.7968,
      "num_input_tokens_seen": 58886392,
      "step": 101420
    },
    {
      "epoch": 15.106493893357165,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0051465840889487035,
      "loss": 0.7812,
      "num_input_tokens_seen": 58889272,
      "step": 101425
    },
    {
      "epoch": 15.107238605898123,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00514511416785005,
      "loss": 0.8003,
      "num_input_tokens_seen": 58892216,
      "step": 101430
    },
    {
      "epoch": 15.107983318439082,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.00514364441323968,
      "loss": 0.8139,
      "num_input_tokens_seen": 58895096,
      "step": 101435
    },
    {
      "epoch": 15.108728030980041,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0051421748251424156,
      "loss": 0.7784,
      "num_input_tokens_seen": 58897880,
      "step": 101440
    },
    {
      "epoch": 15.109472743521001,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.005140705403583094,
      "loss": 0.8077,
      "num_input_tokens_seen": 58900440,
      "step": 101445
    },
    {
      "epoch": 15.11021745606196,
      "grad_norm": 0.044921875,
      "learning_rate": 0.005139236148586534,
      "loss": 0.7804,
      "num_input_tokens_seen": 58903352,
      "step": 101450
    },
    {
      "epoch": 15.110962168602919,
      "grad_norm": 0.05078125,
      "learning_rate": 0.005137767060177556,
      "loss": 0.8202,
      "num_input_tokens_seen": 58906360,
      "step": 101455
    },
    {
      "epoch": 15.111706881143878,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.005136298138380982,
      "loss": 0.7972,
      "num_input_tokens_seen": 58909272,
      "step": 101460
    },
    {
      "epoch": 15.112451593684838,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.005134829383221618,
      "loss": 0.7721,
      "num_input_tokens_seen": 58911896,
      "step": 101465
    },
    {
      "epoch": 15.113196306225797,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0051333607947242925,
      "loss": 0.7955,
      "num_input_tokens_seen": 58914680,
      "step": 101470
    },
    {
      "epoch": 15.113941018766756,
      "grad_norm": 0.181640625,
      "learning_rate": 0.005131892372913802,
      "loss": 0.8567,
      "num_input_tokens_seen": 58917688,
      "step": 101475
    },
    {
      "epoch": 15.114685731307715,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.005130424117814967,
      "loss": 0.7983,
      "num_input_tokens_seen": 58920792,
      "step": 101480
    },
    {
      "epoch": 15.115430443848675,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00512895602945258,
      "loss": 0.8217,
      "num_input_tokens_seen": 58923576,
      "step": 101485
    },
    {
      "epoch": 15.116175156389634,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.005127488107851454,
      "loss": 0.785,
      "num_input_tokens_seen": 58926584,
      "step": 101490
    },
    {
      "epoch": 15.116919868930593,
      "grad_norm": 0.083984375,
      "learning_rate": 0.005126020353036381,
      "loss": 0.787,
      "num_input_tokens_seen": 58929432,
      "step": 101495
    },
    {
      "epoch": 15.117664581471551,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.005124552765032153,
      "loss": 0.7904,
      "num_input_tokens_seen": 58932088,
      "step": 101500
    },
    {
      "epoch": 15.118409294012512,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.005123085343863576,
      "loss": 0.8045,
      "num_input_tokens_seen": 58935224,
      "step": 101505
    },
    {
      "epoch": 15.11915400655347,
      "grad_norm": 0.0625,
      "learning_rate": 0.005121618089555428,
      "loss": 0.8085,
      "num_input_tokens_seen": 58938296,
      "step": 101510
    },
    {
      "epoch": 15.11989871909443,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.005120151002132507,
      "loss": 0.8114,
      "num_input_tokens_seen": 58941336,
      "step": 101515
    },
    {
      "epoch": 15.120643431635388,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.005118684081619595,
      "loss": 0.8066,
      "num_input_tokens_seen": 58944216,
      "step": 101520
    },
    {
      "epoch": 15.121388144176349,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.00511721732804147,
      "loss": 0.7839,
      "num_input_tokens_seen": 58947000,
      "step": 101525
    },
    {
      "epoch": 15.122132856717307,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.005115750741422916,
      "loss": 0.7975,
      "num_input_tokens_seen": 58949496,
      "step": 101530
    },
    {
      "epoch": 15.122877569258266,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.005114284321788701,
      "loss": 0.7936,
      "num_input_tokens_seen": 58952376,
      "step": 101535
    },
    {
      "epoch": 15.123622281799225,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0051128180691636105,
      "loss": 0.7909,
      "num_input_tokens_seen": 58955480,
      "step": 101540
    },
    {
      "epoch": 15.124366994340185,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.005111351983572405,
      "loss": 0.7797,
      "num_input_tokens_seen": 58958552,
      "step": 101545
    },
    {
      "epoch": 15.125111706881144,
      "grad_norm": 0.0390625,
      "learning_rate": 0.005109886065039862,
      "loss": 0.8039,
      "num_input_tokens_seen": 58961752,
      "step": 101550
    },
    {
      "epoch": 15.125856419422103,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.005108420313590737,
      "loss": 0.7867,
      "num_input_tokens_seen": 58964632,
      "step": 101555
    },
    {
      "epoch": 15.126601131963062,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.005106954729249803,
      "loss": 0.8033,
      "num_input_tokens_seen": 58967832,
      "step": 101560
    },
    {
      "epoch": 15.127345844504022,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.005105489312041809,
      "loss": 0.7843,
      "num_input_tokens_seen": 58970776,
      "step": 101565
    },
    {
      "epoch": 15.128090557044981,
      "grad_norm": 0.060546875,
      "learning_rate": 0.005104024061991522,
      "loss": 0.7991,
      "num_input_tokens_seen": 58974008,
      "step": 101570
    },
    {
      "epoch": 15.12883526958594,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.005102558979123689,
      "loss": 0.7928,
      "num_input_tokens_seen": 58976632,
      "step": 101575
    },
    {
      "epoch": 15.129579982126899,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.00510109406346306,
      "loss": 0.7732,
      "num_input_tokens_seen": 58979352,
      "step": 101580
    },
    {
      "epoch": 15.130324694667857,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.005099629315034388,
      "loss": 0.7878,
      "num_input_tokens_seen": 58982264,
      "step": 101585
    },
    {
      "epoch": 15.131069407208818,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.005098164733862419,
      "loss": 0.8385,
      "num_input_tokens_seen": 58985112,
      "step": 101590
    },
    {
      "epoch": 15.131814119749777,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.005096700319971892,
      "loss": 0.7794,
      "num_input_tokens_seen": 58988024,
      "step": 101595
    },
    {
      "epoch": 15.132558832290735,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.005095236073387543,
      "loss": 0.7907,
      "num_input_tokens_seen": 58991192,
      "step": 101600
    },
    {
      "epoch": 15.133303544831694,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.005093771994134119,
      "loss": 0.7746,
      "num_input_tokens_seen": 58994040,
      "step": 101605
    },
    {
      "epoch": 15.134048257372655,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.005092308082236349,
      "loss": 0.7769,
      "num_input_tokens_seen": 58997176,
      "step": 101610
    },
    {
      "epoch": 15.134792969913613,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0050908443377189585,
      "loss": 0.7828,
      "num_input_tokens_seen": 58999736,
      "step": 101615
    },
    {
      "epoch": 15.135537682454572,
      "grad_norm": 0.05078125,
      "learning_rate": 0.005089380760606688,
      "loss": 0.8026,
      "num_input_tokens_seen": 59002776,
      "step": 101620
    },
    {
      "epoch": 15.136282394995531,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0050879173509242506,
      "loss": 0.788,
      "num_input_tokens_seen": 59005656,
      "step": 101625
    },
    {
      "epoch": 15.137027107536491,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.005086454108696381,
      "loss": 0.7929,
      "num_input_tokens_seen": 59008728,
      "step": 101630
    },
    {
      "epoch": 15.13777182007745,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.005084991033947789,
      "loss": 0.8391,
      "num_input_tokens_seen": 59011704,
      "step": 101635
    },
    {
      "epoch": 15.138516532618409,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.005083528126703201,
      "loss": 0.8311,
      "num_input_tokens_seen": 59014616,
      "step": 101640
    },
    {
      "epoch": 15.139261245159368,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.005082065386987326,
      "loss": 0.7982,
      "num_input_tokens_seen": 59017368,
      "step": 101645
    },
    {
      "epoch": 15.140005957700328,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.005080602814824875,
      "loss": 0.8003,
      "num_input_tokens_seen": 59020120,
      "step": 101650
    },
    {
      "epoch": 15.140750670241287,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.005079140410240556,
      "loss": 0.8139,
      "num_input_tokens_seen": 59023064,
      "step": 101655
    },
    {
      "epoch": 15.141495382782246,
      "grad_norm": 0.06640625,
      "learning_rate": 0.00507767817325908,
      "loss": 0.7768,
      "num_input_tokens_seen": 59025944,
      "step": 101660
    },
    {
      "epoch": 15.142240095323205,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.005076216103905146,
      "loss": 0.8157,
      "num_input_tokens_seen": 59028696,
      "step": 101665
    },
    {
      "epoch": 15.142984807864165,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.005074754202203451,
      "loss": 0.8001,
      "num_input_tokens_seen": 59031704,
      "step": 101670
    },
    {
      "epoch": 15.143729520405124,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.005073292468178701,
      "loss": 0.7925,
      "num_input_tokens_seen": 59034520,
      "step": 101675
    },
    {
      "epoch": 15.144474232946083,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.005071830901855579,
      "loss": 0.7997,
      "num_input_tokens_seen": 59037368,
      "step": 101680
    },
    {
      "epoch": 15.145218945487041,
      "grad_norm": 0.041015625,
      "learning_rate": 0.005070369503258791,
      "loss": 0.7999,
      "num_input_tokens_seen": 59039960,
      "step": 101685
    },
    {
      "epoch": 15.145963658028002,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.005068908272413017,
      "loss": 0.7972,
      "num_input_tokens_seen": 59042808,
      "step": 101690
    },
    {
      "epoch": 15.14670837056896,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.005067447209342939,
      "loss": 0.7843,
      "num_input_tokens_seen": 59045592,
      "step": 101695
    },
    {
      "epoch": 15.14745308310992,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.00506598631407325,
      "loss": 0.8024,
      "num_input_tokens_seen": 59048216,
      "step": 101700
    },
    {
      "epoch": 15.148197795650878,
      "grad_norm": 0.05078125,
      "learning_rate": 0.005064525586628621,
      "loss": 0.8024,
      "num_input_tokens_seen": 59050936,
      "step": 101705
    },
    {
      "epoch": 15.148942508191839,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.005063065027033739,
      "loss": 0.7962,
      "num_input_tokens_seen": 59053784,
      "step": 101710
    },
    {
      "epoch": 15.149687220732797,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.005061604635313276,
      "loss": 0.7935,
      "num_input_tokens_seen": 59056568,
      "step": 101715
    },
    {
      "epoch": 15.150431933273756,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.0050601444114919,
      "loss": 0.7915,
      "num_input_tokens_seen": 59059576,
      "step": 101720
    },
    {
      "epoch": 15.151176645814715,
      "grad_norm": 0.064453125,
      "learning_rate": 0.005058684355594278,
      "loss": 0.7966,
      "num_input_tokens_seen": 59062296,
      "step": 101725
    },
    {
      "epoch": 15.151921358355676,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.005057224467645082,
      "loss": 0.791,
      "num_input_tokens_seen": 59065304,
      "step": 101730
    },
    {
      "epoch": 15.152666070896634,
      "grad_norm": 0.037109375,
      "learning_rate": 0.005055764747668975,
      "loss": 0.8071,
      "num_input_tokens_seen": 59068184,
      "step": 101735
    },
    {
      "epoch": 15.153410783437593,
      "grad_norm": 0.0546875,
      "learning_rate": 0.005054305195690612,
      "loss": 0.8094,
      "num_input_tokens_seen": 59070840,
      "step": 101740
    },
    {
      "epoch": 15.154155495978552,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.005052845811734657,
      "loss": 0.7936,
      "num_input_tokens_seen": 59073624,
      "step": 101745
    },
    {
      "epoch": 15.15490020851951,
      "grad_norm": 0.0625,
      "learning_rate": 0.00505138659582576,
      "loss": 0.7777,
      "num_input_tokens_seen": 59076664,
      "step": 101750
    },
    {
      "epoch": 15.155644921060471,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.005049927547988578,
      "loss": 0.7986,
      "num_input_tokens_seen": 59079416,
      "step": 101755
    },
    {
      "epoch": 15.15638963360143,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.005048468668247753,
      "loss": 0.7955,
      "num_input_tokens_seen": 59082296,
      "step": 101760
    },
    {
      "epoch": 15.157134346142389,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.005047009956627941,
      "loss": 0.7854,
      "num_input_tokens_seen": 59085272,
      "step": 101765
    },
    {
      "epoch": 15.157879058683347,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.005045551413153778,
      "loss": 0.8013,
      "num_input_tokens_seen": 59088216,
      "step": 101770
    },
    {
      "epoch": 15.158623771224308,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.005044093037849903,
      "loss": 0.7873,
      "num_input_tokens_seen": 59091000,
      "step": 101775
    },
    {
      "epoch": 15.159368483765267,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.005042634830740963,
      "loss": 0.7936,
      "num_input_tokens_seen": 59093880,
      "step": 101780
    },
    {
      "epoch": 15.160113196306225,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.005041176791851587,
      "loss": 0.7893,
      "num_input_tokens_seen": 59096984,
      "step": 101785
    },
    {
      "epoch": 15.160857908847184,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0050397189212064075,
      "loss": 0.8013,
      "num_input_tokens_seen": 59100184,
      "step": 101790
    },
    {
      "epoch": 15.161602621388145,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0050382612188300474,
      "loss": 0.7729,
      "num_input_tokens_seen": 59103224,
      "step": 101795
    },
    {
      "epoch": 15.162347333929103,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0050368036847471456,
      "loss": 0.7951,
      "num_input_tokens_seen": 59106424,
      "step": 101800
    },
    {
      "epoch": 15.163092046470062,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.005035346318982315,
      "loss": 0.771,
      "num_input_tokens_seen": 59109080,
      "step": 101805
    },
    {
      "epoch": 15.163836759011021,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0050338891215601835,
      "loss": 0.8312,
      "num_input_tokens_seen": 59111672,
      "step": 101810
    },
    {
      "epoch": 15.164581471551982,
      "grad_norm": 0.037109375,
      "learning_rate": 0.005032432092505368,
      "loss": 0.7882,
      "num_input_tokens_seen": 59114456,
      "step": 101815
    },
    {
      "epoch": 15.16532618409294,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.005030975231842475,
      "loss": 0.7922,
      "num_input_tokens_seen": 59117432,
      "step": 101820
    },
    {
      "epoch": 15.166070896633899,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.00502951853959613,
      "loss": 0.8338,
      "num_input_tokens_seen": 59120184,
      "step": 101825
    },
    {
      "epoch": 15.166815609174858,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.005028062015790932,
      "loss": 0.808,
      "num_input_tokens_seen": 59123256,
      "step": 101830
    },
    {
      "epoch": 15.167560321715818,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.005026605660451494,
      "loss": 0.8074,
      "num_input_tokens_seen": 59126008,
      "step": 101835
    },
    {
      "epoch": 15.168305034256777,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.005025149473602411,
      "loss": 0.7747,
      "num_input_tokens_seen": 59129368,
      "step": 101840
    },
    {
      "epoch": 15.169049746797736,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.005023693455268297,
      "loss": 0.8116,
      "num_input_tokens_seen": 59132056,
      "step": 101845
    },
    {
      "epoch": 15.169794459338695,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.005022237605473741,
      "loss": 0.8159,
      "num_input_tokens_seen": 59134872,
      "step": 101850
    },
    {
      "epoch": 15.170539171879655,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.00502078192424334,
      "loss": 0.8195,
      "num_input_tokens_seen": 59137752,
      "step": 101855
    },
    {
      "epoch": 15.171283884420614,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.005019326411601688,
      "loss": 0.8005,
      "num_input_tokens_seen": 59140696,
      "step": 101860
    },
    {
      "epoch": 15.172028596961573,
      "grad_norm": 0.0390625,
      "learning_rate": 0.005017871067573365,
      "loss": 0.8025,
      "num_input_tokens_seen": 59144312,
      "step": 101865
    },
    {
      "epoch": 15.172773309502531,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0050164158921829705,
      "loss": 0.7992,
      "num_input_tokens_seen": 59147512,
      "step": 101870
    },
    {
      "epoch": 15.173518022043492,
      "grad_norm": 0.037109375,
      "learning_rate": 0.005014960885455077,
      "loss": 0.7817,
      "num_input_tokens_seen": 59150360,
      "step": 101875
    },
    {
      "epoch": 15.17426273458445,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.005013506047414278,
      "loss": 0.8237,
      "num_input_tokens_seen": 59153208,
      "step": 101880
    },
    {
      "epoch": 15.17500744712541,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.005012051378085139,
      "loss": 0.7885,
      "num_input_tokens_seen": 59156088,
      "step": 101885
    },
    {
      "epoch": 15.175752159666368,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.005010596877492244,
      "loss": 0.7905,
      "num_input_tokens_seen": 59159000,
      "step": 101890
    },
    {
      "epoch": 15.176496872207329,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.005009142545660163,
      "loss": 0.7979,
      "num_input_tokens_seen": 59162072,
      "step": 101895
    },
    {
      "epoch": 15.177241584748288,
      "grad_norm": 0.046875,
      "learning_rate": 0.00500768838261346,
      "loss": 0.8116,
      "num_input_tokens_seen": 59164792,
      "step": 101900
    },
    {
      "epoch": 15.177986297289246,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.005006234388376709,
      "loss": 0.7946,
      "num_input_tokens_seen": 59167640,
      "step": 101905
    },
    {
      "epoch": 15.178731009830205,
      "grad_norm": 0.060546875,
      "learning_rate": 0.005004780562974468,
      "loss": 0.8042,
      "num_input_tokens_seen": 59170648,
      "step": 101910
    },
    {
      "epoch": 15.179475722371166,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.005003326906431305,
      "loss": 0.7814,
      "num_input_tokens_seen": 59173432,
      "step": 101915
    },
    {
      "epoch": 15.180220434912124,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.005001873418771773,
      "loss": 0.8269,
      "num_input_tokens_seen": 59176120,
      "step": 101920
    },
    {
      "epoch": 15.180965147453083,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00500042010002043,
      "loss": 0.8004,
      "num_input_tokens_seen": 59178904,
      "step": 101925
    },
    {
      "epoch": 15.181709859994042,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.004998966950201819,
      "loss": 0.799,
      "num_input_tokens_seen": 59181624,
      "step": 101930
    },
    {
      "epoch": 15.182454572535,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.004997513969340503,
      "loss": 0.8,
      "num_input_tokens_seen": 59184504,
      "step": 101935
    },
    {
      "epoch": 15.183199285075961,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.004996061157461021,
      "loss": 0.7912,
      "num_input_tokens_seen": 59187128,
      "step": 101940
    },
    {
      "epoch": 15.18394399761692,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.004994608514587915,
      "loss": 0.7957,
      "num_input_tokens_seen": 59189944,
      "step": 101945
    },
    {
      "epoch": 15.184688710157879,
      "grad_norm": 0.041015625,
      "learning_rate": 0.004993156040745734,
      "loss": 0.7848,
      "num_input_tokens_seen": 59192952,
      "step": 101950
    },
    {
      "epoch": 15.185433422698837,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.004991703735959005,
      "loss": 0.7936,
      "num_input_tokens_seen": 59195672,
      "step": 101955
    },
    {
      "epoch": 15.186178135239798,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004990251600252275,
      "loss": 0.7939,
      "num_input_tokens_seen": 59198616,
      "step": 101960
    },
    {
      "epoch": 15.186922847780757,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.004988799633650067,
      "loss": 0.7987,
      "num_input_tokens_seen": 59201464,
      "step": 101965
    },
    {
      "epoch": 15.187667560321715,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.004987347836176916,
      "loss": 0.8161,
      "num_input_tokens_seen": 59204792,
      "step": 101970
    },
    {
      "epoch": 15.188412272862674,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.0049858962078573495,
      "loss": 0.7841,
      "num_input_tokens_seen": 59207864,
      "step": 101975
    },
    {
      "epoch": 15.189156985403635,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.004984444748715887,
      "loss": 0.7799,
      "num_input_tokens_seen": 59210680,
      "step": 101980
    },
    {
      "epoch": 15.189901697944594,
      "grad_norm": 0.076171875,
      "learning_rate": 0.004982993458777049,
      "loss": 0.8125,
      "num_input_tokens_seen": 59213464,
      "step": 101985
    },
    {
      "epoch": 15.190646410485552,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.004981542338065358,
      "loss": 0.7956,
      "num_input_tokens_seen": 59216376,
      "step": 101990
    },
    {
      "epoch": 15.191391123026511,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.004980091386605329,
      "loss": 0.8066,
      "num_input_tokens_seen": 59219320,
      "step": 101995
    },
    {
      "epoch": 15.192135835567472,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0049786406044214655,
      "loss": 0.8153,
      "num_input_tokens_seen": 59222072,
      "step": 102000
    },
    {
      "epoch": 15.19288054810843,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00497718999153829,
      "loss": 0.7938,
      "num_input_tokens_seen": 59225304,
      "step": 102005
    },
    {
      "epoch": 15.19362526064939,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.004975739547980296,
      "loss": 0.7716,
      "num_input_tokens_seen": 59228120,
      "step": 102010
    },
    {
      "epoch": 15.194369973190348,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.004974289273772001,
      "loss": 0.7836,
      "num_input_tokens_seen": 59230904,
      "step": 102015
    },
    {
      "epoch": 15.195114685731308,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.004972839168937898,
      "loss": 0.7939,
      "num_input_tokens_seen": 59233752,
      "step": 102020
    },
    {
      "epoch": 15.195859398272267,
      "grad_norm": 0.052734375,
      "learning_rate": 0.004971389233502481,
      "loss": 0.7984,
      "num_input_tokens_seen": 59236888,
      "step": 102025
    },
    {
      "epoch": 15.196604110813226,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.004969939467490255,
      "loss": 0.7977,
      "num_input_tokens_seen": 59239864,
      "step": 102030
    },
    {
      "epoch": 15.197348823354185,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.004968489870925704,
      "loss": 0.7899,
      "num_input_tokens_seen": 59242648,
      "step": 102035
    },
    {
      "epoch": 15.198093535895145,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.004967040443833327,
      "loss": 0.7891,
      "num_input_tokens_seen": 59245592,
      "step": 102040
    },
    {
      "epoch": 15.198838248436104,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.004965591186237602,
      "loss": 0.8115,
      "num_input_tokens_seen": 59248344,
      "step": 102045
    },
    {
      "epoch": 15.199582960977063,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.004964142098163015,
      "loss": 0.7946,
      "num_input_tokens_seen": 59251224,
      "step": 102050
    },
    {
      "epoch": 15.200327673518021,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.004962693179634049,
      "loss": 0.8123,
      "num_input_tokens_seen": 59253880,
      "step": 102055
    },
    {
      "epoch": 15.201072386058982,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.004961244430675175,
      "loss": 0.7914,
      "num_input_tokens_seen": 59256664,
      "step": 102060
    },
    {
      "epoch": 15.20181709859994,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.00495979585131088,
      "loss": 0.7983,
      "num_input_tokens_seen": 59259384,
      "step": 102065
    },
    {
      "epoch": 15.2025618111409,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.004958347441565621,
      "loss": 0.785,
      "num_input_tokens_seen": 59262392,
      "step": 102070
    },
    {
      "epoch": 15.203306523681858,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.004956899201463884,
      "loss": 0.8092,
      "num_input_tokens_seen": 59265432,
      "step": 102075
    },
    {
      "epoch": 15.204051236222819,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.004955451131030121,
      "loss": 0.7882,
      "num_input_tokens_seen": 59268184,
      "step": 102080
    },
    {
      "epoch": 15.204795948763778,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.004954003230288808,
      "loss": 0.8007,
      "num_input_tokens_seen": 59271000,
      "step": 102085
    },
    {
      "epoch": 15.205540661304736,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.004952555499264396,
      "loss": 0.8053,
      "num_input_tokens_seen": 59273720,
      "step": 102090
    },
    {
      "epoch": 15.206285373845695,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.004951107937981352,
      "loss": 0.7888,
      "num_input_tokens_seen": 59276856,
      "step": 102095
    },
    {
      "epoch": 15.207030086386654,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.004949660546464124,
      "loss": 0.8187,
      "num_input_tokens_seen": 59279736,
      "step": 102100
    },
    {
      "epoch": 15.207774798927614,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.004948213324737162,
      "loss": 0.7743,
      "num_input_tokens_seen": 59282648,
      "step": 102105
    },
    {
      "epoch": 15.208519511468573,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.004946766272824927,
      "loss": 0.7869,
      "num_input_tokens_seen": 59285592,
      "step": 102110
    },
    {
      "epoch": 15.209264224009532,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.004945319390751856,
      "loss": 0.7776,
      "num_input_tokens_seen": 59288664,
      "step": 102115
    },
    {
      "epoch": 15.21000893655049,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.004943872678542396,
      "loss": 0.8144,
      "num_input_tokens_seen": 59291512,
      "step": 102120
    },
    {
      "epoch": 15.210753649091451,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0049424261362209804,
      "loss": 0.7883,
      "num_input_tokens_seen": 59294232,
      "step": 102125
    },
    {
      "epoch": 15.21149836163241,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0049409797638120595,
      "loss": 0.8098,
      "num_input_tokens_seen": 59297208,
      "step": 102130
    },
    {
      "epoch": 15.212243074173369,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.004939533561340061,
      "loss": 0.7899,
      "num_input_tokens_seen": 59300024,
      "step": 102135
    },
    {
      "epoch": 15.212987786714327,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.004938087528829413,
      "loss": 0.8018,
      "num_input_tokens_seen": 59302712,
      "step": 102140
    },
    {
      "epoch": 15.213732499255288,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.004936641666304554,
      "loss": 0.7833,
      "num_input_tokens_seen": 59305496,
      "step": 102145
    },
    {
      "epoch": 15.214477211796247,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.004935195973789904,
      "loss": 0.7883,
      "num_input_tokens_seen": 59308472,
      "step": 102150
    },
    {
      "epoch": 15.215221924337206,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.00493375045130989,
      "loss": 0.7976,
      "num_input_tokens_seen": 59311544,
      "step": 102155
    },
    {
      "epoch": 15.215966636878164,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.004932305098888928,
      "loss": 0.8137,
      "num_input_tokens_seen": 59314648,
      "step": 102160
    },
    {
      "epoch": 15.216711349419125,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.004930859916551443,
      "loss": 0.808,
      "num_input_tokens_seen": 59317176,
      "step": 102165
    },
    {
      "epoch": 15.217456061960084,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.004929414904321842,
      "loss": 0.7784,
      "num_input_tokens_seen": 59320248,
      "step": 102170
    },
    {
      "epoch": 15.218200774501042,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.004927970062224546,
      "loss": 0.8104,
      "num_input_tokens_seen": 59323096,
      "step": 102175
    },
    {
      "epoch": 15.218945487042001,
      "grad_norm": 0.083984375,
      "learning_rate": 0.004926525390283959,
      "loss": 0.7986,
      "num_input_tokens_seen": 59326232,
      "step": 102180
    },
    {
      "epoch": 15.219690199582962,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.004925080888524488,
      "loss": 0.7992,
      "num_input_tokens_seen": 59329240,
      "step": 102185
    },
    {
      "epoch": 15.22043491212392,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004923636556970533,
      "loss": 0.8026,
      "num_input_tokens_seen": 59332056,
      "step": 102190
    },
    {
      "epoch": 15.22117962466488,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.004922192395646495,
      "loss": 0.797,
      "num_input_tokens_seen": 59335096,
      "step": 102195
    },
    {
      "epoch": 15.221924337205838,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.004920748404576777,
      "loss": 0.8064,
      "num_input_tokens_seen": 59337976,
      "step": 102200
    },
    {
      "epoch": 15.222669049746798,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.004919304583785767,
      "loss": 0.7931,
      "num_input_tokens_seen": 59340696,
      "step": 102205
    },
    {
      "epoch": 15.223413762287757,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.004917860933297864,
      "loss": 0.8238,
      "num_input_tokens_seen": 59343416,
      "step": 102210
    },
    {
      "epoch": 15.224158474828716,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.004916417453137454,
      "loss": 0.7926,
      "num_input_tokens_seen": 59346168,
      "step": 102215
    },
    {
      "epoch": 15.224903187369675,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.004914974143328919,
      "loss": 0.795,
      "num_input_tokens_seen": 59348696,
      "step": 102220
    },
    {
      "epoch": 15.225647899910635,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00491353100389665,
      "loss": 0.8008,
      "num_input_tokens_seen": 59351736,
      "step": 102225
    },
    {
      "epoch": 15.226392612451594,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.004912088034865018,
      "loss": 0.7868,
      "num_input_tokens_seen": 59354680,
      "step": 102230
    },
    {
      "epoch": 15.227137324992553,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.004910645236258412,
      "loss": 0.7942,
      "num_input_tokens_seen": 59357720,
      "step": 102235
    },
    {
      "epoch": 15.227882037533512,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.004909202608101193,
      "loss": 0.7922,
      "num_input_tokens_seen": 59360376,
      "step": 102240
    },
    {
      "epoch": 15.228626750074472,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0049077601504177466,
      "loss": 0.8074,
      "num_input_tokens_seen": 59363256,
      "step": 102245
    },
    {
      "epoch": 15.22937146261543,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.004906317863232435,
      "loss": 0.8027,
      "num_input_tokens_seen": 59366200,
      "step": 102250
    },
    {
      "epoch": 15.23011617515639,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.004904875746569624,
      "loss": 0.7957,
      "num_input_tokens_seen": 59368824,
      "step": 102255
    },
    {
      "epoch": 15.230860887697348,
      "grad_norm": 0.07421875,
      "learning_rate": 0.004903433800453678,
      "loss": 0.805,
      "num_input_tokens_seen": 59371704,
      "step": 102260
    },
    {
      "epoch": 15.231605600238307,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.00490199202490895,
      "loss": 0.7969,
      "num_input_tokens_seen": 59374616,
      "step": 102265
    },
    {
      "epoch": 15.232350312779268,
      "grad_norm": 0.041015625,
      "learning_rate": 0.004900550419959808,
      "loss": 0.8139,
      "num_input_tokens_seen": 59377432,
      "step": 102270
    },
    {
      "epoch": 15.233095025320226,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0048991089856306,
      "loss": 0.7956,
      "num_input_tokens_seen": 59380312,
      "step": 102275
    },
    {
      "epoch": 15.233839737861185,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.004897667721945682,
      "loss": 0.8298,
      "num_input_tokens_seen": 59383448,
      "step": 102280
    },
    {
      "epoch": 15.234584450402144,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0048962266289293965,
      "loss": 0.7989,
      "num_input_tokens_seen": 59386360,
      "step": 102285
    },
    {
      "epoch": 15.235329162943104,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.004894785706606097,
      "loss": 0.7936,
      "num_input_tokens_seen": 59389208,
      "step": 102290
    },
    {
      "epoch": 15.236073875484063,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.004893344955000123,
      "loss": 0.8033,
      "num_input_tokens_seen": 59392184,
      "step": 102295
    },
    {
      "epoch": 15.236818588025022,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.00489190437413581,
      "loss": 0.7913,
      "num_input_tokens_seen": 59395352,
      "step": 102300
    },
    {
      "epoch": 15.23756330056598,
      "grad_norm": 0.0390625,
      "learning_rate": 0.004890463964037504,
      "loss": 0.8022,
      "num_input_tokens_seen": 59398232,
      "step": 102305
    },
    {
      "epoch": 15.238308013106941,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.0048890237247295335,
      "loss": 0.7979,
      "num_input_tokens_seen": 59401016,
      "step": 102310
    },
    {
      "epoch": 15.2390527256479,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.004887583656236226,
      "loss": 0.7763,
      "num_input_tokens_seen": 59403800,
      "step": 102315
    },
    {
      "epoch": 15.239797438188859,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.00488614375858192,
      "loss": 0.8144,
      "num_input_tokens_seen": 59406456,
      "step": 102320
    },
    {
      "epoch": 15.240542150729818,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0048847040317909355,
      "loss": 0.7983,
      "num_input_tokens_seen": 59409880,
      "step": 102325
    },
    {
      "epoch": 15.241286863270778,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.0048832644758875915,
      "loss": 0.8068,
      "num_input_tokens_seen": 59412632,
      "step": 102330
    },
    {
      "epoch": 15.242031575811737,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.004881825090896216,
      "loss": 0.8042,
      "num_input_tokens_seen": 59415192,
      "step": 102335
    },
    {
      "epoch": 15.242776288352696,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.004880385876841122,
      "loss": 0.8124,
      "num_input_tokens_seen": 59417784,
      "step": 102340
    },
    {
      "epoch": 15.243521000893654,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.004878946833746619,
      "loss": 0.7952,
      "num_input_tokens_seen": 59420728,
      "step": 102345
    },
    {
      "epoch": 15.244265713434615,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.004877507961637028,
      "loss": 0.8101,
      "num_input_tokens_seen": 59423448,
      "step": 102350
    },
    {
      "epoch": 15.245010425975574,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.004876069260536646,
      "loss": 0.7867,
      "num_input_tokens_seen": 59426392,
      "step": 102355
    },
    {
      "epoch": 15.245755138516532,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.00487463073046979,
      "loss": 0.7874,
      "num_input_tokens_seen": 59429336,
      "step": 102360
    },
    {
      "epoch": 15.246499851057491,
      "grad_norm": 0.044921875,
      "learning_rate": 0.004873192371460752,
      "loss": 0.7988,
      "num_input_tokens_seen": 59432472,
      "step": 102365
    },
    {
      "epoch": 15.247244563598452,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0048717541835338425,
      "loss": 0.8135,
      "num_input_tokens_seen": 59435352,
      "step": 102370
    },
    {
      "epoch": 15.24798927613941,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.00487031616671335,
      "loss": 0.7808,
      "num_input_tokens_seen": 59438072,
      "step": 102375
    },
    {
      "epoch": 15.24873398868037,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0048688783210235715,
      "loss": 0.8021,
      "num_input_tokens_seen": 59441432,
      "step": 102380
    },
    {
      "epoch": 15.249478701221328,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.004867440646488797,
      "loss": 0.8138,
      "num_input_tokens_seen": 59444152,
      "step": 102385
    },
    {
      "epoch": 15.250223413762289,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.004866003143133311,
      "loss": 0.8152,
      "num_input_tokens_seen": 59447032,
      "step": 102390
    },
    {
      "epoch": 15.250968126303247,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.004864565810981405,
      "loss": 0.8058,
      "num_input_tokens_seen": 59449624,
      "step": 102395
    },
    {
      "epoch": 15.251712838844206,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.004863128650057353,
      "loss": 0.7985,
      "num_input_tokens_seen": 59452760,
      "step": 102400
    },
    {
      "epoch": 15.252457551385165,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.004861691660385446,
      "loss": 0.8062,
      "num_input_tokens_seen": 59456152,
      "step": 102405
    },
    {
      "epoch": 15.253202263926125,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.00486025484198995,
      "loss": 0.7917,
      "num_input_tokens_seen": 59459000,
      "step": 102410
    },
    {
      "epoch": 15.253946976467084,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.004858818194895147,
      "loss": 0.788,
      "num_input_tokens_seen": 59461656,
      "step": 102415
    },
    {
      "epoch": 15.254691689008043,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.004857381719125303,
      "loss": 0.8121,
      "num_input_tokens_seen": 59464408,
      "step": 102420
    },
    {
      "epoch": 15.255436401549002,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.004855945414704681,
      "loss": 0.7898,
      "num_input_tokens_seen": 59467288,
      "step": 102425
    },
    {
      "epoch": 15.256181114089962,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.004854509281657557,
      "loss": 0.7973,
      "num_input_tokens_seen": 59470136,
      "step": 102430
    },
    {
      "epoch": 15.256925826630921,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.004853073320008183,
      "loss": 0.7963,
      "num_input_tokens_seen": 59472568,
      "step": 102435
    },
    {
      "epoch": 15.25767053917188,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0048516375297808275,
      "loss": 0.7962,
      "num_input_tokens_seen": 59475576,
      "step": 102440
    },
    {
      "epoch": 15.258415251712838,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.00485020191099974,
      "loss": 0.7896,
      "num_input_tokens_seen": 59478232,
      "step": 102445
    },
    {
      "epoch": 15.259159964253797,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.004848766463689176,
      "loss": 0.8054,
      "num_input_tokens_seen": 59481336,
      "step": 102450
    },
    {
      "epoch": 15.259904676794758,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.004847331187873386,
      "loss": 0.7867,
      "num_input_tokens_seen": 59484216,
      "step": 102455
    },
    {
      "epoch": 15.260649389335716,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0048458960835766115,
      "loss": 0.7986,
      "num_input_tokens_seen": 59487384,
      "step": 102460
    },
    {
      "epoch": 15.261394101876675,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.004844461150823107,
      "loss": 0.8065,
      "num_input_tokens_seen": 59490328,
      "step": 102465
    },
    {
      "epoch": 15.262138814417634,
      "grad_norm": 0.04296875,
      "learning_rate": 0.004843026389637107,
      "loss": 0.799,
      "num_input_tokens_seen": 59493304,
      "step": 102470
    },
    {
      "epoch": 15.262883526958595,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.004841591800042855,
      "loss": 0.7973,
      "num_input_tokens_seen": 59496184,
      "step": 102475
    },
    {
      "epoch": 15.263628239499553,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.004840157382064582,
      "loss": 0.7796,
      "num_input_tokens_seen": 59499128,
      "step": 102480
    },
    {
      "epoch": 15.264372952040512,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.004838723135726528,
      "loss": 0.7998,
      "num_input_tokens_seen": 59502104,
      "step": 102485
    },
    {
      "epoch": 15.26511766458147,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.004837289061052913,
      "loss": 0.8033,
      "num_input_tokens_seen": 59504824,
      "step": 102490
    },
    {
      "epoch": 15.265862377122431,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.004835855158067979,
      "loss": 0.781,
      "num_input_tokens_seen": 59507896,
      "step": 102495
    },
    {
      "epoch": 15.26660708966339,
      "grad_norm": 0.07421875,
      "learning_rate": 0.004834421426795939,
      "loss": 0.8045,
      "num_input_tokens_seen": 59511032,
      "step": 102500
    },
    {
      "epoch": 15.267351802204349,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.004832987867261012,
      "loss": 0.7913,
      "num_input_tokens_seen": 59513688,
      "step": 102505
    },
    {
      "epoch": 15.268096514745308,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.004831554479487429,
      "loss": 0.8,
      "num_input_tokens_seen": 59516824,
      "step": 102510
    },
    {
      "epoch": 15.268841227286268,
      "grad_norm": 0.037109375,
      "learning_rate": 0.004830121263499396,
      "loss": 0.8067,
      "num_input_tokens_seen": 59519608,
      "step": 102515
    },
    {
      "epoch": 15.269585939827227,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.004828688219321128,
      "loss": 0.7914,
      "num_input_tokens_seen": 59522520,
      "step": 102520
    },
    {
      "epoch": 15.270330652368186,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.004827255346976831,
      "loss": 0.8016,
      "num_input_tokens_seen": 59525848,
      "step": 102525
    },
    {
      "epoch": 15.271075364909144,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.004825822646490721,
      "loss": 0.7884,
      "num_input_tokens_seen": 59529016,
      "step": 102530
    },
    {
      "epoch": 15.271820077450105,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.004824390117886991,
      "loss": 0.8116,
      "num_input_tokens_seen": 59531640,
      "step": 102535
    },
    {
      "epoch": 15.272564789991064,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0048229577611898534,
      "loss": 0.7864,
      "num_input_tokens_seen": 59534360,
      "step": 102540
    },
    {
      "epoch": 15.273309502532022,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.004821525576423501,
      "loss": 0.7884,
      "num_input_tokens_seen": 59537016,
      "step": 102545
    },
    {
      "epoch": 15.274054215072981,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.004820093563612123,
      "loss": 0.8113,
      "num_input_tokens_seen": 59539672,
      "step": 102550
    },
    {
      "epoch": 15.274798927613942,
      "grad_norm": 0.0546875,
      "learning_rate": 0.004818661722779925,
      "loss": 0.7858,
      "num_input_tokens_seen": 59542552,
      "step": 102555
    },
    {
      "epoch": 15.2755436401549,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.004817230053951083,
      "loss": 0.8057,
      "num_input_tokens_seen": 59545688,
      "step": 102560
    },
    {
      "epoch": 15.27628835269586,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.004815798557149795,
      "loss": 0.8032,
      "num_input_tokens_seen": 59548376,
      "step": 102565
    },
    {
      "epoch": 15.277033065236818,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0048143672324002336,
      "loss": 0.7894,
      "num_input_tokens_seen": 59551576,
      "step": 102570
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.004812936079726592,
      "loss": 0.7898,
      "num_input_tokens_seen": 59554552,
      "step": 102575
    },
    {
      "epoch": 15.278522490318737,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.004811505099153041,
      "loss": 0.7725,
      "num_input_tokens_seen": 59557368,
      "step": 102580
    },
    {
      "epoch": 15.279267202859696,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.004810074290703756,
      "loss": 0.7935,
      "num_input_tokens_seen": 59560120,
      "step": 102585
    },
    {
      "epoch": 15.280011915400655,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.00480864365440291,
      "loss": 0.8044,
      "num_input_tokens_seen": 59563032,
      "step": 102590
    },
    {
      "epoch": 15.280756627941615,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.004807213190274667,
      "loss": 0.7973,
      "num_input_tokens_seen": 59565976,
      "step": 102595
    },
    {
      "epoch": 15.281501340482574,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.004805782898343202,
      "loss": 0.8017,
      "num_input_tokens_seen": 59568760,
      "step": 102600
    },
    {
      "epoch": 15.282246053023533,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.004804352778632668,
      "loss": 0.8114,
      "num_input_tokens_seen": 59571672,
      "step": 102605
    },
    {
      "epoch": 15.282990765564492,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.004802922831167239,
      "loss": 0.7965,
      "num_input_tokens_seen": 59574488,
      "step": 102610
    },
    {
      "epoch": 15.283735478105452,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.004801493055971057,
      "loss": 0.8076,
      "num_input_tokens_seen": 59577176,
      "step": 102615
    },
    {
      "epoch": 15.284480190646411,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.004800063453068291,
      "loss": 0.8031,
      "num_input_tokens_seen": 59580472,
      "step": 102620
    },
    {
      "epoch": 15.28522490318737,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.004798634022483086,
      "loss": 0.7995,
      "num_input_tokens_seen": 59583192,
      "step": 102625
    },
    {
      "epoch": 15.285969615728328,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.004797204764239585,
      "loss": 0.7984,
      "num_input_tokens_seen": 59586232,
      "step": 102630
    },
    {
      "epoch": 15.286714328269287,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.004795775678361946,
      "loss": 0.8036,
      "num_input_tokens_seen": 59589144,
      "step": 102635
    },
    {
      "epoch": 15.287459040810248,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.004794346764874305,
      "loss": 0.8115,
      "num_input_tokens_seen": 59592056,
      "step": 102640
    },
    {
      "epoch": 15.288203753351207,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.004792918023800799,
      "loss": 0.7914,
      "num_input_tokens_seen": 59594936,
      "step": 102645
    },
    {
      "epoch": 15.288948465892165,
      "grad_norm": 0.05078125,
      "learning_rate": 0.004791489455165573,
      "loss": 0.7895,
      "num_input_tokens_seen": 59597560,
      "step": 102650
    },
    {
      "epoch": 15.289693178433124,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.004790061058992757,
      "loss": 0.8049,
      "num_input_tokens_seen": 59600344,
      "step": 102655
    },
    {
      "epoch": 15.290437890974085,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.004788632835306483,
      "loss": 0.8033,
      "num_input_tokens_seen": 59604216,
      "step": 102660
    },
    {
      "epoch": 15.291182603515043,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.004787204784130875,
      "loss": 0.779,
      "num_input_tokens_seen": 59607032,
      "step": 102665
    },
    {
      "epoch": 15.291927316056002,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.004785776905490067,
      "loss": 0.7892,
      "num_input_tokens_seen": 59609912,
      "step": 102670
    },
    {
      "epoch": 15.29267202859696,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00478434919940817,
      "loss": 0.8219,
      "num_input_tokens_seen": 59613016,
      "step": 102675
    },
    {
      "epoch": 15.293416741137921,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.004782921665909317,
      "loss": 0.8005,
      "num_input_tokens_seen": 59616184,
      "step": 102680
    },
    {
      "epoch": 15.29416145367888,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.004781494305017612,
      "loss": 0.7827,
      "num_input_tokens_seen": 59618968,
      "step": 102685
    },
    {
      "epoch": 15.294906166219839,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.004780067116757182,
      "loss": 0.7945,
      "num_input_tokens_seen": 59621784,
      "step": 102690
    },
    {
      "epoch": 15.295650878760798,
      "grad_norm": 0.060546875,
      "learning_rate": 0.004778640101152125,
      "loss": 0.8092,
      "num_input_tokens_seen": 59624824,
      "step": 102695
    },
    {
      "epoch": 15.296395591301758,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.00477721325822656,
      "loss": 0.7854,
      "num_input_tokens_seen": 59627480,
      "step": 102700
    },
    {
      "epoch": 15.297140303842717,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.004775786588004586,
      "loss": 0.7856,
      "num_input_tokens_seen": 59630552,
      "step": 102705
    },
    {
      "epoch": 15.297885016383676,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0047743600905103085,
      "loss": 0.7971,
      "num_input_tokens_seen": 59633304,
      "step": 102710
    },
    {
      "epoch": 15.298629728924634,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.004772933765767823,
      "loss": 0.7923,
      "num_input_tokens_seen": 59636120,
      "step": 102715
    },
    {
      "epoch": 15.299374441465595,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.004771507613801222,
      "loss": 0.7794,
      "num_input_tokens_seen": 59639000,
      "step": 102720
    },
    {
      "epoch": 15.300119154006554,
      "grad_norm": 0.078125,
      "learning_rate": 0.004770081634634609,
      "loss": 0.8033,
      "num_input_tokens_seen": 59641912,
      "step": 102725
    },
    {
      "epoch": 15.300863866547513,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.004768655828292066,
      "loss": 0.7962,
      "num_input_tokens_seen": 59645048,
      "step": 102730
    },
    {
      "epoch": 15.301608579088471,
      "grad_norm": 0.037109375,
      "learning_rate": 0.004767230194797688,
      "loss": 0.7851,
      "num_input_tokens_seen": 59647896,
      "step": 102735
    },
    {
      "epoch": 15.302353291629432,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.004765804734175555,
      "loss": 0.7904,
      "num_input_tokens_seen": 59650840,
      "step": 102740
    },
    {
      "epoch": 15.30309800417039,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.004764379446449745,
      "loss": 0.8132,
      "num_input_tokens_seen": 59654040,
      "step": 102745
    },
    {
      "epoch": 15.30384271671135,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.004762954331644348,
      "loss": 0.8055,
      "num_input_tokens_seen": 59657240,
      "step": 102750
    },
    {
      "epoch": 15.304587429252308,
      "grad_norm": 0.05078125,
      "learning_rate": 0.004761529389783426,
      "loss": 0.7959,
      "num_input_tokens_seen": 59660248,
      "step": 102755
    },
    {
      "epoch": 15.305332141793269,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.004760104620891064,
      "loss": 0.8073,
      "num_input_tokens_seen": 59663064,
      "step": 102760
    },
    {
      "epoch": 15.306076854334227,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.004758680024991323,
      "loss": 0.8042,
      "num_input_tokens_seen": 59665816,
      "step": 102765
    },
    {
      "epoch": 15.306821566875186,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.004757255602108279,
      "loss": 0.7978,
      "num_input_tokens_seen": 59668792,
      "step": 102770
    },
    {
      "epoch": 15.307566279416145,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.004755831352265991,
      "loss": 0.808,
      "num_input_tokens_seen": 59671768,
      "step": 102775
    },
    {
      "epoch": 15.308310991957104,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.004754407275488522,
      "loss": 0.771,
      "num_input_tokens_seen": 59674616,
      "step": 102780
    },
    {
      "epoch": 15.309055704498064,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.004752983371799927,
      "loss": 0.7839,
      "num_input_tokens_seen": 59677528,
      "step": 102785
    },
    {
      "epoch": 15.309800417039023,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.004751559641224258,
      "loss": 0.7946,
      "num_input_tokens_seen": 59680760,
      "step": 102790
    },
    {
      "epoch": 15.310545129579982,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0047501360837855795,
      "loss": 0.7967,
      "num_input_tokens_seen": 59683768,
      "step": 102795
    },
    {
      "epoch": 15.31128984212094,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.004748712699507929,
      "loss": 0.7994,
      "num_input_tokens_seen": 59686744,
      "step": 102800
    },
    {
      "epoch": 15.312034554661901,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.004747289488415363,
      "loss": 0.7835,
      "num_input_tokens_seen": 59689816,
      "step": 102805
    },
    {
      "epoch": 15.31277926720286,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.004745866450531916,
      "loss": 0.7891,
      "num_input_tokens_seen": 59692600,
      "step": 102810
    },
    {
      "epoch": 15.313523979743819,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0047444435858816384,
      "loss": 0.793,
      "num_input_tokens_seen": 59695480,
      "step": 102815
    },
    {
      "epoch": 15.314268692284777,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.004743020894488563,
      "loss": 0.7736,
      "num_input_tokens_seen": 59698360,
      "step": 102820
    },
    {
      "epoch": 15.315013404825738,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00474159837637672,
      "loss": 0.796,
      "num_input_tokens_seen": 59701368,
      "step": 102825
    },
    {
      "epoch": 15.315758117366697,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0047401760315701515,
      "loss": 0.7978,
      "num_input_tokens_seen": 59703992,
      "step": 102830
    },
    {
      "epoch": 15.316502829907655,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.004738753860092876,
      "loss": 0.809,
      "num_input_tokens_seen": 59706968,
      "step": 102835
    },
    {
      "epoch": 15.317247542448614,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.004737331861968931,
      "loss": 0.8057,
      "num_input_tokens_seen": 59710008,
      "step": 102840
    },
    {
      "epoch": 15.317992254989575,
      "grad_norm": 0.08203125,
      "learning_rate": 0.004735910037222332,
      "loss": 0.7998,
      "num_input_tokens_seen": 59712728,
      "step": 102845
    },
    {
      "epoch": 15.318736967530533,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.004734488385877101,
      "loss": 0.8084,
      "num_input_tokens_seen": 59715864,
      "step": 102850
    },
    {
      "epoch": 15.319481680071492,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.004733066907957253,
      "loss": 0.8235,
      "num_input_tokens_seen": 59718616,
      "step": 102855
    },
    {
      "epoch": 15.320226392612451,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.004731645603486808,
      "loss": 0.863,
      "num_input_tokens_seen": 59721400,
      "step": 102860
    },
    {
      "epoch": 15.320971105153411,
      "grad_norm": 0.041015625,
      "learning_rate": 0.004730224472489773,
      "loss": 0.8138,
      "num_input_tokens_seen": 59724440,
      "step": 102865
    },
    {
      "epoch": 15.32171581769437,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.004728803514990153,
      "loss": 0.7985,
      "num_input_tokens_seen": 59727352,
      "step": 102870
    },
    {
      "epoch": 15.322460530235329,
      "grad_norm": 0.044921875,
      "learning_rate": 0.004727382731011965,
      "loss": 0.8026,
      "num_input_tokens_seen": 59730200,
      "step": 102875
    },
    {
      "epoch": 15.323205242776288,
      "grad_norm": 0.0390625,
      "learning_rate": 0.004725962120579199,
      "loss": 0.7794,
      "num_input_tokens_seen": 59732824,
      "step": 102880
    },
    {
      "epoch": 15.323949955317248,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0047245416837158665,
      "loss": 0.7988,
      "num_input_tokens_seen": 59735864,
      "step": 102885
    },
    {
      "epoch": 15.324694667858207,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0047231214204459535,
      "loss": 0.7983,
      "num_input_tokens_seen": 59739192,
      "step": 102890
    },
    {
      "epoch": 15.325439380399166,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.004721701330793465,
      "loss": 0.8188,
      "num_input_tokens_seen": 59742168,
      "step": 102895
    },
    {
      "epoch": 15.326184092940125,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.004720281414782385,
      "loss": 0.797,
      "num_input_tokens_seen": 59745432,
      "step": 102900
    },
    {
      "epoch": 15.326928805481085,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.004718861672436698,
      "loss": 0.8158,
      "num_input_tokens_seen": 59748760,
      "step": 102905
    },
    {
      "epoch": 15.327673518022044,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.004717442103780398,
      "loss": 0.8152,
      "num_input_tokens_seen": 59751544,
      "step": 102910
    },
    {
      "epoch": 15.328418230563003,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.004716022708837464,
      "loss": 0.8145,
      "num_input_tokens_seen": 59754392,
      "step": 102915
    },
    {
      "epoch": 15.329162943103961,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0047146034876318745,
      "loss": 0.8043,
      "num_input_tokens_seen": 59757208,
      "step": 102920
    },
    {
      "epoch": 15.329907655644922,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0047131844401875995,
      "loss": 0.7735,
      "num_input_tokens_seen": 59760472,
      "step": 102925
    },
    {
      "epoch": 15.33065236818588,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.004711765566528623,
      "loss": 0.811,
      "num_input_tokens_seen": 59763288,
      "step": 102930
    },
    {
      "epoch": 15.33139708072684,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.004710346866678907,
      "loss": 0.7958,
      "num_input_tokens_seen": 59766168,
      "step": 102935
    },
    {
      "epoch": 15.332141793267798,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.004708928340662427,
      "loss": 0.8055,
      "num_input_tokens_seen": 59769112,
      "step": 102940
    },
    {
      "epoch": 15.332886505808759,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.004707509988503144,
      "loss": 0.7891,
      "num_input_tokens_seen": 59772568,
      "step": 102945
    },
    {
      "epoch": 15.333631218349717,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004706091810225014,
      "loss": 0.8063,
      "num_input_tokens_seen": 59775384,
      "step": 102950
    },
    {
      "epoch": 15.334375930890676,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0047046738058520035,
      "loss": 0.812,
      "num_input_tokens_seen": 59778488,
      "step": 102955
    },
    {
      "epoch": 15.335120643431635,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.004703255975408063,
      "loss": 0.7838,
      "num_input_tokens_seen": 59781272,
      "step": 102960
    },
    {
      "epoch": 15.335865355972594,
      "grad_norm": 0.037109375,
      "learning_rate": 0.004701838318917151,
      "loss": 0.7981,
      "num_input_tokens_seen": 59784504,
      "step": 102965
    },
    {
      "epoch": 15.336610068513554,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0047004208364032135,
      "loss": 0.8067,
      "num_input_tokens_seen": 59787448,
      "step": 102970
    },
    {
      "epoch": 15.337354781054513,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.004699003527890195,
      "loss": 0.7877,
      "num_input_tokens_seen": 59790296,
      "step": 102975
    },
    {
      "epoch": 15.338099493595472,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.004697586393402046,
      "loss": 0.7947,
      "num_input_tokens_seen": 59792984,
      "step": 102980
    },
    {
      "epoch": 15.33884420613643,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004696169432962704,
      "loss": 0.7983,
      "num_input_tokens_seen": 59796120,
      "step": 102985
    },
    {
      "epoch": 15.339588918677391,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.004694752646596108,
      "loss": 0.8152,
      "num_input_tokens_seen": 59799352,
      "step": 102990
    },
    {
      "epoch": 15.34033363121835,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.004693336034326186,
      "loss": 0.8177,
      "num_input_tokens_seen": 59802264,
      "step": 102995
    },
    {
      "epoch": 15.341078343759309,
      "grad_norm": 0.041015625,
      "learning_rate": 0.004691919596176883,
      "loss": 0.8023,
      "num_input_tokens_seen": 59805112,
      "step": 103000
    },
    {
      "epoch": 15.341823056300267,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.004690503332172116,
      "loss": 0.7973,
      "num_input_tokens_seen": 59807832,
      "step": 103005
    },
    {
      "epoch": 15.342567768841228,
      "grad_norm": 0.0625,
      "learning_rate": 0.004689087242335822,
      "loss": 0.8048,
      "num_input_tokens_seen": 59810712,
      "step": 103010
    },
    {
      "epoch": 15.343312481382187,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.004687671326691915,
      "loss": 0.7979,
      "num_input_tokens_seen": 59813432,
      "step": 103015
    },
    {
      "epoch": 15.344057193923145,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.004686255585264324,
      "loss": 0.7871,
      "num_input_tokens_seen": 59816280,
      "step": 103020
    },
    {
      "epoch": 15.344801906464104,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004684840018076963,
      "loss": 0.8015,
      "num_input_tokens_seen": 59819160,
      "step": 103025
    },
    {
      "epoch": 15.345546619005065,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.004683424625153741,
      "loss": 0.8014,
      "num_input_tokens_seen": 59822104,
      "step": 103030
    },
    {
      "epoch": 15.346291331546023,
      "grad_norm": 0.072265625,
      "learning_rate": 0.004682009406518579,
      "loss": 0.7912,
      "num_input_tokens_seen": 59825080,
      "step": 103035
    },
    {
      "epoch": 15.347036044086982,
      "grad_norm": 0.072265625,
      "learning_rate": 0.004680594362195381,
      "loss": 0.8019,
      "num_input_tokens_seen": 59827960,
      "step": 103040
    },
    {
      "epoch": 15.347780756627941,
      "grad_norm": 0.068359375,
      "learning_rate": 0.004679179492208056,
      "loss": 0.8041,
      "num_input_tokens_seen": 59831224,
      "step": 103045
    },
    {
      "epoch": 15.348525469168901,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.004677764796580497,
      "loss": 0.7906,
      "num_input_tokens_seen": 59834008,
      "step": 103050
    },
    {
      "epoch": 15.34927018170986,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.004676350275336616,
      "loss": 0.7917,
      "num_input_tokens_seen": 59836760,
      "step": 103055
    },
    {
      "epoch": 15.350014894250819,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004674935928500303,
      "loss": 0.8027,
      "num_input_tokens_seen": 59839448,
      "step": 103060
    },
    {
      "epoch": 15.350759606791778,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.00467352175609545,
      "loss": 0.8054,
      "num_input_tokens_seen": 59842424,
      "step": 103065
    },
    {
      "epoch": 15.351504319332738,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004672107758145957,
      "loss": 0.798,
      "num_input_tokens_seen": 59845304,
      "step": 103070
    },
    {
      "epoch": 15.352249031873697,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.004670693934675701,
      "loss": 0.8103,
      "num_input_tokens_seen": 59848344,
      "step": 103075
    },
    {
      "epoch": 15.352993744414656,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.004669280285708578,
      "loss": 0.8399,
      "num_input_tokens_seen": 59851224,
      "step": 103080
    },
    {
      "epoch": 15.353738456955615,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.004667866811268458,
      "loss": 0.8015,
      "num_input_tokens_seen": 59854264,
      "step": 103085
    },
    {
      "epoch": 15.354483169496575,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.004666453511379234,
      "loss": 0.8175,
      "num_input_tokens_seen": 59857240,
      "step": 103090
    },
    {
      "epoch": 15.355227882037534,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.004665040386064772,
      "loss": 0.8019,
      "num_input_tokens_seen": 59860504,
      "step": 103095
    },
    {
      "epoch": 15.355972594578493,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.0046636274353489505,
      "loss": 0.795,
      "num_input_tokens_seen": 59863480,
      "step": 103100
    },
    {
      "epoch": 15.356717307119451,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00466221465925564,
      "loss": 0.8004,
      "num_input_tokens_seen": 59866456,
      "step": 103105
    },
    {
      "epoch": 15.357462019660412,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.004660802057808705,
      "loss": 0.785,
      "num_input_tokens_seen": 59868856,
      "step": 103110
    },
    {
      "epoch": 15.35820673220137,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.004659389631032011,
      "loss": 0.8015,
      "num_input_tokens_seen": 59871864,
      "step": 103115
    },
    {
      "epoch": 15.35895144474233,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.004657977378949415,
      "loss": 0.7875,
      "num_input_tokens_seen": 59874712,
      "step": 103120
    },
    {
      "epoch": 15.359696157283288,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.004656565301584785,
      "loss": 0.8,
      "num_input_tokens_seen": 59877528,
      "step": 103125
    },
    {
      "epoch": 15.360440869824249,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.004655153398961966,
      "loss": 0.8046,
      "num_input_tokens_seen": 59880888,
      "step": 103130
    },
    {
      "epoch": 15.361185582365207,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.004653741671104821,
      "loss": 0.7792,
      "num_input_tokens_seen": 59883672,
      "step": 103135
    },
    {
      "epoch": 15.361930294906166,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.004652330118037192,
      "loss": 0.7938,
      "num_input_tokens_seen": 59886680,
      "step": 103140
    },
    {
      "epoch": 15.362675007447125,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.004650918739782932,
      "loss": 0.8031,
      "num_input_tokens_seen": 59889912,
      "step": 103145
    },
    {
      "epoch": 15.363419719988084,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.004649507536365882,
      "loss": 0.7882,
      "num_input_tokens_seen": 59892952,
      "step": 103150
    },
    {
      "epoch": 15.364164432529044,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0046480965078098775,
      "loss": 0.7819,
      "num_input_tokens_seen": 59896056,
      "step": 103155
    },
    {
      "epoch": 15.364909145070003,
      "grad_norm": 0.02294921875,
      "learning_rate": 0.004646685654138765,
      "loss": 0.7869,
      "num_input_tokens_seen": 59898616,
      "step": 103160
    },
    {
      "epoch": 15.365653857610962,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.004645274975376373,
      "loss": 0.8044,
      "num_input_tokens_seen": 59901912,
      "step": 103165
    },
    {
      "epoch": 15.36639857015192,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.00464386447154654,
      "loss": 0.7915,
      "num_input_tokens_seen": 59904952,
      "step": 103170
    },
    {
      "epoch": 15.367143282692881,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.004642454142673091,
      "loss": 0.7922,
      "num_input_tokens_seen": 59907704,
      "step": 103175
    },
    {
      "epoch": 15.36788799523384,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.004641043988779853,
      "loss": 0.8119,
      "num_input_tokens_seen": 59910328,
      "step": 103180
    },
    {
      "epoch": 15.368632707774799,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0046396340098906474,
      "loss": 0.8096,
      "num_input_tokens_seen": 59913656,
      "step": 103185
    },
    {
      "epoch": 15.369377420315757,
      "grad_norm": 0.037109375,
      "learning_rate": 0.004638224206029292,
      "loss": 0.8192,
      "num_input_tokens_seen": 59916600,
      "step": 103190
    },
    {
      "epoch": 15.370122132856718,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.00463681457721961,
      "loss": 0.7994,
      "num_input_tokens_seen": 59919416,
      "step": 103195
    },
    {
      "epoch": 15.370866845397677,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.004635405123485411,
      "loss": 0.7895,
      "num_input_tokens_seen": 59922424,
      "step": 103200
    },
    {
      "epoch": 15.371611557938635,
      "grad_norm": 0.033203125,
      "learning_rate": 0.004633995844850511,
      "loss": 0.7871,
      "num_input_tokens_seen": 59925496,
      "step": 103205
    },
    {
      "epoch": 15.372356270479594,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.004632586741338713,
      "loss": 0.796,
      "num_input_tokens_seen": 59928152,
      "step": 103210
    },
    {
      "epoch": 15.373100983020555,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.004631177812973827,
      "loss": 0.7867,
      "num_input_tokens_seen": 59931064,
      "step": 103215
    },
    {
      "epoch": 15.373845695561513,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.004629769059779651,
      "loss": 0.7751,
      "num_input_tokens_seen": 59934040,
      "step": 103220
    },
    {
      "epoch": 15.374590408102472,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.004628360481779991,
      "loss": 0.8081,
      "num_input_tokens_seen": 59936600,
      "step": 103225
    },
    {
      "epoch": 15.375335120643431,
      "grad_norm": 0.048828125,
      "learning_rate": 0.004626952078998639,
      "loss": 0.7992,
      "num_input_tokens_seen": 59939320,
      "step": 103230
    },
    {
      "epoch": 15.376079833184392,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.004625543851459385,
      "loss": 0.7956,
      "num_input_tokens_seen": 59942232,
      "step": 103235
    },
    {
      "epoch": 15.37682454572535,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.004624135799186028,
      "loss": 0.7968,
      "num_input_tokens_seen": 59945304,
      "step": 103240
    },
    {
      "epoch": 15.377569258266309,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00462272792220235,
      "loss": 0.8108,
      "num_input_tokens_seen": 59948344,
      "step": 103245
    },
    {
      "epoch": 15.378313970807268,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.004621320220532138,
      "loss": 0.8076,
      "num_input_tokens_seen": 59951256,
      "step": 103250
    },
    {
      "epoch": 15.379058683348228,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.004619912694199168,
      "loss": 0.7923,
      "num_input_tokens_seen": 59954200,
      "step": 103255
    },
    {
      "epoch": 15.379803395889187,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.004618505343227226,
      "loss": 0.7935,
      "num_input_tokens_seen": 59957048,
      "step": 103260
    },
    {
      "epoch": 15.380548108430146,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.004617098167640088,
      "loss": 0.7963,
      "num_input_tokens_seen": 59959608,
      "step": 103265
    },
    {
      "epoch": 15.381292820971105,
      "grad_norm": 0.046875,
      "learning_rate": 0.004615691167461517,
      "loss": 0.8199,
      "num_input_tokens_seen": 59962328,
      "step": 103270
    },
    {
      "epoch": 15.382037533512065,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.004614284342715294,
      "loss": 0.7793,
      "num_input_tokens_seen": 59965560,
      "step": 103275
    },
    {
      "epoch": 15.382782246053024,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.004612877693425177,
      "loss": 0.7662,
      "num_input_tokens_seen": 59968280,
      "step": 103280
    },
    {
      "epoch": 15.383526958593983,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.00461147121961494,
      "loss": 0.8054,
      "num_input_tokens_seen": 59971256,
      "step": 103285
    },
    {
      "epoch": 15.384271671134941,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.004610064921308332,
      "loss": 0.7766,
      "num_input_tokens_seen": 59974168,
      "step": 103290
    },
    {
      "epoch": 15.3850163836759,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.004608658798529123,
      "loss": 0.7961,
      "num_input_tokens_seen": 59977400,
      "step": 103295
    },
    {
      "epoch": 15.38576109621686,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.004607252851301063,
      "loss": 0.7926,
      "num_input_tokens_seen": 59979960,
      "step": 103300
    },
    {
      "epoch": 15.38650580875782,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.004605847079647897,
      "loss": 0.7931,
      "num_input_tokens_seen": 59982936,
      "step": 103305
    },
    {
      "epoch": 15.387250521298778,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0046044414835933875,
      "loss": 0.7822,
      "num_input_tokens_seen": 59985592,
      "step": 103310
    },
    {
      "epoch": 15.387995233839739,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.004603036063161272,
      "loss": 0.806,
      "num_input_tokens_seen": 59988568,
      "step": 103315
    },
    {
      "epoch": 15.388739946380698,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.004601630818375295,
      "loss": 0.8167,
      "num_input_tokens_seen": 59991352,
      "step": 103320
    },
    {
      "epoch": 15.389484658921656,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.004600225749259192,
      "loss": 0.7884,
      "num_input_tokens_seen": 59994360,
      "step": 103325
    },
    {
      "epoch": 15.390229371462615,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0045988208558367105,
      "loss": 0.7818,
      "num_input_tokens_seen": 59997528,
      "step": 103330
    },
    {
      "epoch": 15.390974084003574,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.004597416138131575,
      "loss": 0.816,
      "num_input_tokens_seen": 60000248,
      "step": 103335
    },
    {
      "epoch": 15.391718796544534,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.004596011596167525,
      "loss": 0.7889,
      "num_input_tokens_seen": 60003128,
      "step": 103340
    },
    {
      "epoch": 15.392463509085493,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0045946072299682846,
      "loss": 0.8012,
      "num_input_tokens_seen": 60005784,
      "step": 103345
    },
    {
      "epoch": 15.393208221626452,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.004593203039557576,
      "loss": 0.8104,
      "num_input_tokens_seen": 60008952,
      "step": 103350
    },
    {
      "epoch": 15.39395293416741,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.004591799024959129,
      "loss": 0.7805,
      "num_input_tokens_seen": 60011768,
      "step": 103355
    },
    {
      "epoch": 15.394697646708371,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.004590395186196654,
      "loss": 0.8057,
      "num_input_tokens_seen": 60014680,
      "step": 103360
    },
    {
      "epoch": 15.39544235924933,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0045889915232938784,
      "loss": 0.8105,
      "num_input_tokens_seen": 60017624,
      "step": 103365
    },
    {
      "epoch": 15.396187071790289,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0045875880362745095,
      "loss": 0.833,
      "num_input_tokens_seen": 60020600,
      "step": 103370
    },
    {
      "epoch": 15.396931784331247,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.004586184725162258,
      "loss": 0.8077,
      "num_input_tokens_seen": 60023480,
      "step": 103375
    },
    {
      "epoch": 15.397676496872208,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.004584781589980826,
      "loss": 0.7956,
      "num_input_tokens_seen": 60026520,
      "step": 103380
    },
    {
      "epoch": 15.398421209413167,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.004583378630753928,
      "loss": 0.7999,
      "num_input_tokens_seen": 60029400,
      "step": 103385
    },
    {
      "epoch": 15.399165921954125,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.004581975847505262,
      "loss": 0.7879,
      "num_input_tokens_seen": 60032216,
      "step": 103390
    },
    {
      "epoch": 15.399910634495084,
      "grad_norm": 0.0546875,
      "learning_rate": 0.00458057324025852,
      "loss": 0.7963,
      "num_input_tokens_seen": 60035416,
      "step": 103395
    },
    {
      "epoch": 15.400655347036045,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.00457917080903741,
      "loss": 0.7994,
      "num_input_tokens_seen": 60038456,
      "step": 103400
    },
    {
      "epoch": 15.401400059577004,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.004577768553865611,
      "loss": 0.7796,
      "num_input_tokens_seen": 60041496,
      "step": 103405
    },
    {
      "epoch": 15.402144772117962,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.004576366474766823,
      "loss": 0.806,
      "num_input_tokens_seen": 60044472,
      "step": 103410
    },
    {
      "epoch": 15.402889484658921,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.004574964571764728,
      "loss": 0.8424,
      "num_input_tokens_seen": 60047224,
      "step": 103415
    },
    {
      "epoch": 15.403634197199882,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.004573562844883015,
      "loss": 0.7908,
      "num_input_tokens_seen": 60050200,
      "step": 103420
    },
    {
      "epoch": 15.40437890974084,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.004572161294145359,
      "loss": 0.7803,
      "num_input_tokens_seen": 60053304,
      "step": 103425
    },
    {
      "epoch": 15.405123622281799,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.004570759919575437,
      "loss": 0.7908,
      "num_input_tokens_seen": 60056280,
      "step": 103430
    },
    {
      "epoch": 15.405868334822758,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.00456935872119693,
      "loss": 0.8126,
      "num_input_tokens_seen": 60059224,
      "step": 103435
    },
    {
      "epoch": 15.406613047363718,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.004567957699033505,
      "loss": 0.7975,
      "num_input_tokens_seen": 60062072,
      "step": 103440
    },
    {
      "epoch": 15.407357759904677,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.004566556853108834,
      "loss": 0.7916,
      "num_input_tokens_seen": 60064696,
      "step": 103445
    },
    {
      "epoch": 15.408102472445636,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0045651561834465755,
      "loss": 0.7826,
      "num_input_tokens_seen": 60067736,
      "step": 103450
    },
    {
      "epoch": 15.408847184986595,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.004563755690070402,
      "loss": 0.7831,
      "num_input_tokens_seen": 60070392,
      "step": 103455
    },
    {
      "epoch": 15.409591897527555,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.004562355373003964,
      "loss": 0.7885,
      "num_input_tokens_seen": 60073080,
      "step": 103460
    },
    {
      "epoch": 15.410336610068514,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.00456095523227093,
      "loss": 0.775,
      "num_input_tokens_seen": 60076120,
      "step": 103465
    },
    {
      "epoch": 15.411081322609473,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0045595552678949465,
      "loss": 0.7782,
      "num_input_tokens_seen": 60079256,
      "step": 103470
    },
    {
      "epoch": 15.411826035150431,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.004558155479899661,
      "loss": 0.8356,
      "num_input_tokens_seen": 60082072,
      "step": 103475
    },
    {
      "epoch": 15.41257074769139,
      "grad_norm": 0.08203125,
      "learning_rate": 0.004556755868308731,
      "loss": 0.8082,
      "num_input_tokens_seen": 60084920,
      "step": 103480
    },
    {
      "epoch": 15.41331546023235,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.004555356433145793,
      "loss": 0.8157,
      "num_input_tokens_seen": 60087448,
      "step": 103485
    },
    {
      "epoch": 15.41406017277331,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0045539571744344955,
      "loss": 0.7977,
      "num_input_tokens_seen": 60090360,
      "step": 103490
    },
    {
      "epoch": 15.414804885314268,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.004552558092198469,
      "loss": 0.7979,
      "num_input_tokens_seen": 60093336,
      "step": 103495
    },
    {
      "epoch": 15.415549597855227,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.004551159186461363,
      "loss": 0.7823,
      "num_input_tokens_seen": 60096152,
      "step": 103500
    },
    {
      "epoch": 15.416294310396188,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.004549760457246802,
      "loss": 0.7811,
      "num_input_tokens_seen": 60098840,
      "step": 103505
    },
    {
      "epoch": 15.417039022937146,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0045483619045784155,
      "loss": 0.7874,
      "num_input_tokens_seen": 60101752,
      "step": 103510
    },
    {
      "epoch": 15.417783735478105,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.004546963528479833,
      "loss": 0.774,
      "num_input_tokens_seen": 60104536,
      "step": 103515
    },
    {
      "epoch": 15.418528448019064,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.004545565328974672,
      "loss": 0.8113,
      "num_input_tokens_seen": 60107576,
      "step": 103520
    },
    {
      "epoch": 15.419273160560024,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.004544167306086566,
      "loss": 0.8169,
      "num_input_tokens_seen": 60110776,
      "step": 103525
    },
    {
      "epoch": 15.420017873100983,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.004542769459839119,
      "loss": 0.8034,
      "num_input_tokens_seen": 60113592,
      "step": 103530
    },
    {
      "epoch": 15.420762585641942,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.004541371790255961,
      "loss": 0.7623,
      "num_input_tokens_seen": 60116568,
      "step": 103535
    },
    {
      "epoch": 15.4215072981829,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.004539974297360691,
      "loss": 0.8051,
      "num_input_tokens_seen": 60119320,
      "step": 103540
    },
    {
      "epoch": 15.422252010723861,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0045385769811769295,
      "loss": 0.7642,
      "num_input_tokens_seen": 60122168,
      "step": 103545
    },
    {
      "epoch": 15.42299672326482,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.004537179841728276,
      "loss": 0.8329,
      "num_input_tokens_seen": 60125240,
      "step": 103550
    },
    {
      "epoch": 15.423741435805779,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.004535782879038332,
      "loss": 0.8006,
      "num_input_tokens_seen": 60128312,
      "step": 103555
    },
    {
      "epoch": 15.424486148346737,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.004534386093130705,
      "loss": 0.7728,
      "num_input_tokens_seen": 60131160,
      "step": 103560
    },
    {
      "epoch": 15.425230860887698,
      "grad_norm": 0.046875,
      "learning_rate": 0.004532989484028983,
      "loss": 0.7858,
      "num_input_tokens_seen": 60133656,
      "step": 103565
    },
    {
      "epoch": 15.425975573428657,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.004531593051756772,
      "loss": 0.7841,
      "num_input_tokens_seen": 60136632,
      "step": 103570
    },
    {
      "epoch": 15.426720285969616,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.004530196796337655,
      "loss": 0.8028,
      "num_input_tokens_seen": 60139320,
      "step": 103575
    },
    {
      "epoch": 15.427464998510574,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.004528800717795223,
      "loss": 0.8184,
      "num_input_tokens_seen": 60142200,
      "step": 103580
    },
    {
      "epoch": 15.428209711051535,
      "grad_norm": 0.1796875,
      "learning_rate": 0.00452740481615306,
      "loss": 0.8388,
      "num_input_tokens_seen": 60145304,
      "step": 103585
    },
    {
      "epoch": 15.428954423592494,
      "grad_norm": 0.0546875,
      "learning_rate": 0.004526009091434743,
      "loss": 0.798,
      "num_input_tokens_seen": 60148248,
      "step": 103590
    },
    {
      "epoch": 15.429699136133452,
      "grad_norm": 0.03515625,
      "learning_rate": 0.004524613543663864,
      "loss": 0.7983,
      "num_input_tokens_seen": 60151352,
      "step": 103595
    },
    {
      "epoch": 15.430443848674411,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.004523218172863986,
      "loss": 0.7977,
      "num_input_tokens_seen": 60154200,
      "step": 103600
    },
    {
      "epoch": 15.431188561215372,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004521822979058694,
      "loss": 0.793,
      "num_input_tokens_seen": 60156824,
      "step": 103605
    },
    {
      "epoch": 15.43193327375633,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.004520427962271549,
      "loss": 0.7985,
      "num_input_tokens_seen": 60159512,
      "step": 103610
    },
    {
      "epoch": 15.43267798629729,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.004519033122526127,
      "loss": 0.7908,
      "num_input_tokens_seen": 60162456,
      "step": 103615
    },
    {
      "epoch": 15.433422698838248,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0045176384598459825,
      "loss": 0.7698,
      "num_input_tokens_seen": 60165144,
      "step": 103620
    },
    {
      "epoch": 15.434167411379208,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.004516243974254688,
      "loss": 0.8043,
      "num_input_tokens_seen": 60168120,
      "step": 103625
    },
    {
      "epoch": 15.434912123920167,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.004514849665775796,
      "loss": 0.8148,
      "num_input_tokens_seen": 60170680,
      "step": 103630
    },
    {
      "epoch": 15.435656836461126,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.004513455534432859,
      "loss": 0.7951,
      "num_input_tokens_seen": 60173816,
      "step": 103635
    },
    {
      "epoch": 15.436401549002085,
      "grad_norm": 0.05078125,
      "learning_rate": 0.004512061580249436,
      "loss": 0.8554,
      "num_input_tokens_seen": 60176728,
      "step": 103640
    },
    {
      "epoch": 15.437146261543045,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0045106678032490735,
      "loss": 0.7948,
      "num_input_tokens_seen": 60179608,
      "step": 103645
    },
    {
      "epoch": 15.437890974084004,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0045092742034553174,
      "loss": 0.7684,
      "num_input_tokens_seen": 60182808,
      "step": 103650
    },
    {
      "epoch": 15.438635686624963,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.004507880780891706,
      "loss": 0.8075,
      "num_input_tokens_seen": 60186008,
      "step": 103655
    },
    {
      "epoch": 15.439380399165922,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.004506487535581791,
      "loss": 0.802,
      "num_input_tokens_seen": 60188856,
      "step": 103660
    },
    {
      "epoch": 15.44012511170688,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.004505094467549103,
      "loss": 0.7954,
      "num_input_tokens_seen": 60191864,
      "step": 103665
    },
    {
      "epoch": 15.44086982424784,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.004503701576817174,
      "loss": 0.8018,
      "num_input_tokens_seen": 60194584,
      "step": 103670
    },
    {
      "epoch": 15.4416145367888,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.004502308863409542,
      "loss": 0.7877,
      "num_input_tokens_seen": 60197624,
      "step": 103675
    },
    {
      "epoch": 15.442359249329758,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.004500916327349729,
      "loss": 0.804,
      "num_input_tokens_seen": 60200632,
      "step": 103680
    },
    {
      "epoch": 15.443103961870717,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.004499523968661268,
      "loss": 0.7887,
      "num_input_tokens_seen": 60203608,
      "step": 103685
    },
    {
      "epoch": 15.443848674411678,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.004498131787367673,
      "loss": 0.8005,
      "num_input_tokens_seen": 60206264,
      "step": 103690
    },
    {
      "epoch": 15.444593386952636,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.004496739783492473,
      "loss": 0.7945,
      "num_input_tokens_seen": 60209176,
      "step": 103695
    },
    {
      "epoch": 15.445338099493595,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.004495347957059177,
      "loss": 0.79,
      "num_input_tokens_seen": 60212056,
      "step": 103700
    },
    {
      "epoch": 15.446082812034554,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.004493956308091304,
      "loss": 0.8071,
      "num_input_tokens_seen": 60214968,
      "step": 103705
    },
    {
      "epoch": 15.446827524575514,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0044925648366123585,
      "loss": 0.7832,
      "num_input_tokens_seen": 60217880,
      "step": 103710
    },
    {
      "epoch": 15.447572237116473,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.004491173542645847,
      "loss": 0.7939,
      "num_input_tokens_seen": 60221112,
      "step": 103715
    },
    {
      "epoch": 15.448316949657432,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.004489782426215283,
      "loss": 0.7827,
      "num_input_tokens_seen": 60223896,
      "step": 103720
    },
    {
      "epoch": 15.44906166219839,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.004488391487344158,
      "loss": 0.7869,
      "num_input_tokens_seen": 60226616,
      "step": 103725
    },
    {
      "epoch": 15.449806374739351,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.004487000726055979,
      "loss": 0.7862,
      "num_input_tokens_seen": 60229464,
      "step": 103730
    },
    {
      "epoch": 15.45055108728031,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.004485610142374235,
      "loss": 0.7826,
      "num_input_tokens_seen": 60232536,
      "step": 103735
    },
    {
      "epoch": 15.451295799821269,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.004484219736322425,
      "loss": 0.8206,
      "num_input_tokens_seen": 60235512,
      "step": 103740
    },
    {
      "epoch": 15.452040512362228,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.00448282950792403,
      "loss": 0.8069,
      "num_input_tokens_seen": 60238328,
      "step": 103745
    },
    {
      "epoch": 15.452785224903188,
      "grad_norm": 0.052734375,
      "learning_rate": 0.004481439457202545,
      "loss": 0.8115,
      "num_input_tokens_seen": 60240824,
      "step": 103750
    },
    {
      "epoch": 15.453529937444147,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.004480049584181452,
      "loss": 0.813,
      "num_input_tokens_seen": 60243352,
      "step": 103755
    },
    {
      "epoch": 15.454274649985106,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004478659888884222,
      "loss": 0.799,
      "num_input_tokens_seen": 60246328,
      "step": 103760
    },
    {
      "epoch": 15.455019362526064,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.004477270371334347,
      "loss": 0.8016,
      "num_input_tokens_seen": 60249304,
      "step": 103765
    },
    {
      "epoch": 15.455764075067025,
      "grad_norm": 0.056640625,
      "learning_rate": 0.004475881031555292,
      "loss": 0.8019,
      "num_input_tokens_seen": 60252056,
      "step": 103770
    },
    {
      "epoch": 15.456508787607984,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.004474491869570532,
      "loss": 0.7974,
      "num_input_tokens_seen": 60255064,
      "step": 103775
    },
    {
      "epoch": 15.457253500148942,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0044731028854035305,
      "loss": 0.811,
      "num_input_tokens_seen": 60257944,
      "step": 103780
    },
    {
      "epoch": 15.457998212689901,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00447171407907776,
      "loss": 0.8051,
      "num_input_tokens_seen": 60260696,
      "step": 103785
    },
    {
      "epoch": 15.458742925230862,
      "grad_norm": 0.0390625,
      "learning_rate": 0.00447032545061668,
      "loss": 0.798,
      "num_input_tokens_seen": 60263800,
      "step": 103790
    },
    {
      "epoch": 15.45948763777182,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.004468937000043744,
      "loss": 0.7813,
      "num_input_tokens_seen": 60266840,
      "step": 103795
    },
    {
      "epoch": 15.46023235031278,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00446754872738242,
      "loss": 0.825,
      "num_input_tokens_seen": 60270040,
      "step": 103800
    },
    {
      "epoch": 15.460977062853738,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.004466160632656151,
      "loss": 0.806,
      "num_input_tokens_seen": 60272824,
      "step": 103805
    },
    {
      "epoch": 15.461721775394698,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.004464772715888396,
      "loss": 0.7987,
      "num_input_tokens_seen": 60275960,
      "step": 103810
    },
    {
      "epoch": 15.462466487935657,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.004463384977102593,
      "loss": 0.8107,
      "num_input_tokens_seen": 60278456,
      "step": 103815
    },
    {
      "epoch": 15.463211200476616,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.004461997416322198,
      "loss": 0.7893,
      "num_input_tokens_seen": 60281464,
      "step": 103820
    },
    {
      "epoch": 15.463955913017575,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.004460610033570641,
      "loss": 0.7816,
      "num_input_tokens_seen": 60284440,
      "step": 103825
    },
    {
      "epoch": 15.464700625558535,
      "grad_norm": 0.0546875,
      "learning_rate": 0.004459222828871371,
      "loss": 0.809,
      "num_input_tokens_seen": 60287256,
      "step": 103830
    },
    {
      "epoch": 15.465445338099494,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.004457835802247817,
      "loss": 0.7875,
      "num_input_tokens_seen": 60289976,
      "step": 103835
    },
    {
      "epoch": 15.466190050640453,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.004456448953723413,
      "loss": 0.7946,
      "num_input_tokens_seen": 60292920,
      "step": 103840
    },
    {
      "epoch": 15.466934763181412,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.004455062283321589,
      "loss": 0.7913,
      "num_input_tokens_seen": 60296184,
      "step": 103845
    },
    {
      "epoch": 15.46767947572237,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.004453675791065765,
      "loss": 0.781,
      "num_input_tokens_seen": 60299128,
      "step": 103850
    },
    {
      "epoch": 15.46842418826333,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0044522894769793726,
      "loss": 0.8069,
      "num_input_tokens_seen": 60301944,
      "step": 103855
    },
    {
      "epoch": 15.46916890080429,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.004450903341085826,
      "loss": 0.8208,
      "num_input_tokens_seen": 60304600,
      "step": 103860
    },
    {
      "epoch": 15.469913613345248,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.004449517383408552,
      "loss": 0.7941,
      "num_input_tokens_seen": 60307320,
      "step": 103865
    },
    {
      "epoch": 15.470658325886207,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.004448131603970957,
      "loss": 0.7927,
      "num_input_tokens_seen": 60309976,
      "step": 103870
    },
    {
      "epoch": 15.471403038427168,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.00444674600279645,
      "loss": 0.787,
      "num_input_tokens_seen": 60312824,
      "step": 103875
    },
    {
      "epoch": 15.472147750968126,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.004445360579908448,
      "loss": 0.7905,
      "num_input_tokens_seen": 60315480,
      "step": 103880
    },
    {
      "epoch": 15.472892463509085,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.004443975335330347,
      "loss": 0.7988,
      "num_input_tokens_seen": 60318456,
      "step": 103885
    },
    {
      "epoch": 15.473637176050044,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.004442590269085558,
      "loss": 0.7898,
      "num_input_tokens_seen": 60321144,
      "step": 103890
    },
    {
      "epoch": 15.474381888591004,
      "grad_norm": 0.078125,
      "learning_rate": 0.004441205381197472,
      "loss": 0.8039,
      "num_input_tokens_seen": 60324312,
      "step": 103895
    },
    {
      "epoch": 15.475126601131963,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.004439820671689496,
      "loss": 0.8461,
      "num_input_tokens_seen": 60327160,
      "step": 103900
    },
    {
      "epoch": 15.475871313672922,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.004438436140585016,
      "loss": 0.794,
      "num_input_tokens_seen": 60329848,
      "step": 103905
    },
    {
      "epoch": 15.47661602621388,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.004437051787907422,
      "loss": 0.8127,
      "num_input_tokens_seen": 60332632,
      "step": 103910
    },
    {
      "epoch": 15.477360738754841,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.004435667613680103,
      "loss": 0.7747,
      "num_input_tokens_seen": 60335544,
      "step": 103915
    },
    {
      "epoch": 15.4781054512958,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.004434283617926439,
      "loss": 0.7998,
      "num_input_tokens_seen": 60338392,
      "step": 103920
    },
    {
      "epoch": 15.478850163836759,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.004432899800669818,
      "loss": 0.7936,
      "num_input_tokens_seen": 60341048,
      "step": 103925
    },
    {
      "epoch": 15.479594876377718,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.00443151616193361,
      "loss": 0.7944,
      "num_input_tokens_seen": 60343896,
      "step": 103930
    },
    {
      "epoch": 15.480339588918678,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0044301327017412006,
      "loss": 0.8185,
      "num_input_tokens_seen": 60346744,
      "step": 103935
    },
    {
      "epoch": 15.481084301459637,
      "grad_norm": 0.044921875,
      "learning_rate": 0.004428749420115951,
      "loss": 0.7756,
      "num_input_tokens_seen": 60349784,
      "step": 103940
    },
    {
      "epoch": 15.481829014000596,
      "grad_norm": 0.033203125,
      "learning_rate": 0.004427366317081242,
      "loss": 0.7957,
      "num_input_tokens_seen": 60352728,
      "step": 103945
    },
    {
      "epoch": 15.482573726541554,
      "grad_norm": 0.03125,
      "learning_rate": 0.004425983392660432,
      "loss": 0.8053,
      "num_input_tokens_seen": 60355672,
      "step": 103950
    },
    {
      "epoch": 15.483318439082515,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.004424600646876882,
      "loss": 0.7892,
      "num_input_tokens_seen": 60358552,
      "step": 103955
    },
    {
      "epoch": 15.484063151623474,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.004423218079753959,
      "loss": 0.7854,
      "num_input_tokens_seen": 60361656,
      "step": 103960
    },
    {
      "epoch": 15.484807864164432,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.004421835691315015,
      "loss": 0.7893,
      "num_input_tokens_seen": 60364504,
      "step": 103965
    },
    {
      "epoch": 15.485552576705391,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.004420453481583407,
      "loss": 0.7985,
      "num_input_tokens_seen": 60367288,
      "step": 103970
    },
    {
      "epoch": 15.486297289246352,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004419071450582486,
      "loss": 0.8068,
      "num_input_tokens_seen": 60370168,
      "step": 103975
    },
    {
      "epoch": 15.48704200178731,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0044176895983356,
      "loss": 0.7847,
      "num_input_tokens_seen": 60373016,
      "step": 103980
    },
    {
      "epoch": 15.48778671432827,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.004416307924866089,
      "loss": 0.8164,
      "num_input_tokens_seen": 60376088,
      "step": 103985
    },
    {
      "epoch": 15.488531426869228,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0044149264301973,
      "loss": 0.8023,
      "num_input_tokens_seen": 60378968,
      "step": 103990
    },
    {
      "epoch": 15.489276139410187,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.004413545114352574,
      "loss": 0.8024,
      "num_input_tokens_seen": 60381752,
      "step": 103995
    },
    {
      "epoch": 15.490020851951147,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.004412163977355238,
      "loss": 0.794,
      "num_input_tokens_seen": 60384600,
      "step": 104000
    },
    {
      "epoch": 15.490765564492106,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.004410783019228635,
      "loss": 0.7908,
      "num_input_tokens_seen": 60387416,
      "step": 104005
    },
    {
      "epoch": 15.491510277033065,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.004409402239996085,
      "loss": 0.8033,
      "num_input_tokens_seen": 60390584,
      "step": 104010
    },
    {
      "epoch": 15.492254989574024,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.004408021639680926,
      "loss": 0.7966,
      "num_input_tokens_seen": 60393624,
      "step": 104015
    },
    {
      "epoch": 15.492999702114984,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004406641218306471,
      "loss": 0.7934,
      "num_input_tokens_seen": 60396600,
      "step": 104020
    },
    {
      "epoch": 15.493744414655943,
      "grad_norm": 0.052734375,
      "learning_rate": 0.004405260975896052,
      "loss": 0.8288,
      "num_input_tokens_seen": 60399160,
      "step": 104025
    },
    {
      "epoch": 15.494489127196902,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.00440388091247298,
      "loss": 0.8089,
      "num_input_tokens_seen": 60401976,
      "step": 104030
    },
    {
      "epoch": 15.49523383973786,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.00440250102806057,
      "loss": 0.799,
      "num_input_tokens_seen": 60404760,
      "step": 104035
    },
    {
      "epoch": 15.495978552278821,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.004401121322682135,
      "loss": 0.7905,
      "num_input_tokens_seen": 60407896,
      "step": 104040
    },
    {
      "epoch": 15.49672326481978,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.0043997417963609784,
      "loss": 0.7737,
      "num_input_tokens_seen": 60410712,
      "step": 104045
    },
    {
      "epoch": 15.497467977360738,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004398362449120416,
      "loss": 0.7723,
      "num_input_tokens_seen": 60413752,
      "step": 104050
    },
    {
      "epoch": 15.498212689901697,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.004396983280983737,
      "loss": 0.7902,
      "num_input_tokens_seen": 60416568,
      "step": 104055
    },
    {
      "epoch": 15.498957402442658,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.004395604291974257,
      "loss": 0.7802,
      "num_input_tokens_seen": 60419640,
      "step": 104060
    },
    {
      "epoch": 15.499702114983616,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.004394225482115258,
      "loss": 0.8015,
      "num_input_tokens_seen": 60422712,
      "step": 104065
    },
    {
      "epoch": 15.500446827524575,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.004392846851430047,
      "loss": 0.8131,
      "num_input_tokens_seen": 60425688,
      "step": 104070
    },
    {
      "epoch": 15.501191540065534,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.004391468399941906,
      "loss": 0.7938,
      "num_input_tokens_seen": 60428472,
      "step": 104075
    },
    {
      "epoch": 15.501936252606495,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.004390090127674121,
      "loss": 0.7871,
      "num_input_tokens_seen": 60431640,
      "step": 104080
    },
    {
      "epoch": 15.502680965147453,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.004388712034649984,
      "loss": 0.7973,
      "num_input_tokens_seen": 60434360,
      "step": 104085
    },
    {
      "epoch": 15.503425677688412,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.004387334120892768,
      "loss": 0.7896,
      "num_input_tokens_seen": 60437208,
      "step": 104090
    },
    {
      "epoch": 15.50417039022937,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.004385956386425759,
      "loss": 0.7872,
      "num_input_tokens_seen": 60440184,
      "step": 104095
    },
    {
      "epoch": 15.504915102770331,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00438457883127223,
      "loss": 0.8136,
      "num_input_tokens_seen": 60442968,
      "step": 104100
    },
    {
      "epoch": 15.50565981531129,
      "grad_norm": 0.03515625,
      "learning_rate": 0.004383201455455454,
      "loss": 0.8217,
      "num_input_tokens_seen": 60446200,
      "step": 104105
    },
    {
      "epoch": 15.506404527852249,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.004381824258998697,
      "loss": 0.7995,
      "num_input_tokens_seen": 60448920,
      "step": 104110
    },
    {
      "epoch": 15.507149240393208,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.004380447241925223,
      "loss": 0.7857,
      "num_input_tokens_seen": 60451416,
      "step": 104115
    },
    {
      "epoch": 15.507893952934168,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.004379070404258305,
      "loss": 0.8031,
      "num_input_tokens_seen": 60454456,
      "step": 104120
    },
    {
      "epoch": 15.508638665475127,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.004377693746021194,
      "loss": 0.7779,
      "num_input_tokens_seen": 60457368,
      "step": 104125
    },
    {
      "epoch": 15.509383378016086,
      "grad_norm": 0.0264892578125,
      "learning_rate": 0.0043763172672371545,
      "loss": 0.8369,
      "num_input_tokens_seen": 60459992,
      "step": 104130
    },
    {
      "epoch": 15.510128090557044,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.004374940967929433,
      "loss": 0.7976,
      "num_input_tokens_seen": 60462840,
      "step": 104135
    },
    {
      "epoch": 15.510872803098005,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.004373564848121291,
      "loss": 0.7985,
      "num_input_tokens_seen": 60465816,
      "step": 104140
    },
    {
      "epoch": 15.511617515638964,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0043721889078359635,
      "loss": 0.8008,
      "num_input_tokens_seen": 60468568,
      "step": 104145
    },
    {
      "epoch": 15.512362228179922,
      "grad_norm": 0.052734375,
      "learning_rate": 0.004370813147096708,
      "loss": 0.7949,
      "num_input_tokens_seen": 60471256,
      "step": 104150
    },
    {
      "epoch": 15.513106940720881,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0043694375659267615,
      "loss": 0.79,
      "num_input_tokens_seen": 60474456,
      "step": 104155
    },
    {
      "epoch": 15.513851653261842,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0043680621643493595,
      "loss": 0.803,
      "num_input_tokens_seen": 60476952,
      "step": 104160
    },
    {
      "epoch": 15.5145963658028,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.004366686942387745,
      "loss": 0.8027,
      "num_input_tokens_seen": 60479864,
      "step": 104165
    },
    {
      "epoch": 15.51534107834376,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.004365311900065149,
      "loss": 0.8053,
      "num_input_tokens_seen": 60482904,
      "step": 104170
    },
    {
      "epoch": 15.516085790884718,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.004363937037404798,
      "loss": 0.7956,
      "num_input_tokens_seen": 60485496,
      "step": 104175
    },
    {
      "epoch": 15.516830503425677,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.004362562354429916,
      "loss": 0.7829,
      "num_input_tokens_seen": 60488408,
      "step": 104180
    },
    {
      "epoch": 15.517575215966637,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.004361187851163737,
      "loss": 0.7981,
      "num_input_tokens_seen": 60491288,
      "step": 104185
    },
    {
      "epoch": 15.518319928507596,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.004359813527629476,
      "loss": 0.7785,
      "num_input_tokens_seen": 60494200,
      "step": 104190
    },
    {
      "epoch": 15.519064641048555,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.004358439383850348,
      "loss": 0.7754,
      "num_input_tokens_seen": 60496952,
      "step": 104195
    },
    {
      "epoch": 15.519809353589514,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.004357065419849574,
      "loss": 0.7985,
      "num_input_tokens_seen": 60499928,
      "step": 104200
    },
    {
      "epoch": 15.520554066130474,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.004355691635650358,
      "loss": 0.8012,
      "num_input_tokens_seen": 60503128,
      "step": 104205
    },
    {
      "epoch": 15.521298778671433,
      "grad_norm": 0.0390625,
      "learning_rate": 0.00435431803127592,
      "loss": 0.82,
      "num_input_tokens_seen": 60506264,
      "step": 104210
    },
    {
      "epoch": 15.522043491212392,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.004352944606749454,
      "loss": 0.8176,
      "num_input_tokens_seen": 60509176,
      "step": 104215
    },
    {
      "epoch": 15.52278820375335,
      "grad_norm": 0.041015625,
      "learning_rate": 0.004351571362094171,
      "loss": 0.843,
      "num_input_tokens_seen": 60511960,
      "step": 104220
    },
    {
      "epoch": 15.523532916294311,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.004350198297333263,
      "loss": 0.7961,
      "num_input_tokens_seen": 60514904,
      "step": 104225
    },
    {
      "epoch": 15.52427762883527,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.004348825412489935,
      "loss": 0.8434,
      "num_input_tokens_seen": 60517752,
      "step": 104230
    },
    {
      "epoch": 15.525022341376228,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.004347452707587377,
      "loss": 0.7947,
      "num_input_tokens_seen": 60520824,
      "step": 104235
    },
    {
      "epoch": 15.525767053917187,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.0043460801826487785,
      "loss": 0.8007,
      "num_input_tokens_seen": 60523512,
      "step": 104240
    },
    {
      "epoch": 15.526511766458148,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.004344707837697327,
      "loss": 0.8072,
      "num_input_tokens_seen": 60526456,
      "step": 104245
    },
    {
      "epoch": 15.527256478999107,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.004343335672756202,
      "loss": 0.7706,
      "num_input_tokens_seen": 60529048,
      "step": 104250
    },
    {
      "epoch": 15.528001191540065,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.004341963687848596,
      "loss": 0.7908,
      "num_input_tokens_seen": 60532024,
      "step": 104255
    },
    {
      "epoch": 15.528745904081024,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.004340591882997675,
      "loss": 0.787,
      "num_input_tokens_seen": 60534744,
      "step": 104260
    },
    {
      "epoch": 15.529490616621985,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.004339220258226625,
      "loss": 0.8304,
      "num_input_tokens_seen": 60537688,
      "step": 104265
    },
    {
      "epoch": 15.530235329162943,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0043378488135586155,
      "loss": 0.7828,
      "num_input_tokens_seen": 60540600,
      "step": 104270
    },
    {
      "epoch": 15.530980041703902,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.004336477549016808,
      "loss": 0.8373,
      "num_input_tokens_seen": 60543192,
      "step": 104275
    },
    {
      "epoch": 15.53172475424486,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.004335106464624378,
      "loss": 0.8054,
      "num_input_tokens_seen": 60545880,
      "step": 104280
    },
    {
      "epoch": 15.532469466785821,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.004333735560404483,
      "loss": 0.7752,
      "num_input_tokens_seen": 60548792,
      "step": 104285
    },
    {
      "epoch": 15.53321417932678,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.004332364836380287,
      "loss": 0.7925,
      "num_input_tokens_seen": 60551864,
      "step": 104290
    },
    {
      "epoch": 15.533958891867739,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.004330994292574941,
      "loss": 0.7934,
      "num_input_tokens_seen": 60554808,
      "step": 104295
    },
    {
      "epoch": 15.534703604408698,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.004329623929011608,
      "loss": 0.7871,
      "num_input_tokens_seen": 60557784,
      "step": 104300
    },
    {
      "epoch": 15.535448316949658,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0043282537457134335,
      "loss": 0.7936,
      "num_input_tokens_seen": 60560696,
      "step": 104305
    },
    {
      "epoch": 15.536193029490617,
      "grad_norm": 0.04296875,
      "learning_rate": 0.004326883742703566,
      "loss": 0.7978,
      "num_input_tokens_seen": 60563512,
      "step": 104310
    },
    {
      "epoch": 15.536937742031576,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.00432551392000515,
      "loss": 0.7963,
      "num_input_tokens_seen": 60566552,
      "step": 104315
    },
    {
      "epoch": 15.537682454572534,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0043241442776413225,
      "loss": 0.8117,
      "num_input_tokens_seen": 60569464,
      "step": 104320
    },
    {
      "epoch": 15.538427167113493,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.004322774815635232,
      "loss": 0.8018,
      "num_input_tokens_seen": 60572280,
      "step": 104325
    },
    {
      "epoch": 15.539171879654454,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.004321405534010003,
      "loss": 0.7934,
      "num_input_tokens_seen": 60575192,
      "step": 104330
    },
    {
      "epoch": 15.539916592195413,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0043200364327887816,
      "loss": 0.7766,
      "num_input_tokens_seen": 60577912,
      "step": 104335
    },
    {
      "epoch": 15.540661304736371,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.004318667511994683,
      "loss": 0.7725,
      "num_input_tokens_seen": 60580664,
      "step": 104340
    },
    {
      "epoch": 15.541406017277332,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.0043172987716508454,
      "loss": 0.775,
      "num_input_tokens_seen": 60583640,
      "step": 104345
    },
    {
      "epoch": 15.54215072981829,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.004315930211780384,
      "loss": 0.8179,
      "num_input_tokens_seen": 60586552,
      "step": 104350
    },
    {
      "epoch": 15.54289544235925,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.004314561832406427,
      "loss": 0.7888,
      "num_input_tokens_seen": 60589272,
      "step": 104355
    },
    {
      "epoch": 15.543640154900208,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.004313193633552088,
      "loss": 0.7931,
      "num_input_tokens_seen": 60592376,
      "step": 104360
    },
    {
      "epoch": 15.544384867441167,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.004311825615240482,
      "loss": 0.7776,
      "num_input_tokens_seen": 60595160,
      "step": 104365
    },
    {
      "epoch": 15.545129579982127,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.004310457777494717,
      "loss": 0.815,
      "num_input_tokens_seen": 60597912,
      "step": 104370
    },
    {
      "epoch": 15.545874292523086,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0043090901203379,
      "loss": 0.781,
      "num_input_tokens_seen": 60600792,
      "step": 104375
    },
    {
      "epoch": 15.546619005064045,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0043077226437931455,
      "loss": 0.8121,
      "num_input_tokens_seen": 60603864,
      "step": 104380
    },
    {
      "epoch": 15.547363717605004,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.004306355347883544,
      "loss": 0.8016,
      "num_input_tokens_seen": 60606392,
      "step": 104385
    },
    {
      "epoch": 15.548108430145964,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.0043049882326322035,
      "loss": 0.7877,
      "num_input_tokens_seen": 60609464,
      "step": 104390
    },
    {
      "epoch": 15.548853142686923,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.00430362129806222,
      "loss": 0.7908,
      "num_input_tokens_seen": 60612376,
      "step": 104395
    },
    {
      "epoch": 15.549597855227882,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.004302254544196677,
      "loss": 0.81,
      "num_input_tokens_seen": 60615384,
      "step": 104400
    },
    {
      "epoch": 15.55034256776884,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.004300887971058675,
      "loss": 0.7873,
      "num_input_tokens_seen": 60618424,
      "step": 104405
    },
    {
      "epoch": 15.551087280309801,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.004299521578671292,
      "loss": 0.7993,
      "num_input_tokens_seen": 60621048,
      "step": 104410
    },
    {
      "epoch": 15.55183199285076,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004298155367057622,
      "loss": 0.8175,
      "num_input_tokens_seen": 60623896,
      "step": 104415
    },
    {
      "epoch": 15.552576705391719,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0042967893362407346,
      "loss": 0.8055,
      "num_input_tokens_seen": 60626616,
      "step": 104420
    },
    {
      "epoch": 15.553321417932677,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0042954234862437185,
      "loss": 0.7939,
      "num_input_tokens_seen": 60629240,
      "step": 104425
    },
    {
      "epoch": 15.554066130473638,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.004294057817089642,
      "loss": 0.7843,
      "num_input_tokens_seen": 60632248,
      "step": 104430
    },
    {
      "epoch": 15.554810843014597,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.00429269232880158,
      "loss": 0.8052,
      "num_input_tokens_seen": 60635192,
      "step": 104435
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.004291327021402596,
      "loss": 0.773,
      "num_input_tokens_seen": 60637784,
      "step": 104440
    },
    {
      "epoch": 15.556300268096514,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.004289961894915754,
      "loss": 0.822,
      "num_input_tokens_seen": 60640664,
      "step": 104445
    },
    {
      "epoch": 15.557044980637475,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.004288596949364127,
      "loss": 0.8611,
      "num_input_tokens_seen": 60643448,
      "step": 104450
    },
    {
      "epoch": 15.557789693178433,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.004287232184770762,
      "loss": 0.7703,
      "num_input_tokens_seen": 60646168,
      "step": 104455
    },
    {
      "epoch": 15.558534405719392,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.004285867601158725,
      "loss": 0.7958,
      "num_input_tokens_seen": 60649208,
      "step": 104460
    },
    {
      "epoch": 15.559279118260351,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.004284503198551063,
      "loss": 0.8042,
      "num_input_tokens_seen": 60651928,
      "step": 104465
    },
    {
      "epoch": 15.560023830801311,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.004283138976970833,
      "loss": 0.7843,
      "num_input_tokens_seen": 60655000,
      "step": 104470
    },
    {
      "epoch": 15.56076854334227,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.004281774936441077,
      "loss": 0.8023,
      "num_input_tokens_seen": 60657752,
      "step": 104475
    },
    {
      "epoch": 15.561513255883229,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.004280411076984835,
      "loss": 0.7909,
      "num_input_tokens_seen": 60660408,
      "step": 104480
    },
    {
      "epoch": 15.562257968424188,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.004279047398625158,
      "loss": 0.8076,
      "num_input_tokens_seen": 60663256,
      "step": 104485
    },
    {
      "epoch": 15.563002680965148,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.004277683901385073,
      "loss": 0.8065,
      "num_input_tokens_seen": 60666136,
      "step": 104490
    },
    {
      "epoch": 15.563747393506107,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004276320585287628,
      "loss": 0.7783,
      "num_input_tokens_seen": 60668824,
      "step": 104495
    },
    {
      "epoch": 15.564492106047066,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.004274957450355846,
      "loss": 0.7861,
      "num_input_tokens_seen": 60671480,
      "step": 104500
    },
    {
      "epoch": 15.565236818588025,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.0042735944966127565,
      "loss": 0.7961,
      "num_input_tokens_seen": 60674072,
      "step": 104505
    },
    {
      "epoch": 15.565981531128983,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.004272231724081384,
      "loss": 0.7958,
      "num_input_tokens_seen": 60677752,
      "step": 104510
    },
    {
      "epoch": 15.566726243669944,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.004270869132784756,
      "loss": 0.7912,
      "num_input_tokens_seen": 60680824,
      "step": 104515
    },
    {
      "epoch": 15.567470956210903,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.004269506722745891,
      "loss": 0.7856,
      "num_input_tokens_seen": 60684408,
      "step": 104520
    },
    {
      "epoch": 15.568215668751861,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0042681444939877975,
      "loss": 0.8002,
      "num_input_tokens_seen": 60687192,
      "step": 104525
    },
    {
      "epoch": 15.568960381292822,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.004266782446533501,
      "loss": 0.8015,
      "num_input_tokens_seen": 60690296,
      "step": 104530
    },
    {
      "epoch": 15.56970509383378,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.004265420580406002,
      "loss": 0.7859,
      "num_input_tokens_seen": 60693240,
      "step": 104535
    },
    {
      "epoch": 15.57044980637474,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.004264058895628317,
      "loss": 0.7869,
      "num_input_tokens_seen": 60696056,
      "step": 104540
    },
    {
      "epoch": 15.571194518915698,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00426269739222344,
      "loss": 0.7939,
      "num_input_tokens_seen": 60698840,
      "step": 104545
    },
    {
      "epoch": 15.571939231456657,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.004261336070214382,
      "loss": 0.8272,
      "num_input_tokens_seen": 60701464,
      "step": 104550
    },
    {
      "epoch": 15.572683943997617,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.004259974929624139,
      "loss": 0.8078,
      "num_input_tokens_seen": 60704856,
      "step": 104555
    },
    {
      "epoch": 15.573428656538576,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.004258613970475698,
      "loss": 0.794,
      "num_input_tokens_seen": 60707640,
      "step": 104560
    },
    {
      "epoch": 15.574173369079535,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00425725319279206,
      "loss": 0.7905,
      "num_input_tokens_seen": 60710264,
      "step": 104565
    },
    {
      "epoch": 15.574918081620494,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0042558925965962115,
      "loss": 0.7823,
      "num_input_tokens_seen": 60713048,
      "step": 104570
    },
    {
      "epoch": 15.575662794161454,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.004254532181911138,
      "loss": 0.7843,
      "num_input_tokens_seen": 60716120,
      "step": 104575
    },
    {
      "epoch": 15.576407506702413,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.004253171948759818,
      "loss": 0.7977,
      "num_input_tokens_seen": 60719032,
      "step": 104580
    },
    {
      "epoch": 15.577152219243372,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.004251811897165239,
      "loss": 0.8087,
      "num_input_tokens_seen": 60721560,
      "step": 104585
    },
    {
      "epoch": 15.57789693178433,
      "grad_norm": 0.0625,
      "learning_rate": 0.004250452027150372,
      "loss": 0.8044,
      "num_input_tokens_seen": 60724408,
      "step": 104590
    },
    {
      "epoch": 15.578641644325291,
      "grad_norm": 0.23046875,
      "learning_rate": 0.004249092338738194,
      "loss": 0.8233,
      "num_input_tokens_seen": 60727224,
      "step": 104595
    },
    {
      "epoch": 15.57938635686625,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.004247732831951677,
      "loss": 0.8093,
      "num_input_tokens_seen": 60731064,
      "step": 104600
    },
    {
      "epoch": 15.580131069407209,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.00424637350681378,
      "loss": 0.7974,
      "num_input_tokens_seen": 60733944,
      "step": 104605
    },
    {
      "epoch": 15.580875781948167,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.004245014363347476,
      "loss": 0.7972,
      "num_input_tokens_seen": 60736856,
      "step": 104610
    },
    {
      "epoch": 15.581620494489128,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0042436554015757195,
      "loss": 0.7865,
      "num_input_tokens_seen": 60739928,
      "step": 104615
    },
    {
      "epoch": 15.582365207030087,
      "grad_norm": 0.02734375,
      "learning_rate": 0.004242296621521479,
      "loss": 0.7906,
      "num_input_tokens_seen": 60742552,
      "step": 104620
    },
    {
      "epoch": 15.583109919571045,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.004240938023207698,
      "loss": 0.8047,
      "num_input_tokens_seen": 60745400,
      "step": 104625
    },
    {
      "epoch": 15.583854632112004,
      "grad_norm": 0.033203125,
      "learning_rate": 0.004239579606657338,
      "loss": 0.7947,
      "num_input_tokens_seen": 60748536,
      "step": 104630
    },
    {
      "epoch": 15.584599344652965,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0042382213718933455,
      "loss": 0.7857,
      "num_input_tokens_seen": 60751672,
      "step": 104635
    },
    {
      "epoch": 15.585344057193923,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.004236863318938664,
      "loss": 0.7938,
      "num_input_tokens_seen": 60754456,
      "step": 104640
    },
    {
      "epoch": 15.586088769734882,
      "grad_norm": 0.048828125,
      "learning_rate": 0.004235505447816238,
      "loss": 0.7877,
      "num_input_tokens_seen": 60757592,
      "step": 104645
    },
    {
      "epoch": 15.586833482275841,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.004234147758549003,
      "loss": 0.8211,
      "num_input_tokens_seen": 60760216,
      "step": 104650
    },
    {
      "epoch": 15.587578194816802,
      "grad_norm": 0.400390625,
      "learning_rate": 0.004232790251159902,
      "loss": 0.8161,
      "num_input_tokens_seen": 60763128,
      "step": 104655
    },
    {
      "epoch": 15.58832290735776,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004231432925671863,
      "loss": 0.8085,
      "num_input_tokens_seen": 60765944,
      "step": 104660
    },
    {
      "epoch": 15.589067619898719,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.004230075782107826,
      "loss": 0.8177,
      "num_input_tokens_seen": 60768856,
      "step": 104665
    },
    {
      "epoch": 15.589812332439678,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.004228718820490706,
      "loss": 0.8155,
      "num_input_tokens_seen": 60771672,
      "step": 104670
    },
    {
      "epoch": 15.590557044980638,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.00422736204084344,
      "loss": 0.8055,
      "num_input_tokens_seen": 60774648,
      "step": 104675
    },
    {
      "epoch": 15.591301757521597,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.004226005443188944,
      "loss": 0.7928,
      "num_input_tokens_seen": 60777464,
      "step": 104680
    },
    {
      "epoch": 15.592046470062556,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.004224649027550129,
      "loss": 0.7948,
      "num_input_tokens_seen": 60780120,
      "step": 104685
    },
    {
      "epoch": 15.592791182603515,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.004223292793949925,
      "loss": 0.7883,
      "num_input_tokens_seen": 60782904,
      "step": 104690
    },
    {
      "epoch": 15.593535895144473,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.004221936742411234,
      "loss": 0.7825,
      "num_input_tokens_seen": 60785976,
      "step": 104695
    },
    {
      "epoch": 15.594280607685434,
      "grad_norm": 0.02734375,
      "learning_rate": 0.004220580872956968,
      "loss": 0.7722,
      "num_input_tokens_seen": 60788728,
      "step": 104700
    },
    {
      "epoch": 15.595025320226393,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0042192251856100265,
      "loss": 0.7879,
      "num_input_tokens_seen": 60791608,
      "step": 104705
    },
    {
      "epoch": 15.595770032767351,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.004217869680393324,
      "loss": 0.7731,
      "num_input_tokens_seen": 60794584,
      "step": 104710
    },
    {
      "epoch": 15.59651474530831,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.004216514357329755,
      "loss": 0.798,
      "num_input_tokens_seen": 60797336,
      "step": 104715
    },
    {
      "epoch": 15.59725945784927,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.004215159216442213,
      "loss": 0.7873,
      "num_input_tokens_seen": 60799992,
      "step": 104720
    },
    {
      "epoch": 15.59800417039023,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.004213804257753597,
      "loss": 0.7988,
      "num_input_tokens_seen": 60803000,
      "step": 104725
    },
    {
      "epoch": 15.598748882931188,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.004212449481286791,
      "loss": 0.7737,
      "num_input_tokens_seen": 60805752,
      "step": 104730
    },
    {
      "epoch": 15.599493595472147,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.004211094887064693,
      "loss": 0.801,
      "num_input_tokens_seen": 60808760,
      "step": 104735
    },
    {
      "epoch": 15.600238308013108,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.004209740475110176,
      "loss": 0.7946,
      "num_input_tokens_seen": 60811928,
      "step": 104740
    },
    {
      "epoch": 15.600983020554066,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.004208386245446131,
      "loss": 0.8025,
      "num_input_tokens_seen": 60814424,
      "step": 104745
    },
    {
      "epoch": 15.601727733095025,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.004207032198095429,
      "loss": 0.798,
      "num_input_tokens_seen": 60817368,
      "step": 104750
    },
    {
      "epoch": 15.602472445635984,
      "grad_norm": 0.072265625,
      "learning_rate": 0.004205678333080952,
      "loss": 0.7997,
      "num_input_tokens_seen": 60820088,
      "step": 104755
    },
    {
      "epoch": 15.603217158176944,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.004204324650425569,
      "loss": 0.7978,
      "num_input_tokens_seen": 60822776,
      "step": 104760
    },
    {
      "epoch": 15.603961870717903,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.004202971150152149,
      "loss": 0.8,
      "num_input_tokens_seen": 60825720,
      "step": 104765
    },
    {
      "epoch": 15.604706583258862,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.004201617832283559,
      "loss": 0.8009,
      "num_input_tokens_seen": 60828568,
      "step": 104770
    },
    {
      "epoch": 15.60545129579982,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.004200264696842655,
      "loss": 0.7766,
      "num_input_tokens_seen": 60831640,
      "step": 104775
    },
    {
      "epoch": 15.606196008340781,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.004198911743852307,
      "loss": 0.8035,
      "num_input_tokens_seen": 60834648,
      "step": 104780
    },
    {
      "epoch": 15.60694072088174,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0041975589733353625,
      "loss": 0.799,
      "num_input_tokens_seen": 60837784,
      "step": 104785
    },
    {
      "epoch": 15.607685433422699,
      "grad_norm": 0.05078125,
      "learning_rate": 0.004196206385314686,
      "loss": 0.8027,
      "num_input_tokens_seen": 60840632,
      "step": 104790
    },
    {
      "epoch": 15.608430145963657,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0041948539798131205,
      "loss": 0.7723,
      "num_input_tokens_seen": 60843608,
      "step": 104795
    },
    {
      "epoch": 15.609174858504618,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.004193501756853512,
      "loss": 0.8098,
      "num_input_tokens_seen": 60846328,
      "step": 104800
    },
    {
      "epoch": 15.609919571045577,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.004192149716458712,
      "loss": 0.7809,
      "num_input_tokens_seen": 60849336,
      "step": 104805
    },
    {
      "epoch": 15.610664283586535,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004190797858651553,
      "loss": 0.8031,
      "num_input_tokens_seen": 60852248,
      "step": 104810
    },
    {
      "epoch": 15.611408996127494,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.004189446183454883,
      "loss": 0.8102,
      "num_input_tokens_seen": 60855512,
      "step": 104815
    },
    {
      "epoch": 15.612153708668455,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.004188094690891528,
      "loss": 0.8037,
      "num_input_tokens_seen": 60858648,
      "step": 104820
    },
    {
      "epoch": 15.612898421209414,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.004186743380984328,
      "loss": 0.7831,
      "num_input_tokens_seen": 60861624,
      "step": 104825
    },
    {
      "epoch": 15.613643133750372,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.0041853922537561095,
      "loss": 0.8076,
      "num_input_tokens_seen": 60864568,
      "step": 104830
    },
    {
      "epoch": 15.614387846291331,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.004184041309229697,
      "loss": 0.7967,
      "num_input_tokens_seen": 60867512,
      "step": 104835
    },
    {
      "epoch": 15.615132558832292,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.004182690547427915,
      "loss": 0.8152,
      "num_input_tokens_seen": 60870328,
      "step": 104840
    },
    {
      "epoch": 15.61587727137325,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.004181339968373577,
      "loss": 0.7931,
      "num_input_tokens_seen": 60872952,
      "step": 104845
    },
    {
      "epoch": 15.616621983914209,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.004179989572089507,
      "loss": 0.7915,
      "num_input_tokens_seen": 60875928,
      "step": 104850
    },
    {
      "epoch": 15.617366696455168,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.004178639358598513,
      "loss": 0.8032,
      "num_input_tokens_seen": 60878872,
      "step": 104855
    },
    {
      "epoch": 15.618111408996128,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.004177289327923414,
      "loss": 0.7839,
      "num_input_tokens_seen": 60881848,
      "step": 104860
    },
    {
      "epoch": 15.618856121537087,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.004175939480087008,
      "loss": 0.7985,
      "num_input_tokens_seen": 60884824,
      "step": 104865
    },
    {
      "epoch": 15.619600834078046,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.004174589815112107,
      "loss": 0.7789,
      "num_input_tokens_seen": 60887416,
      "step": 104870
    },
    {
      "epoch": 15.620345546619005,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.004173240333021509,
      "loss": 0.7886,
      "num_input_tokens_seen": 60890520,
      "step": 104875
    },
    {
      "epoch": 15.621090259159963,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.004171891033838007,
      "loss": 0.8021,
      "num_input_tokens_seen": 60893304,
      "step": 104880
    },
    {
      "epoch": 15.621834971700924,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.004170541917584405,
      "loss": 0.7856,
      "num_input_tokens_seen": 60896344,
      "step": 104885
    },
    {
      "epoch": 15.622579684241883,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.004169192984283486,
      "loss": 0.7793,
      "num_input_tokens_seen": 60899544,
      "step": 104890
    },
    {
      "epoch": 15.623324396782841,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.004167844233958049,
      "loss": 0.7918,
      "num_input_tokens_seen": 60902488,
      "step": 104895
    },
    {
      "epoch": 15.6240691093238,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.004166495666630874,
      "loss": 0.8029,
      "num_input_tokens_seen": 60905400,
      "step": 104900
    },
    {
      "epoch": 15.62481382186476,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0041651472823247436,
      "loss": 0.7981,
      "num_input_tokens_seen": 60908024,
      "step": 104905
    },
    {
      "epoch": 15.62555853440572,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.004163799081062434,
      "loss": 0.8282,
      "num_input_tokens_seen": 60910936,
      "step": 104910
    },
    {
      "epoch": 15.626303246946678,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0041624510628667285,
      "loss": 0.8015,
      "num_input_tokens_seen": 60913624,
      "step": 104915
    },
    {
      "epoch": 15.627047959487637,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.004161103227760399,
      "loss": 0.7915,
      "num_input_tokens_seen": 60916696,
      "step": 104920
    },
    {
      "epoch": 15.627792672028598,
      "grad_norm": 0.041015625,
      "learning_rate": 0.00415975557576621,
      "loss": 0.8035,
      "num_input_tokens_seen": 60919384,
      "step": 104925
    },
    {
      "epoch": 15.628537384569556,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.004158408106906937,
      "loss": 0.7895,
      "num_input_tokens_seen": 60922168,
      "step": 104930
    },
    {
      "epoch": 15.629282097110515,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004157060821205335,
      "loss": 0.8099,
      "num_input_tokens_seen": 60925240,
      "step": 104935
    },
    {
      "epoch": 15.630026809651474,
      "grad_norm": 0.068359375,
      "learning_rate": 0.004155713718684175,
      "loss": 0.8334,
      "num_input_tokens_seen": 60927832,
      "step": 104940
    },
    {
      "epoch": 15.630771522192434,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.004154366799366206,
      "loss": 0.8038,
      "num_input_tokens_seen": 60931032,
      "step": 104945
    },
    {
      "epoch": 15.631516234733393,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.004153020063274191,
      "loss": 0.7892,
      "num_input_tokens_seen": 60933560,
      "step": 104950
    },
    {
      "epoch": 15.632260947274352,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.004151673510430874,
      "loss": 0.7945,
      "num_input_tokens_seen": 60936376,
      "step": 104955
    },
    {
      "epoch": 15.63300565981531,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.00415032714085901,
      "loss": 0.7812,
      "num_input_tokens_seen": 60939192,
      "step": 104960
    },
    {
      "epoch": 15.633750372356271,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.004148980954581344,
      "loss": 0.7952,
      "num_input_tokens_seen": 60942008,
      "step": 104965
    },
    {
      "epoch": 15.63449508489723,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.004147634951620615,
      "loss": 0.7825,
      "num_input_tokens_seen": 60944696,
      "step": 104970
    },
    {
      "epoch": 15.635239797438189,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.004146289131999563,
      "loss": 0.794,
      "num_input_tokens_seen": 60947416,
      "step": 104975
    },
    {
      "epoch": 15.635984509979147,
      "grad_norm": 0.091796875,
      "learning_rate": 0.004144943495740921,
      "loss": 0.801,
      "num_input_tokens_seen": 60950488,
      "step": 104980
    },
    {
      "epoch": 15.636729222520108,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.004143598042867431,
      "loss": 0.808,
      "num_input_tokens_seen": 60953272,
      "step": 104985
    },
    {
      "epoch": 15.637473935061067,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.004142252773401815,
      "loss": 0.8049,
      "num_input_tokens_seen": 60955928,
      "step": 104990
    },
    {
      "epoch": 15.638218647602026,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.004140907687366806,
      "loss": 0.8086,
      "num_input_tokens_seen": 60958840,
      "step": 104995
    },
    {
      "epoch": 15.638963360142984,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.004139562784785124,
      "loss": 0.8138,
      "num_input_tokens_seen": 60961560,
      "step": 105000
    },
    {
      "epoch": 15.639708072683945,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.004138218065679488,
      "loss": 0.805,
      "num_input_tokens_seen": 60964312,
      "step": 105005
    },
    {
      "epoch": 15.640452785224904,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004136873530072621,
      "loss": 0.7958,
      "num_input_tokens_seen": 60967576,
      "step": 105010
    },
    {
      "epoch": 15.641197497765862,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.004135529177987232,
      "loss": 0.7952,
      "num_input_tokens_seen": 60970680,
      "step": 105015
    },
    {
      "epoch": 15.641942210306821,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004134185009446039,
      "loss": 0.7933,
      "num_input_tokens_seen": 60973528,
      "step": 105020
    },
    {
      "epoch": 15.64268692284778,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0041328410244717475,
      "loss": 0.8015,
      "num_input_tokens_seen": 60976376,
      "step": 105025
    },
    {
      "epoch": 15.64343163538874,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.0041314972230870614,
      "loss": 0.7763,
      "num_input_tokens_seen": 60979128,
      "step": 105030
    },
    {
      "epoch": 15.6441763479297,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.004130153605314679,
      "loss": 0.7928,
      "num_input_tokens_seen": 60981848,
      "step": 105035
    },
    {
      "epoch": 15.644921060470658,
      "grad_norm": 0.0625,
      "learning_rate": 0.0041288101711773086,
      "loss": 0.7911,
      "num_input_tokens_seen": 60984856,
      "step": 105040
    },
    {
      "epoch": 15.645665773011618,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.004127466920697641,
      "loss": 0.7853,
      "num_input_tokens_seen": 60987832,
      "step": 105045
    },
    {
      "epoch": 15.646410485552577,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.004126123853898366,
      "loss": 0.8064,
      "num_input_tokens_seen": 60990584,
      "step": 105050
    },
    {
      "epoch": 15.647155198093536,
      "grad_norm": 0.0625,
      "learning_rate": 0.00412478097080218,
      "loss": 0.7853,
      "num_input_tokens_seen": 60993304,
      "step": 105055
    },
    {
      "epoch": 15.647899910634495,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.004123438271431762,
      "loss": 0.7982,
      "num_input_tokens_seen": 60996504,
      "step": 105060
    },
    {
      "epoch": 15.648644623175453,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0041220957558098055,
      "loss": 0.8046,
      "num_input_tokens_seen": 60999544,
      "step": 105065
    },
    {
      "epoch": 15.649389335716414,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00412075342395898,
      "loss": 0.7993,
      "num_input_tokens_seen": 61002232,
      "step": 105070
    },
    {
      "epoch": 15.650134048257373,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.004119411275901972,
      "loss": 0.8049,
      "num_input_tokens_seen": 61004888,
      "step": 105075
    },
    {
      "epoch": 15.650878760798332,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.004118069311661455,
      "loss": 0.8041,
      "num_input_tokens_seen": 61007672,
      "step": 105080
    },
    {
      "epoch": 15.65162347333929,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.0041167275312600895,
      "loss": 0.8023,
      "num_input_tokens_seen": 61010680,
      "step": 105085
    },
    {
      "epoch": 15.65236818588025,
      "grad_norm": 0.052734375,
      "learning_rate": 0.004115385934720556,
      "loss": 0.7719,
      "num_input_tokens_seen": 61014008,
      "step": 105090
    },
    {
      "epoch": 15.65311289842121,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.004114044522065516,
      "loss": 0.7623,
      "num_input_tokens_seen": 61016920,
      "step": 105095
    },
    {
      "epoch": 15.653857610962168,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0041127032933176285,
      "loss": 0.8099,
      "num_input_tokens_seen": 61020120,
      "step": 105100
    },
    {
      "epoch": 15.654602323503127,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0041113622484995485,
      "loss": 0.8586,
      "num_input_tokens_seen": 61023064,
      "step": 105105
    },
    {
      "epoch": 15.655347036044088,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.0041100213876339425,
      "loss": 0.788,
      "num_input_tokens_seen": 61026104,
      "step": 105110
    },
    {
      "epoch": 15.656091748585046,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.004108680710743453,
      "loss": 0.8038,
      "num_input_tokens_seen": 61029112,
      "step": 105115
    },
    {
      "epoch": 15.656836461126005,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.004107340217850738,
      "loss": 0.7875,
      "num_input_tokens_seen": 61031896,
      "step": 105120
    },
    {
      "epoch": 15.657581173666964,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.004105999908978437,
      "loss": 0.8238,
      "num_input_tokens_seen": 61034872,
      "step": 105125
    },
    {
      "epoch": 15.658325886207924,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.004104659784149193,
      "loss": 0.818,
      "num_input_tokens_seen": 61037528,
      "step": 105130
    },
    {
      "epoch": 15.659070598748883,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.004103319843385651,
      "loss": 0.7863,
      "num_input_tokens_seen": 61040376,
      "step": 105135
    },
    {
      "epoch": 15.659815311289842,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0041019800867104425,
      "loss": 0.8034,
      "num_input_tokens_seen": 61043128,
      "step": 105140
    },
    {
      "epoch": 15.6605600238308,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.004100640514146208,
      "loss": 0.7845,
      "num_input_tokens_seen": 61046424,
      "step": 105145
    },
    {
      "epoch": 15.661304736371761,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0040993011257155695,
      "loss": 0.8173,
      "num_input_tokens_seen": 61049688,
      "step": 105150
    },
    {
      "epoch": 15.66204944891272,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.004097961921441164,
      "loss": 0.7912,
      "num_input_tokens_seen": 61052280,
      "step": 105155
    },
    {
      "epoch": 15.662794161453679,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.004096622901345612,
      "loss": 0.7953,
      "num_input_tokens_seen": 61055288,
      "step": 105160
    },
    {
      "epoch": 15.663538873994638,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.004095284065451535,
      "loss": 0.7944,
      "num_input_tokens_seen": 61058072,
      "step": 105165
    },
    {
      "epoch": 15.664283586535598,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00409394541378155,
      "loss": 0.8182,
      "num_input_tokens_seen": 61060920,
      "step": 105170
    },
    {
      "epoch": 15.665028299076557,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.004092606946358268,
      "loss": 0.7917,
      "num_input_tokens_seen": 61063768,
      "step": 105175
    },
    {
      "epoch": 15.665773011617516,
      "grad_norm": 0.052734375,
      "learning_rate": 0.004091268663204311,
      "loss": 0.8032,
      "num_input_tokens_seen": 61066808,
      "step": 105180
    },
    {
      "epoch": 15.666517724158474,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.004089930564342277,
      "loss": 0.7923,
      "num_input_tokens_seen": 61069976,
      "step": 105185
    },
    {
      "epoch": 15.667262436699435,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.004088592649794783,
      "loss": 0.8007,
      "num_input_tokens_seen": 61072792,
      "step": 105190
    },
    {
      "epoch": 15.668007149240394,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.004087254919584423,
      "loss": 0.792,
      "num_input_tokens_seen": 61075896,
      "step": 105195
    },
    {
      "epoch": 15.668751861781352,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.004085917373733804,
      "loss": 0.8331,
      "num_input_tokens_seen": 61078456,
      "step": 105200
    },
    {
      "epoch": 15.669496574322311,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004084580012265516,
      "loss": 0.7936,
      "num_input_tokens_seen": 61081272,
      "step": 105205
    },
    {
      "epoch": 15.67024128686327,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.004083242835202153,
      "loss": 0.8389,
      "num_input_tokens_seen": 61084088,
      "step": 105210
    },
    {
      "epoch": 15.67098599940423,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.00408190584256631,
      "loss": 0.7985,
      "num_input_tokens_seen": 61087032,
      "step": 105215
    },
    {
      "epoch": 15.67173071194519,
      "grad_norm": 0.025390625,
      "learning_rate": 0.004080569034380567,
      "loss": 0.8059,
      "num_input_tokens_seen": 61089752,
      "step": 105220
    },
    {
      "epoch": 15.672475424486148,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.004079232410667516,
      "loss": 0.7897,
      "num_input_tokens_seen": 61092984,
      "step": 105225
    },
    {
      "epoch": 15.673220137027108,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.004077895971449736,
      "loss": 0.8023,
      "num_input_tokens_seen": 61095864,
      "step": 105230
    },
    {
      "epoch": 15.673964849568067,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.004076559716749802,
      "loss": 0.8032,
      "num_input_tokens_seen": 61098776,
      "step": 105235
    },
    {
      "epoch": 15.674709562109026,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.004075223646590289,
      "loss": 0.7862,
      "num_input_tokens_seen": 61101752,
      "step": 105240
    },
    {
      "epoch": 15.675454274649985,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.004073887760993765,
      "loss": 0.8103,
      "num_input_tokens_seen": 61104792,
      "step": 105245
    },
    {
      "epoch": 15.676198987190944,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.004072552059982805,
      "loss": 0.8075,
      "num_input_tokens_seen": 61107512,
      "step": 105250
    },
    {
      "epoch": 15.676943699731904,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.00407121654357997,
      "loss": 0.8098,
      "num_input_tokens_seen": 61110328,
      "step": 105255
    },
    {
      "epoch": 15.677688412272863,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0040698812118078265,
      "loss": 0.7993,
      "num_input_tokens_seen": 61113528,
      "step": 105260
    },
    {
      "epoch": 15.678433124813822,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.004068546064688928,
      "loss": 0.7919,
      "num_input_tokens_seen": 61116248,
      "step": 105265
    },
    {
      "epoch": 15.67917783735478,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004067211102245837,
      "loss": 0.7909,
      "num_input_tokens_seen": 61119032,
      "step": 105270
    },
    {
      "epoch": 15.67992254989574,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0040658763245011,
      "loss": 0.8018,
      "num_input_tokens_seen": 61121912,
      "step": 105275
    },
    {
      "epoch": 15.6806672624367,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0040645417314772725,
      "loss": 0.7836,
      "num_input_tokens_seen": 61124664,
      "step": 105280
    },
    {
      "epoch": 15.681411974977658,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.004063207323196899,
      "loss": 0.7993,
      "num_input_tokens_seen": 61127608,
      "step": 105285
    },
    {
      "epoch": 15.682156687518617,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.004061873099682516,
      "loss": 0.8108,
      "num_input_tokens_seen": 61130424,
      "step": 105290
    },
    {
      "epoch": 15.682901400059578,
      "grad_norm": 0.32421875,
      "learning_rate": 0.0040605390609566765,
      "loss": 0.8231,
      "num_input_tokens_seen": 61133304,
      "step": 105295
    },
    {
      "epoch": 15.683646112600536,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.004059205207041909,
      "loss": 0.7863,
      "num_input_tokens_seen": 61135960,
      "step": 105300
    },
    {
      "epoch": 15.684390825141495,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.004057871537960753,
      "loss": 0.7839,
      "num_input_tokens_seen": 61138680,
      "step": 105305
    },
    {
      "epoch": 15.685135537682454,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.004056538053735729,
      "loss": 0.835,
      "num_input_tokens_seen": 61141336,
      "step": 105310
    },
    {
      "epoch": 15.685880250223414,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004055204754389378,
      "loss": 0.8021,
      "num_input_tokens_seen": 61144280,
      "step": 105315
    },
    {
      "epoch": 15.686624962764373,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.004053871639944219,
      "loss": 0.8101,
      "num_input_tokens_seen": 61147128,
      "step": 105320
    },
    {
      "epoch": 15.687369675305332,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.004052538710422767,
      "loss": 0.8095,
      "num_input_tokens_seen": 61150008,
      "step": 105325
    },
    {
      "epoch": 15.68811438784629,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.004051205965847554,
      "loss": 0.7949,
      "num_input_tokens_seen": 61153112,
      "step": 105330
    },
    {
      "epoch": 15.688859100387251,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.004049873406241083,
      "loss": 0.8186,
      "num_input_tokens_seen": 61156056,
      "step": 105335
    },
    {
      "epoch": 15.68960381292821,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.004048541031625876,
      "loss": 0.8092,
      "num_input_tokens_seen": 61159352,
      "step": 105340
    },
    {
      "epoch": 15.690348525469169,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.004047208842024433,
      "loss": 0.8004,
      "num_input_tokens_seen": 61162168,
      "step": 105345
    },
    {
      "epoch": 15.691093238010128,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.004045876837459268,
      "loss": 0.7825,
      "num_input_tokens_seen": 61165112,
      "step": 105350
    },
    {
      "epoch": 15.691837950551088,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.004044545017952883,
      "loss": 0.7955,
      "num_input_tokens_seen": 61167864,
      "step": 105355
    },
    {
      "epoch": 15.692582663092047,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.004043213383527773,
      "loss": 0.7849,
      "num_input_tokens_seen": 61170936,
      "step": 105360
    },
    {
      "epoch": 15.693327375633006,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.004041881934206434,
      "loss": 0.8108,
      "num_input_tokens_seen": 61173592,
      "step": 105365
    },
    {
      "epoch": 15.694072088173964,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.004040550670011366,
      "loss": 0.7969,
      "num_input_tokens_seen": 61176248,
      "step": 105370
    },
    {
      "epoch": 15.694816800714925,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.0040392195909650565,
      "loss": 0.7866,
      "num_input_tokens_seen": 61179064,
      "step": 105375
    },
    {
      "epoch": 15.695561513255884,
      "grad_norm": 0.033203125,
      "learning_rate": 0.004037888697089986,
      "loss": 0.8213,
      "num_input_tokens_seen": 61181592,
      "step": 105380
    },
    {
      "epoch": 15.696306225796842,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0040365579884086514,
      "loss": 0.7888,
      "num_input_tokens_seen": 61184088,
      "step": 105385
    },
    {
      "epoch": 15.697050938337801,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.004035227464943521,
      "loss": 0.7813,
      "num_input_tokens_seen": 61186872,
      "step": 105390
    },
    {
      "epoch": 15.69779565087876,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.004033897126717083,
      "loss": 0.804,
      "num_input_tokens_seen": 61190136,
      "step": 105395
    },
    {
      "epoch": 15.69854036341972,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.004032566973751806,
      "loss": 0.8017,
      "num_input_tokens_seen": 61193048,
      "step": 105400
    },
    {
      "epoch": 15.69928507596068,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.004031237006070162,
      "loss": 0.7975,
      "num_input_tokens_seen": 61195992,
      "step": 105405
    },
    {
      "epoch": 15.700029788501638,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.004029907223694623,
      "loss": 0.7964,
      "num_input_tokens_seen": 61198808,
      "step": 105410
    },
    {
      "epoch": 15.700774501042597,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004028577626647647,
      "loss": 0.794,
      "num_input_tokens_seen": 61201880,
      "step": 105415
    },
    {
      "epoch": 15.701519213583557,
      "grad_norm": 0.046875,
      "learning_rate": 0.004027248214951706,
      "loss": 0.8024,
      "num_input_tokens_seen": 61205208,
      "step": 105420
    },
    {
      "epoch": 15.702263926124516,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0040259189886292564,
      "loss": 0.7769,
      "num_input_tokens_seen": 61207896,
      "step": 105425
    },
    {
      "epoch": 15.703008638665475,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.004024589947702751,
      "loss": 0.8083,
      "num_input_tokens_seen": 61210744,
      "step": 105430
    },
    {
      "epoch": 15.703753351206434,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0040232610921946375,
      "loss": 0.8216,
      "num_input_tokens_seen": 61213720,
      "step": 105435
    },
    {
      "epoch": 15.704498063747394,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.004021932422127377,
      "loss": 0.7886,
      "num_input_tokens_seen": 61216408,
      "step": 105440
    },
    {
      "epoch": 15.705242776288353,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0040206039375234115,
      "loss": 0.7951,
      "num_input_tokens_seen": 61219192,
      "step": 105445
    },
    {
      "epoch": 15.705987488829312,
      "grad_norm": 0.283203125,
      "learning_rate": 0.004019275638405178,
      "loss": 0.8176,
      "num_input_tokens_seen": 61222488,
      "step": 105450
    },
    {
      "epoch": 15.70673220137027,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.004017947524795126,
      "loss": 0.7853,
      "num_input_tokens_seen": 61225656,
      "step": 105455
    },
    {
      "epoch": 15.707476913911231,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.004016619596715685,
      "loss": 0.8056,
      "num_input_tokens_seen": 61228664,
      "step": 105460
    },
    {
      "epoch": 15.70822162645219,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.004015291854189298,
      "loss": 0.8035,
      "num_input_tokens_seen": 61231384,
      "step": 105465
    },
    {
      "epoch": 15.708966338993148,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.004013964297238385,
      "loss": 0.7882,
      "num_input_tokens_seen": 61234296,
      "step": 105470
    },
    {
      "epoch": 15.709711051534107,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.004012636925885386,
      "loss": 0.8002,
      "num_input_tokens_seen": 61237208,
      "step": 105475
    },
    {
      "epoch": 15.710455764075068,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.004011309740152717,
      "loss": 0.8043,
      "num_input_tokens_seen": 61240312,
      "step": 105480
    },
    {
      "epoch": 15.711200476616026,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.004009982740062798,
      "loss": 0.7918,
      "num_input_tokens_seen": 61243288,
      "step": 105485
    },
    {
      "epoch": 15.711945189156985,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004008655925638056,
      "loss": 0.7946,
      "num_input_tokens_seen": 61246392,
      "step": 105490
    },
    {
      "epoch": 15.712689901697944,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0040073292969009,
      "loss": 0.7836,
      "num_input_tokens_seen": 61249624,
      "step": 105495
    },
    {
      "epoch": 15.713434614238905,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.004006002853873744,
      "loss": 0.8306,
      "num_input_tokens_seen": 61252568,
      "step": 105500
    },
    {
      "epoch": 15.714179326779863,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.004004676596578991,
      "loss": 0.8337,
      "num_input_tokens_seen": 61255672,
      "step": 105505
    },
    {
      "epoch": 15.714924039320822,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.004003350525039057,
      "loss": 0.8119,
      "num_input_tokens_seen": 61258360,
      "step": 105510
    },
    {
      "epoch": 15.71566875186178,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.004002024639276333,
      "loss": 0.7984,
      "num_input_tokens_seen": 61261336,
      "step": 105515
    },
    {
      "epoch": 15.716413464402741,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.00400069893931323,
      "loss": 0.8008,
      "num_input_tokens_seen": 61264184,
      "step": 105520
    },
    {
      "epoch": 15.7171581769437,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00399937342517214,
      "loss": 0.8015,
      "num_input_tokens_seen": 61267128,
      "step": 105525
    },
    {
      "epoch": 15.717902889484659,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.003998048096875452,
      "loss": 0.8015,
      "num_input_tokens_seen": 61270008,
      "step": 105530
    },
    {
      "epoch": 15.718647602025618,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.003996722954445562,
      "loss": 0.7928,
      "num_input_tokens_seen": 61273272,
      "step": 105535
    },
    {
      "epoch": 15.719392314566576,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.003995397997904851,
      "loss": 0.8042,
      "num_input_tokens_seen": 61276024,
      "step": 105540
    },
    {
      "epoch": 15.720137027107537,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.003994073227275709,
      "loss": 0.7893,
      "num_input_tokens_seen": 61279128,
      "step": 105545
    },
    {
      "epoch": 15.720881739648496,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.003992748642580511,
      "loss": 0.7933,
      "num_input_tokens_seen": 61281912,
      "step": 105550
    },
    {
      "epoch": 15.721626452189454,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.003991424243841642,
      "loss": 0.7897,
      "num_input_tokens_seen": 61284984,
      "step": 105555
    },
    {
      "epoch": 15.722371164730415,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.0039901000310814715,
      "loss": 0.8044,
      "num_input_tokens_seen": 61287928,
      "step": 105560
    },
    {
      "epoch": 15.723115877271374,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.003988776004322371,
      "loss": 0.7886,
      "num_input_tokens_seen": 61290584,
      "step": 105565
    },
    {
      "epoch": 15.723860589812332,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.003987452163586708,
      "loss": 0.7834,
      "num_input_tokens_seen": 61293176,
      "step": 105570
    },
    {
      "epoch": 15.724605302353291,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.003986128508896844,
      "loss": 0.8051,
      "num_input_tokens_seen": 61296280,
      "step": 105575
    },
    {
      "epoch": 15.72535001489425,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.00398480504027515,
      "loss": 0.7955,
      "num_input_tokens_seen": 61299352,
      "step": 105580
    },
    {
      "epoch": 15.72609472743521,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.003983481757743975,
      "loss": 0.7843,
      "num_input_tokens_seen": 61302104,
      "step": 105585
    },
    {
      "epoch": 15.72683943997617,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.003982158661325682,
      "loss": 0.7874,
      "num_input_tokens_seen": 61305144,
      "step": 105590
    },
    {
      "epoch": 15.727584152517128,
      "grad_norm": 0.046875,
      "learning_rate": 0.003980835751042618,
      "loss": 0.7981,
      "num_input_tokens_seen": 61308280,
      "step": 105595
    },
    {
      "epoch": 15.728328865058087,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.003979513026917138,
      "loss": 0.7919,
      "num_input_tokens_seen": 61311128,
      "step": 105600
    },
    {
      "epoch": 15.729073577599047,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.003978190488971585,
      "loss": 0.7907,
      "num_input_tokens_seen": 61314104,
      "step": 105605
    },
    {
      "epoch": 15.729818290140006,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.003976868137228296,
      "loss": 0.7994,
      "num_input_tokens_seen": 61317208,
      "step": 105610
    },
    {
      "epoch": 15.730563002680965,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.00397554597170962,
      "loss": 0.7909,
      "num_input_tokens_seen": 61319928,
      "step": 105615
    },
    {
      "epoch": 15.731307715221924,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0039742239924378895,
      "loss": 0.8051,
      "num_input_tokens_seen": 61322712,
      "step": 105620
    },
    {
      "epoch": 15.732052427762884,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.00397290219943544,
      "loss": 0.8038,
      "num_input_tokens_seen": 61325688,
      "step": 105625
    },
    {
      "epoch": 15.732797140303843,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.003971580592724601,
      "loss": 0.8022,
      "num_input_tokens_seen": 61328408,
      "step": 105630
    },
    {
      "epoch": 15.733541852844802,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.003970259172327699,
      "loss": 0.7914,
      "num_input_tokens_seen": 61330872,
      "step": 105635
    },
    {
      "epoch": 15.73428656538576,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.003968937938267055,
      "loss": 0.7993,
      "num_input_tokens_seen": 61333752,
      "step": 105640
    },
    {
      "epoch": 15.735031277926721,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.003967616890564997,
      "loss": 0.7999,
      "num_input_tokens_seen": 61336600,
      "step": 105645
    },
    {
      "epoch": 15.73577599046768,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.003966296029243838,
      "loss": 0.7996,
      "num_input_tokens_seen": 61339704,
      "step": 105650
    },
    {
      "epoch": 15.736520703008638,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.003964975354325888,
      "loss": 0.784,
      "num_input_tokens_seen": 61342296,
      "step": 105655
    },
    {
      "epoch": 15.737265415549597,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0039636548658334705,
      "loss": 0.7885,
      "num_input_tokens_seen": 61345304,
      "step": 105660
    },
    {
      "epoch": 15.738010128090558,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0039623345637888815,
      "loss": 0.7848,
      "num_input_tokens_seen": 61348120,
      "step": 105665
    },
    {
      "epoch": 15.738754840631517,
      "grad_norm": 0.03515625,
      "learning_rate": 0.003961014448214437,
      "loss": 0.7993,
      "num_input_tokens_seen": 61351000,
      "step": 105670
    },
    {
      "epoch": 15.739499553172475,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.003959694519132429,
      "loss": 0.7916,
      "num_input_tokens_seen": 61353720,
      "step": 105675
    },
    {
      "epoch": 15.740244265713434,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.003958374776565165,
      "loss": 0.7768,
      "num_input_tokens_seen": 61356728,
      "step": 105680
    },
    {
      "epoch": 15.740988978254395,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0039570552205349385,
      "loss": 0.7918,
      "num_input_tokens_seen": 61359768,
      "step": 105685
    },
    {
      "epoch": 15.741733690795353,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.00395573585106404,
      "loss": 0.8055,
      "num_input_tokens_seen": 61362808,
      "step": 105690
    },
    {
      "epoch": 15.742478403336312,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.003954416668174754,
      "loss": 0.8121,
      "num_input_tokens_seen": 61365432,
      "step": 105695
    },
    {
      "epoch": 15.74322311587727,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0039530976718893775,
      "loss": 0.7831,
      "num_input_tokens_seen": 61368536,
      "step": 105700
    },
    {
      "epoch": 15.743967828418231,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.003951778862230189,
      "loss": 0.8055,
      "num_input_tokens_seen": 61371608,
      "step": 105705
    },
    {
      "epoch": 15.74471254095919,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.003950460239219461,
      "loss": 0.7943,
      "num_input_tokens_seen": 61374520,
      "step": 105710
    },
    {
      "epoch": 15.745457253500149,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.003949141802879483,
      "loss": 0.8016,
      "num_input_tokens_seen": 61377176,
      "step": 105715
    },
    {
      "epoch": 15.746201966041108,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.003947823553232518,
      "loss": 0.7977,
      "num_input_tokens_seen": 61379992,
      "step": 105720
    },
    {
      "epoch": 15.746946678582066,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.003946505490300845,
      "loss": 0.7816,
      "num_input_tokens_seen": 61382872,
      "step": 105725
    },
    {
      "epoch": 15.747691391123027,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.003945187614106728,
      "loss": 0.7858,
      "num_input_tokens_seen": 61385752,
      "step": 105730
    },
    {
      "epoch": 15.748436103663986,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.003943869924672428,
      "loss": 0.7785,
      "num_input_tokens_seen": 61388984,
      "step": 105735
    },
    {
      "epoch": 15.749180816204944,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00394255242202021,
      "loss": 0.7936,
      "num_input_tokens_seen": 61391832,
      "step": 105740
    },
    {
      "epoch": 15.749925528745905,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.003941235106172327,
      "loss": 0.8321,
      "num_input_tokens_seen": 61394648,
      "step": 105745
    },
    {
      "epoch": 15.750670241286864,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.003939917977151043,
      "loss": 0.8076,
      "num_input_tokens_seen": 61397528,
      "step": 105750
    },
    {
      "epoch": 15.751414953827823,
      "grad_norm": 0.072265625,
      "learning_rate": 0.003938601034978604,
      "loss": 0.7989,
      "num_input_tokens_seen": 61400536,
      "step": 105755
    },
    {
      "epoch": 15.752159666368781,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.003937284279677256,
      "loss": 0.8033,
      "num_input_tokens_seen": 61403416,
      "step": 105760
    },
    {
      "epoch": 15.75290437890974,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.003935967711269249,
      "loss": 0.8187,
      "num_input_tokens_seen": 61406104,
      "step": 105765
    },
    {
      "epoch": 15.7536490914507,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.003934651329776816,
      "loss": 0.7944,
      "num_input_tokens_seen": 61408888,
      "step": 105770
    },
    {
      "epoch": 15.75439380399166,
      "grad_norm": 0.0230712890625,
      "learning_rate": 0.003933335135222208,
      "loss": 0.793,
      "num_input_tokens_seen": 61411896,
      "step": 105775
    },
    {
      "epoch": 15.755138516532618,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.00393201912762765,
      "loss": 0.7874,
      "num_input_tokens_seen": 61414968,
      "step": 105780
    },
    {
      "epoch": 15.755883229073577,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.003930703307015384,
      "loss": 0.7889,
      "num_input_tokens_seen": 61417688,
      "step": 105785
    },
    {
      "epoch": 15.756627941614537,
      "grad_norm": 0.03125,
      "learning_rate": 0.003929387673407633,
      "loss": 0.7805,
      "num_input_tokens_seen": 61420408,
      "step": 105790
    },
    {
      "epoch": 15.757372654155496,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.003928072226826628,
      "loss": 0.809,
      "num_input_tokens_seen": 61423288,
      "step": 105795
    },
    {
      "epoch": 15.758117366696455,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0039267569672945864,
      "loss": 0.8191,
      "num_input_tokens_seen": 61426040,
      "step": 105800
    },
    {
      "epoch": 15.758862079237414,
      "grad_norm": 0.03515625,
      "learning_rate": 0.003925441894833736,
      "loss": 0.7906,
      "num_input_tokens_seen": 61428952,
      "step": 105805
    },
    {
      "epoch": 15.759606791778374,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.003924127009466288,
      "loss": 0.8065,
      "num_input_tokens_seen": 61432280,
      "step": 105810
    },
    {
      "epoch": 15.760351504319333,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.003922812311214452,
      "loss": 0.7833,
      "num_input_tokens_seen": 61434744,
      "step": 105815
    },
    {
      "epoch": 15.761096216860292,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.00392149780010045,
      "loss": 0.7901,
      "num_input_tokens_seen": 61437496,
      "step": 105820
    },
    {
      "epoch": 15.76184092940125,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.003920183476146482,
      "loss": 0.7978,
      "num_input_tokens_seen": 61440312,
      "step": 105825
    },
    {
      "epoch": 15.762585641942211,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.003918869339374753,
      "loss": 0.7708,
      "num_input_tokens_seen": 61443800,
      "step": 105830
    },
    {
      "epoch": 15.76333035448317,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.003917555389807462,
      "loss": 0.7969,
      "num_input_tokens_seen": 61446808,
      "step": 105835
    },
    {
      "epoch": 15.764075067024129,
      "grad_norm": 0.068359375,
      "learning_rate": 0.003916241627466811,
      "loss": 0.8038,
      "num_input_tokens_seen": 61449848,
      "step": 105840
    },
    {
      "epoch": 15.764819779565087,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.003914928052374994,
      "loss": 0.7866,
      "num_input_tokens_seen": 61452440,
      "step": 105845
    },
    {
      "epoch": 15.765564492106048,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.003913614664554195,
      "loss": 0.7874,
      "num_input_tokens_seen": 61455160,
      "step": 105850
    },
    {
      "epoch": 15.766309204647007,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.003912301464026614,
      "loss": 0.7892,
      "num_input_tokens_seen": 61458072,
      "step": 105855
    },
    {
      "epoch": 15.767053917187965,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0039109884508144255,
      "loss": 0.7951,
      "num_input_tokens_seen": 61461048,
      "step": 105860
    },
    {
      "epoch": 15.767798629728924,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0039096756249398235,
      "loss": 0.7906,
      "num_input_tokens_seen": 61464248,
      "step": 105865
    },
    {
      "epoch": 15.768543342269885,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.003908362986424973,
      "loss": 0.7942,
      "num_input_tokens_seen": 61467256,
      "step": 105870
    },
    {
      "epoch": 15.769288054810843,
      "grad_norm": 0.037109375,
      "learning_rate": 0.003907050535292064,
      "loss": 0.8005,
      "num_input_tokens_seen": 61469720,
      "step": 105875
    },
    {
      "epoch": 15.770032767351802,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.003905738271563256,
      "loss": 0.8115,
      "num_input_tokens_seen": 61472632,
      "step": 105880
    },
    {
      "epoch": 15.770777479892761,
      "grad_norm": 0.0228271484375,
      "learning_rate": 0.0039044261952607276,
      "loss": 0.8245,
      "num_input_tokens_seen": 61475256,
      "step": 105885
    },
    {
      "epoch": 15.771522192433721,
      "grad_norm": 0.033203125,
      "learning_rate": 0.003903114306406643,
      "loss": 0.8157,
      "num_input_tokens_seen": 61477944,
      "step": 105890
    },
    {
      "epoch": 15.77226690497468,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.0039018026050231624,
      "loss": 0.8036,
      "num_input_tokens_seen": 61480952,
      "step": 105895
    },
    {
      "epoch": 15.773011617515639,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.003900491091132448,
      "loss": 0.7958,
      "num_input_tokens_seen": 61483800,
      "step": 105900
    },
    {
      "epoch": 15.773756330056598,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.003899179764756652,
      "loss": 0.8117,
      "num_input_tokens_seen": 61486744,
      "step": 105905
    },
    {
      "epoch": 15.774501042597556,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0038978686259179356,
      "loss": 0.7942,
      "num_input_tokens_seen": 61489912,
      "step": 105910
    },
    {
      "epoch": 15.775245755138517,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.003896557674638439,
      "loss": 0.7878,
      "num_input_tokens_seen": 61492760,
      "step": 105915
    },
    {
      "epoch": 15.775990467679476,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0038952469109403208,
      "loss": 0.7826,
      "num_input_tokens_seen": 61495704,
      "step": 105920
    },
    {
      "epoch": 15.776735180220435,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0038939363348457203,
      "loss": 0.8001,
      "num_input_tokens_seen": 61498680,
      "step": 105925
    },
    {
      "epoch": 15.777479892761393,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0038926259463767736,
      "loss": 0.7934,
      "num_input_tokens_seen": 61501656,
      "step": 105930
    },
    {
      "epoch": 15.778224605302354,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0038913157455556257,
      "loss": 0.779,
      "num_input_tokens_seen": 61504600,
      "step": 105935
    },
    {
      "epoch": 15.778969317843313,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.003890005732404404,
      "loss": 0.7747,
      "num_input_tokens_seen": 61507352,
      "step": 105940
    },
    {
      "epoch": 15.779714030384271,
      "grad_norm": 0.0296630859375,
      "learning_rate": 0.0038886959069452474,
      "loss": 0.793,
      "num_input_tokens_seen": 61510072,
      "step": 105945
    },
    {
      "epoch": 15.78045874292523,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0038873862692002767,
      "loss": 0.7921,
      "num_input_tokens_seen": 61513144,
      "step": 105950
    },
    {
      "epoch": 15.78120345546619,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.003886076819191626,
      "loss": 0.8031,
      "num_input_tokens_seen": 61516024,
      "step": 105955
    },
    {
      "epoch": 15.78194816800715,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.003884767556941411,
      "loss": 0.8057,
      "num_input_tokens_seen": 61519320,
      "step": 105960
    },
    {
      "epoch": 15.782692880548108,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00388345848247175,
      "loss": 0.8064,
      "num_input_tokens_seen": 61522424,
      "step": 105965
    },
    {
      "epoch": 15.783437593089067,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0038821495958047607,
      "loss": 0.7912,
      "num_input_tokens_seen": 61525432,
      "step": 105970
    },
    {
      "epoch": 15.784182305630027,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0038808408969625507,
      "loss": 0.7978,
      "num_input_tokens_seen": 61528088,
      "step": 105975
    },
    {
      "epoch": 15.784927018170986,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0038795323859672354,
      "loss": 0.7833,
      "num_input_tokens_seen": 61530936,
      "step": 105980
    },
    {
      "epoch": 15.785671730711945,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.003878224062840915,
      "loss": 0.7998,
      "num_input_tokens_seen": 61533752,
      "step": 105985
    },
    {
      "epoch": 15.786416443252904,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0038769159276056996,
      "loss": 0.7776,
      "num_input_tokens_seen": 61536664,
      "step": 105990
    },
    {
      "epoch": 15.787161155793864,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.003875607980283679,
      "loss": 0.8168,
      "num_input_tokens_seen": 61539736,
      "step": 105995
    },
    {
      "epoch": 15.787905868334823,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0038743002208969607,
      "loss": 0.8252,
      "num_input_tokens_seen": 61542584,
      "step": 106000
    },
    {
      "epoch": 15.788650580875782,
      "grad_norm": 0.033203125,
      "learning_rate": 0.003872992649467632,
      "loss": 0.7692,
      "num_input_tokens_seen": 61545304,
      "step": 106005
    },
    {
      "epoch": 15.78939529341674,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00387168526601778,
      "loss": 0.7868,
      "num_input_tokens_seen": 61548376,
      "step": 106010
    },
    {
      "epoch": 15.790140005957701,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0038703780705694993,
      "loss": 0.7821,
      "num_input_tokens_seen": 61551256,
      "step": 106015
    },
    {
      "epoch": 15.79088471849866,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0038690710631448686,
      "loss": 0.7929,
      "num_input_tokens_seen": 61554424,
      "step": 106020
    },
    {
      "epoch": 15.791629431039619,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0038677642437659654,
      "loss": 0.8174,
      "num_input_tokens_seen": 61557240,
      "step": 106025
    },
    {
      "epoch": 15.792374143580577,
      "grad_norm": 0.0625,
      "learning_rate": 0.003866457612454875,
      "loss": 0.7956,
      "num_input_tokens_seen": 61560248,
      "step": 106030
    },
    {
      "epoch": 15.793118856121538,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.003865151169233668,
      "loss": 0.7963,
      "num_input_tokens_seen": 61562968,
      "step": 106035
    },
    {
      "epoch": 15.793863568662497,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0038638449141244107,
      "loss": 0.7986,
      "num_input_tokens_seen": 61565784,
      "step": 106040
    },
    {
      "epoch": 15.794608281203455,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00386253884714918,
      "loss": 0.7701,
      "num_input_tokens_seen": 61569016,
      "step": 106045
    },
    {
      "epoch": 15.795352993744414,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0038612329683300364,
      "loss": 0.7872,
      "num_input_tokens_seen": 61571704,
      "step": 106050
    },
    {
      "epoch": 15.796097706285373,
      "grad_norm": 0.03125,
      "learning_rate": 0.0038599272776890345,
      "loss": 0.804,
      "num_input_tokens_seen": 61574744,
      "step": 106055
    },
    {
      "epoch": 15.796842418826333,
      "grad_norm": 0.072265625,
      "learning_rate": 0.003858621775248245,
      "loss": 0.7979,
      "num_input_tokens_seen": 61577720,
      "step": 106060
    },
    {
      "epoch": 15.797587131367292,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0038573164610297126,
      "loss": 0.7973,
      "num_input_tokens_seen": 61581080,
      "step": 106065
    },
    {
      "epoch": 15.798331843908251,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0038560113350554976,
      "loss": 0.8024,
      "num_input_tokens_seen": 61583768,
      "step": 106070
    },
    {
      "epoch": 15.799076556449211,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0038547063973476394,
      "loss": 0.7801,
      "num_input_tokens_seen": 61586296,
      "step": 106075
    },
    {
      "epoch": 15.79982126899017,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.003853401647928195,
      "loss": 0.7689,
      "num_input_tokens_seen": 61589144,
      "step": 106080
    },
    {
      "epoch": 15.800565981531129,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.003852097086819201,
      "loss": 0.8083,
      "num_input_tokens_seen": 61592088,
      "step": 106085
    },
    {
      "epoch": 15.801310694072088,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0038507927140426948,
      "loss": 0.8139,
      "num_input_tokens_seen": 61595032,
      "step": 106090
    },
    {
      "epoch": 15.802055406613047,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0038494885296207156,
      "loss": 0.7979,
      "num_input_tokens_seen": 61597848,
      "step": 106095
    },
    {
      "epoch": 15.802800119154007,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0038481845335752896,
      "loss": 0.7919,
      "num_input_tokens_seen": 61600824,
      "step": 106100
    },
    {
      "epoch": 15.803544831694966,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0038468807259284562,
      "loss": 0.803,
      "num_input_tokens_seen": 61603576,
      "step": 106105
    },
    {
      "epoch": 15.804289544235925,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.003845577106702234,
      "loss": 0.8059,
      "num_input_tokens_seen": 61606616,
      "step": 106110
    },
    {
      "epoch": 15.805034256776883,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0038442736759186536,
      "loss": 0.8102,
      "num_input_tokens_seen": 61609400,
      "step": 106115
    },
    {
      "epoch": 15.805778969317844,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0038429704335997267,
      "loss": 0.7823,
      "num_input_tokens_seen": 61612184,
      "step": 106120
    },
    {
      "epoch": 15.806523681858803,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.003841667379767479,
      "loss": 0.7858,
      "num_input_tokens_seen": 61614872,
      "step": 106125
    },
    {
      "epoch": 15.807268394399761,
      "grad_norm": 0.0546875,
      "learning_rate": 0.003840364514443921,
      "loss": 0.8057,
      "num_input_tokens_seen": 61617880,
      "step": 106130
    },
    {
      "epoch": 15.80801310694072,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0038390618376510565,
      "loss": 0.8,
      "num_input_tokens_seen": 61621176,
      "step": 106135
    },
    {
      "epoch": 15.80875781948168,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.003837759349410904,
      "loss": 0.7778,
      "num_input_tokens_seen": 61624120,
      "step": 106140
    },
    {
      "epoch": 15.80950253202264,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.003836457049745459,
      "loss": 0.7772,
      "num_input_tokens_seen": 61627064,
      "step": 106145
    },
    {
      "epoch": 15.810247244563598,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.003835154938676729,
      "loss": 0.7964,
      "num_input_tokens_seen": 61629688,
      "step": 106150
    },
    {
      "epoch": 15.810991957104557,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0038338530162267096,
      "loss": 0.7943,
      "num_input_tokens_seen": 61632600,
      "step": 106155
    },
    {
      "epoch": 15.811736669645517,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.003832551282417395,
      "loss": 0.791,
      "num_input_tokens_seen": 61635576,
      "step": 106160
    },
    {
      "epoch": 15.812481382186476,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.003831249737270771,
      "loss": 0.7906,
      "num_input_tokens_seen": 61638456,
      "step": 106165
    },
    {
      "epoch": 15.813226094727435,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0038299483808088364,
      "loss": 0.7982,
      "num_input_tokens_seen": 61641304,
      "step": 106170
    },
    {
      "epoch": 15.813970807268394,
      "grad_norm": 0.0546875,
      "learning_rate": 0.00382864721305357,
      "loss": 0.7978,
      "num_input_tokens_seen": 61644120,
      "step": 106175
    },
    {
      "epoch": 15.814715519809354,
      "grad_norm": 0.056640625,
      "learning_rate": 0.003827346234026952,
      "loss": 0.8029,
      "num_input_tokens_seen": 61647128,
      "step": 106180
    },
    {
      "epoch": 15.815460232350313,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.003826045443750969,
      "loss": 0.8004,
      "num_input_tokens_seen": 61650008,
      "step": 106185
    },
    {
      "epoch": 15.816204944891272,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0038247448422475848,
      "loss": 0.8154,
      "num_input_tokens_seen": 61653080,
      "step": 106190
    },
    {
      "epoch": 15.81694965743223,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.003823444429538784,
      "loss": 0.7992,
      "num_input_tokens_seen": 61655736,
      "step": 106195
    },
    {
      "epoch": 15.817694369973191,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.003822144205646526,
      "loss": 0.7966,
      "num_input_tokens_seen": 61658552,
      "step": 106200
    },
    {
      "epoch": 15.81843908251415,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.003820844170592786,
      "loss": 0.7735,
      "num_input_tokens_seen": 61661592,
      "step": 106205
    },
    {
      "epoch": 15.819183795055109,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0038195443243995214,
      "loss": 0.7964,
      "num_input_tokens_seen": 61664440,
      "step": 106210
    },
    {
      "epoch": 15.819928507596067,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0038182446670886865,
      "loss": 0.7884,
      "num_input_tokens_seen": 61667192,
      "step": 106215
    },
    {
      "epoch": 15.820673220137028,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0038169451986822486,
      "loss": 0.8034,
      "num_input_tokens_seen": 61670008,
      "step": 106220
    },
    {
      "epoch": 15.821417932677987,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0038156459192021557,
      "loss": 0.7909,
      "num_input_tokens_seen": 61672984,
      "step": 106225
    },
    {
      "epoch": 15.822162645218945,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0038143468286703577,
      "loss": 0.7851,
      "num_input_tokens_seen": 61675704,
      "step": 106230
    },
    {
      "epoch": 15.822907357759904,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0038130479271087964,
      "loss": 0.8103,
      "num_input_tokens_seen": 61678712,
      "step": 106235
    },
    {
      "epoch": 15.823652070300863,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0038117492145394263,
      "loss": 0.8053,
      "num_input_tokens_seen": 61681816,
      "step": 106240
    },
    {
      "epoch": 15.824396782841823,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.003810450690984176,
      "loss": 0.8121,
      "num_input_tokens_seen": 61684600,
      "step": 106245
    },
    {
      "epoch": 15.825141495382782,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0038091523564649944,
      "loss": 0.7808,
      "num_input_tokens_seen": 61687288,
      "step": 106250
    },
    {
      "epoch": 15.825886207923741,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.003807854211003808,
      "loss": 0.8054,
      "num_input_tokens_seen": 61690072,
      "step": 106255
    },
    {
      "epoch": 15.826630920464702,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.003806556254622547,
      "loss": 0.8053,
      "num_input_tokens_seen": 61692792,
      "step": 106260
    },
    {
      "epoch": 15.82737563300566,
      "grad_norm": 0.046875,
      "learning_rate": 0.0038052584873431456,
      "loss": 0.7876,
      "num_input_tokens_seen": 61696056,
      "step": 106265
    },
    {
      "epoch": 15.828120345546619,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0038039609091875203,
      "loss": 0.79,
      "num_input_tokens_seen": 61699128,
      "step": 106270
    },
    {
      "epoch": 15.828865058087578,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0038026635201775993,
      "loss": 0.8018,
      "num_input_tokens_seen": 61702008,
      "step": 106275
    },
    {
      "epoch": 15.829609770628537,
      "grad_norm": 0.03515625,
      "learning_rate": 0.003801366320335293,
      "loss": 0.7764,
      "num_input_tokens_seen": 61705048,
      "step": 106280
    },
    {
      "epoch": 15.830354483169497,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0038000693096825276,
      "loss": 0.7911,
      "num_input_tokens_seen": 61708216,
      "step": 106285
    },
    {
      "epoch": 15.831099195710456,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.003798772488241207,
      "loss": 0.7814,
      "num_input_tokens_seen": 61710904,
      "step": 106290
    },
    {
      "epoch": 15.831843908251415,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0037974758560332406,
      "loss": 0.7974,
      "num_input_tokens_seen": 61713720,
      "step": 106295
    },
    {
      "epoch": 15.832588620792373,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0037961794130805335,
      "loss": 0.8058,
      "num_input_tokens_seen": 61716888,
      "step": 106300
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.003794883159404984,
      "loss": 0.8134,
      "num_input_tokens_seen": 61720024,
      "step": 106305
    },
    {
      "epoch": 15.834078045874293,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.003793587095028499,
      "loss": 0.779,
      "num_input_tokens_seen": 61723288,
      "step": 106310
    },
    {
      "epoch": 15.834822758415251,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0037922912199729666,
      "loss": 0.7965,
      "num_input_tokens_seen": 61726168,
      "step": 106315
    },
    {
      "epoch": 15.83556747095621,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.003790995534260287,
      "loss": 0.7938,
      "num_input_tokens_seen": 61728824,
      "step": 106320
    },
    {
      "epoch": 15.83631218349717,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0037897000379123434,
      "loss": 0.8032,
      "num_input_tokens_seen": 61731544,
      "step": 106325
    },
    {
      "epoch": 15.83705689603813,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.003788404730951026,
      "loss": 0.7962,
      "num_input_tokens_seen": 61734840,
      "step": 106330
    },
    {
      "epoch": 15.837801608579088,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0037871096133982158,
      "loss": 0.7968,
      "num_input_tokens_seen": 61737656,
      "step": 106335
    },
    {
      "epoch": 15.838546321120047,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.003785814685275788,
      "loss": 0.8057,
      "num_input_tokens_seen": 61740792,
      "step": 106340
    },
    {
      "epoch": 15.839291033661008,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.00378451994660563,
      "loss": 0.7982,
      "num_input_tokens_seen": 61743896,
      "step": 106345
    },
    {
      "epoch": 15.840035746201966,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0037832253974096064,
      "loss": 0.7713,
      "num_input_tokens_seen": 61746488,
      "step": 106350
    },
    {
      "epoch": 15.840780458742925,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0037819310377095864,
      "loss": 0.7895,
      "num_input_tokens_seen": 61749304,
      "step": 106355
    },
    {
      "epoch": 15.841525171283884,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.003780636867527443,
      "loss": 0.8264,
      "num_input_tokens_seen": 61752536,
      "step": 106360
    },
    {
      "epoch": 15.842269883824844,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.003779342886885037,
      "loss": 0.8077,
      "num_input_tokens_seen": 61755448,
      "step": 106365
    },
    {
      "epoch": 15.843014596365803,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0037780490958042303,
      "loss": 0.7998,
      "num_input_tokens_seen": 61758200,
      "step": 106370
    },
    {
      "epoch": 15.843759308906762,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.003776755494306873,
      "loss": 0.7925,
      "num_input_tokens_seen": 61760920,
      "step": 106375
    },
    {
      "epoch": 15.84450402144772,
      "grad_norm": 0.0625,
      "learning_rate": 0.0037754620824148287,
      "loss": 0.7794,
      "num_input_tokens_seen": 61763800,
      "step": 106380
    },
    {
      "epoch": 15.845248733988681,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0037741688601499418,
      "loss": 0.7912,
      "num_input_tokens_seen": 61766488,
      "step": 106385
    },
    {
      "epoch": 15.84599344652964,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.003772875827534066,
      "loss": 0.7997,
      "num_input_tokens_seen": 61769240,
      "step": 106390
    },
    {
      "epoch": 15.846738159070599,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.003771582984589038,
      "loss": 0.7827,
      "num_input_tokens_seen": 61772056,
      "step": 106395
    },
    {
      "epoch": 15.847482871611557,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.003770290331336708,
      "loss": 0.803,
      "num_input_tokens_seen": 61775000,
      "step": 106400
    },
    {
      "epoch": 15.848227584152518,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0037689978677989035,
      "loss": 0.7744,
      "num_input_tokens_seen": 61778008,
      "step": 106405
    },
    {
      "epoch": 15.848972296693477,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.003767705593997471,
      "loss": 0.7945,
      "num_input_tokens_seen": 61781080,
      "step": 106410
    },
    {
      "epoch": 15.849717009234435,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.003766413509954238,
      "loss": 0.792,
      "num_input_tokens_seen": 61784184,
      "step": 106415
    },
    {
      "epoch": 15.850461721775394,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.003765121615691029,
      "loss": 0.8035,
      "num_input_tokens_seen": 61787128,
      "step": 106420
    },
    {
      "epoch": 15.851206434316353,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0037638299112296735,
      "loss": 0.8151,
      "num_input_tokens_seen": 61789848,
      "step": 106425
    },
    {
      "epoch": 15.851951146857314,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.003762538396591986,
      "loss": 0.7929,
      "num_input_tokens_seen": 61792728,
      "step": 106430
    },
    {
      "epoch": 15.852695859398272,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0037612470717997963,
      "loss": 0.8093,
      "num_input_tokens_seen": 61795736,
      "step": 106435
    },
    {
      "epoch": 15.853440571939231,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0037599559368749103,
      "loss": 0.7952,
      "num_input_tokens_seen": 61798584,
      "step": 106440
    },
    {
      "epoch": 15.85418528448019,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0037586649918391482,
      "loss": 0.7881,
      "num_input_tokens_seen": 61801720,
      "step": 106445
    },
    {
      "epoch": 15.85492999702115,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.003757374236714317,
      "loss": 0.7946,
      "num_input_tokens_seen": 61804696,
      "step": 106450
    },
    {
      "epoch": 15.85567470956211,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0037560836715222166,
      "loss": 0.775,
      "num_input_tokens_seen": 61807736,
      "step": 106455
    },
    {
      "epoch": 15.856419422103068,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.00375479329628466,
      "loss": 0.7898,
      "num_input_tokens_seen": 61810776,
      "step": 106460
    },
    {
      "epoch": 15.857164134644027,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0037535031110234366,
      "loss": 0.7936,
      "num_input_tokens_seen": 61813656,
      "step": 106465
    },
    {
      "epoch": 15.857908847184987,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.003752213115760352,
      "loss": 0.8021,
      "num_input_tokens_seen": 61816920,
      "step": 106470
    },
    {
      "epoch": 15.858653559725946,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0037509233105171907,
      "loss": 0.8039,
      "num_input_tokens_seen": 61819608,
      "step": 106475
    },
    {
      "epoch": 15.859398272266905,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0037496336953157516,
      "loss": 0.7958,
      "num_input_tokens_seen": 61822552,
      "step": 106480
    },
    {
      "epoch": 15.860142984807863,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0037483442701778174,
      "loss": 0.7886,
      "num_input_tokens_seen": 61825752,
      "step": 106485
    },
    {
      "epoch": 15.860887697348824,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.003747055035125169,
      "loss": 0.8144,
      "num_input_tokens_seen": 61828504,
      "step": 106490
    },
    {
      "epoch": 15.861632409889783,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.003745765990179591,
      "loss": 0.7797,
      "num_input_tokens_seen": 61831480,
      "step": 106495
    },
    {
      "epoch": 15.862377122430741,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0037444771353628537,
      "loss": 0.8082,
      "num_input_tokens_seen": 61834456,
      "step": 106500
    },
    {
      "epoch": 15.8631218349717,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0037431884706967383,
      "loss": 0.7966,
      "num_input_tokens_seen": 61837208,
      "step": 106505
    },
    {
      "epoch": 15.86386654751266,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.003741899996203009,
      "loss": 0.7965,
      "num_input_tokens_seen": 61840024,
      "step": 106510
    },
    {
      "epoch": 15.86461126005362,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.003740611711903442,
      "loss": 0.7933,
      "num_input_tokens_seen": 61842616,
      "step": 106515
    },
    {
      "epoch": 15.865355972594578,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.003739323617819793,
      "loss": 0.8087,
      "num_input_tokens_seen": 61845464,
      "step": 106520
    },
    {
      "epoch": 15.866100685135537,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0037380357139738295,
      "loss": 0.789,
      "num_input_tokens_seen": 61848248,
      "step": 106525
    },
    {
      "epoch": 15.866845397676498,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.0037367480003873076,
      "loss": 0.8022,
      "num_input_tokens_seen": 61850968,
      "step": 106530
    },
    {
      "epoch": 15.867590110217456,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0037354604770819763,
      "loss": 0.7932,
      "num_input_tokens_seen": 61853752,
      "step": 106535
    },
    {
      "epoch": 15.868334822758415,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0037341731440795954,
      "loss": 0.7913,
      "num_input_tokens_seen": 61856952,
      "step": 106540
    },
    {
      "epoch": 15.869079535299374,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0037328860014019054,
      "loss": 0.8036,
      "num_input_tokens_seen": 61859992,
      "step": 106545
    },
    {
      "epoch": 15.869824247840334,
      "grad_norm": 0.03125,
      "learning_rate": 0.003731599049070657,
      "loss": 0.7869,
      "num_input_tokens_seen": 61862936,
      "step": 106550
    },
    {
      "epoch": 15.870568960381293,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.003730312287107592,
      "loss": 0.8261,
      "num_input_tokens_seen": 61865912,
      "step": 106555
    },
    {
      "epoch": 15.871313672922252,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.003729025715534447,
      "loss": 0.8078,
      "num_input_tokens_seen": 61868440,
      "step": 106560
    },
    {
      "epoch": 15.87205838546321,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0037277393343729524,
      "loss": 0.7915,
      "num_input_tokens_seen": 61871448,
      "step": 106565
    },
    {
      "epoch": 15.872803098004171,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0037264531436448497,
      "loss": 0.8256,
      "num_input_tokens_seen": 61874232,
      "step": 106570
    },
    {
      "epoch": 15.87354781054513,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.003725167143371864,
      "loss": 0.802,
      "num_input_tokens_seen": 61877144,
      "step": 106575
    },
    {
      "epoch": 15.874292523086089,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0037238813335757157,
      "loss": 0.7958,
      "num_input_tokens_seen": 61880056,
      "step": 106580
    },
    {
      "epoch": 15.875037235627047,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0037225957142781345,
      "loss": 0.7912,
      "num_input_tokens_seen": 61882840,
      "step": 106585
    },
    {
      "epoch": 15.875781948168008,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0037213102855008346,
      "loss": 0.7997,
      "num_input_tokens_seen": 61885880,
      "step": 106590
    },
    {
      "epoch": 15.876526660708967,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0037200250472655366,
      "loss": 0.8009,
      "num_input_tokens_seen": 61888664,
      "step": 106595
    },
    {
      "epoch": 15.877271373249926,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0037187399995939467,
      "loss": 0.8105,
      "num_input_tokens_seen": 61891320,
      "step": 106600
    },
    {
      "epoch": 15.878016085790884,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0037174551425077844,
      "loss": 0.785,
      "num_input_tokens_seen": 61894296,
      "step": 106605
    },
    {
      "epoch": 15.878760798331843,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0037161704760287507,
      "loss": 0.7982,
      "num_input_tokens_seen": 61897208,
      "step": 106610
    },
    {
      "epoch": 15.879505510872804,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0037148860001785427,
      "loss": 0.8035,
      "num_input_tokens_seen": 61900184,
      "step": 106615
    },
    {
      "epoch": 15.880250223413762,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00371360171497887,
      "loss": 0.7767,
      "num_input_tokens_seen": 61902840,
      "step": 106620
    },
    {
      "epoch": 15.880994935954721,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0037123176204514273,
      "loss": 0.8201,
      "num_input_tokens_seen": 61906104,
      "step": 106625
    },
    {
      "epoch": 15.88173964849568,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.003711033716617905,
      "loss": 0.8004,
      "num_input_tokens_seen": 61908888,
      "step": 106630
    },
    {
      "epoch": 15.88248436103664,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0037097500034999897,
      "loss": 0.7926,
      "num_input_tokens_seen": 61911832,
      "step": 106635
    },
    {
      "epoch": 15.8832290735776,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0037084664811193784,
      "loss": 0.791,
      "num_input_tokens_seen": 61914776,
      "step": 106640
    },
    {
      "epoch": 15.883973786118558,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0037071831494977455,
      "loss": 0.7953,
      "num_input_tokens_seen": 61917432,
      "step": 106645
    },
    {
      "epoch": 15.884718498659517,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0037059000086567795,
      "loss": 0.797,
      "num_input_tokens_seen": 61920344,
      "step": 106650
    },
    {
      "epoch": 15.885463211200477,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0037046170586181536,
      "loss": 0.7972,
      "num_input_tokens_seen": 61922904,
      "step": 106655
    },
    {
      "epoch": 15.886207923741436,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0037033342994035393,
      "loss": 0.7855,
      "num_input_tokens_seen": 61926008,
      "step": 106660
    },
    {
      "epoch": 15.886952636282395,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0037020517310346144,
      "loss": 0.8154,
      "num_input_tokens_seen": 61928792,
      "step": 106665
    },
    {
      "epoch": 15.887697348823353,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.003700769353533038,
      "loss": 0.8029,
      "num_input_tokens_seen": 61931864,
      "step": 106670
    },
    {
      "epoch": 15.888442061364314,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.003699487166920485,
      "loss": 0.8017,
      "num_input_tokens_seen": 61934648,
      "step": 106675
    },
    {
      "epoch": 15.889186773905273,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0036982051712186087,
      "loss": 0.804,
      "num_input_tokens_seen": 61937528,
      "step": 106680
    },
    {
      "epoch": 15.889931486446232,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0036969233664490672,
      "loss": 0.8259,
      "num_input_tokens_seen": 61940760,
      "step": 106685
    },
    {
      "epoch": 15.89067619898719,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0036956417526335206,
      "loss": 0.7785,
      "num_input_tokens_seen": 61943800,
      "step": 106690
    },
    {
      "epoch": 15.89142091152815,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.003694360329793617,
      "loss": 0.815,
      "num_input_tokens_seen": 61946680,
      "step": 106695
    },
    {
      "epoch": 15.89216562406911,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.003693079097951007,
      "loss": 0.7882,
      "num_input_tokens_seen": 61949784,
      "step": 106700
    },
    {
      "epoch": 15.892910336610068,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.003691798057127327,
      "loss": 0.7923,
      "num_input_tokens_seen": 61953048,
      "step": 106705
    },
    {
      "epoch": 15.893655049151027,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0036905172073442324,
      "loss": 0.79,
      "num_input_tokens_seen": 61955736,
      "step": 106710
    },
    {
      "epoch": 15.894399761691988,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.003689236548623349,
      "loss": 0.7896,
      "num_input_tokens_seen": 61958904,
      "step": 106715
    },
    {
      "epoch": 15.895144474232946,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.003687956080986324,
      "loss": 0.8031,
      "num_input_tokens_seen": 61962072,
      "step": 106720
    },
    {
      "epoch": 15.895889186773905,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.003686675804454779,
      "loss": 0.8117,
      "num_input_tokens_seen": 61965144,
      "step": 106725
    },
    {
      "epoch": 15.896633899314864,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0036853957190503515,
      "loss": 0.7974,
      "num_input_tokens_seen": 61968248,
      "step": 106730
    },
    {
      "epoch": 15.897378611855824,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0036841158247946657,
      "loss": 0.7981,
      "num_input_tokens_seen": 61971160,
      "step": 106735
    },
    {
      "epoch": 15.898123324396783,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0036828361217093363,
      "loss": 0.8199,
      "num_input_tokens_seen": 61974008,
      "step": 106740
    },
    {
      "epoch": 15.898868036937742,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0036815566098159923,
      "loss": 0.7924,
      "num_input_tokens_seen": 61976824,
      "step": 106745
    },
    {
      "epoch": 15.8996127494787,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0036802772891362467,
      "loss": 0.7975,
      "num_input_tokens_seen": 61979608,
      "step": 106750
    },
    {
      "epoch": 15.90035746201966,
      "grad_norm": 0.048828125,
      "learning_rate": 0.003678998159691712,
      "loss": 0.8099,
      "num_input_tokens_seen": 61982488,
      "step": 106755
    },
    {
      "epoch": 15.90110217456062,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.003677719221503991,
      "loss": 0.7855,
      "num_input_tokens_seen": 61985400,
      "step": 106760
    },
    {
      "epoch": 15.901846887101579,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0036764404745947016,
      "loss": 0.817,
      "num_input_tokens_seen": 61988280,
      "step": 106765
    },
    {
      "epoch": 15.902591599642538,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.003675161918985438,
      "loss": 0.7945,
      "num_input_tokens_seen": 61991096,
      "step": 106770
    },
    {
      "epoch": 15.903336312183498,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0036738835546978074,
      "loss": 0.812,
      "num_input_tokens_seen": 61994008,
      "step": 106775
    },
    {
      "epoch": 15.904081024724457,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.003672605381753404,
      "loss": 0.8113,
      "num_input_tokens_seen": 61997240,
      "step": 106780
    },
    {
      "epoch": 15.904825737265416,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0036713274001738136,
      "loss": 0.7834,
      "num_input_tokens_seen": 62000184,
      "step": 106785
    },
    {
      "epoch": 15.905570449806374,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.003670049609980639,
      "loss": 0.7898,
      "num_input_tokens_seen": 62003096,
      "step": 106790
    },
    {
      "epoch": 15.906315162347333,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.003668772011195455,
      "loss": 0.7909,
      "num_input_tokens_seen": 62006296,
      "step": 106795
    },
    {
      "epoch": 15.907059874888294,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.003667494603839857,
      "loss": 0.8001,
      "num_input_tokens_seen": 62009336,
      "step": 106800
    },
    {
      "epoch": 15.907804587429252,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0036662173879354154,
      "loss": 0.8047,
      "num_input_tokens_seen": 62012152,
      "step": 106805
    },
    {
      "epoch": 15.908549299970211,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0036649403635037148,
      "loss": 0.7905,
      "num_input_tokens_seen": 62014776,
      "step": 106810
    },
    {
      "epoch": 15.90929401251117,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0036636635305663276,
      "loss": 0.7866,
      "num_input_tokens_seen": 62017752,
      "step": 106815
    },
    {
      "epoch": 15.91003872505213,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.003662386889144822,
      "loss": 0.7927,
      "num_input_tokens_seen": 62020600,
      "step": 106820
    },
    {
      "epoch": 15.91078343759309,
      "grad_norm": 0.03125,
      "learning_rate": 0.0036611104392607697,
      "loss": 0.7896,
      "num_input_tokens_seen": 62023384,
      "step": 106825
    },
    {
      "epoch": 15.911528150134048,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0036598341809357254,
      "loss": 0.7919,
      "num_input_tokens_seen": 62026488,
      "step": 106830
    },
    {
      "epoch": 15.912272862675007,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0036585581141912607,
      "loss": 0.8063,
      "num_input_tokens_seen": 62029080,
      "step": 106835
    },
    {
      "epoch": 15.913017575215967,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0036572822390489273,
      "loss": 0.8048,
      "num_input_tokens_seen": 62032344,
      "step": 106840
    },
    {
      "epoch": 15.913762287756926,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0036560065555302857,
      "loss": 0.7882,
      "num_input_tokens_seen": 62035224,
      "step": 106845
    },
    {
      "epoch": 15.914507000297885,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.00365473106365688,
      "loss": 0.8011,
      "num_input_tokens_seen": 62037976,
      "step": 106850
    },
    {
      "epoch": 15.915251712838844,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0036534557634502673,
      "loss": 0.7935,
      "num_input_tokens_seen": 62040888,
      "step": 106855
    },
    {
      "epoch": 15.915996425379804,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0036521806549319863,
      "loss": 0.8052,
      "num_input_tokens_seen": 62043576,
      "step": 106860
    },
    {
      "epoch": 15.916741137920763,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.0036509057381235747,
      "loss": 0.8032,
      "num_input_tokens_seen": 62046776,
      "step": 106865
    },
    {
      "epoch": 15.917485850461722,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0036496310130465807,
      "loss": 0.7965,
      "num_input_tokens_seen": 62049720,
      "step": 106870
    },
    {
      "epoch": 15.91823056300268,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.00364835647972253,
      "loss": 0.7907,
      "num_input_tokens_seen": 62052760,
      "step": 106875
    },
    {
      "epoch": 15.918975275543641,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0036470821381729642,
      "loss": 0.8081,
      "num_input_tokens_seen": 62055768,
      "step": 106880
    },
    {
      "epoch": 15.9197199880846,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0036458079884194076,
      "loss": 0.7893,
      "num_input_tokens_seen": 62058424,
      "step": 106885
    },
    {
      "epoch": 15.920464700625558,
      "grad_norm": 0.0213623046875,
      "learning_rate": 0.0036445340304833863,
      "loss": 0.8034,
      "num_input_tokens_seen": 62061272,
      "step": 106890
    },
    {
      "epoch": 15.921209413166517,
      "grad_norm": 0.044921875,
      "learning_rate": 0.003643260264386419,
      "loss": 0.8104,
      "num_input_tokens_seen": 62063960,
      "step": 106895
    },
    {
      "epoch": 15.921954125707478,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.003641986690150023,
      "loss": 0.7945,
      "num_input_tokens_seen": 62067128,
      "step": 106900
    },
    {
      "epoch": 15.922698838248436,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.0036407133077957237,
      "loss": 0.8007,
      "num_input_tokens_seen": 62069912,
      "step": 106905
    },
    {
      "epoch": 15.923443550789395,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.0036394401173450225,
      "loss": 0.8137,
      "num_input_tokens_seen": 62072760,
      "step": 106910
    },
    {
      "epoch": 15.924188263330354,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0036381671188194385,
      "loss": 0.8046,
      "num_input_tokens_seen": 62075608,
      "step": 106915
    },
    {
      "epoch": 15.924932975871315,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.00363689431224047,
      "loss": 0.7885,
      "num_input_tokens_seen": 62079032,
      "step": 106920
    },
    {
      "epoch": 15.925677688412273,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.0036356216976296263,
      "loss": 0.8018,
      "num_input_tokens_seen": 62081848,
      "step": 106925
    },
    {
      "epoch": 15.926422400953232,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0036343492750084017,
      "loss": 0.7846,
      "num_input_tokens_seen": 62084568,
      "step": 106930
    },
    {
      "epoch": 15.92716711349419,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0036330770443982956,
      "loss": 0.7965,
      "num_input_tokens_seen": 62087544,
      "step": 106935
    },
    {
      "epoch": 15.92791182603515,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0036318050058208027,
      "loss": 0.7852,
      "num_input_tokens_seen": 62090488,
      "step": 106940
    },
    {
      "epoch": 15.92865653857611,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.0036305331592974043,
      "loss": 0.803,
      "num_input_tokens_seen": 62093336,
      "step": 106945
    },
    {
      "epoch": 15.929401251117069,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.003629261504849598,
      "loss": 0.785,
      "num_input_tokens_seen": 62096088,
      "step": 106950
    },
    {
      "epoch": 15.930145963658028,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0036279900424988614,
      "loss": 0.8049,
      "num_input_tokens_seen": 62098968,
      "step": 106955
    },
    {
      "epoch": 15.930890676198988,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0036267187722666743,
      "loss": 0.7826,
      "num_input_tokens_seen": 62101880,
      "step": 106960
    },
    {
      "epoch": 15.931635388739947,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0036254476941745112,
      "loss": 0.8077,
      "num_input_tokens_seen": 62104824,
      "step": 106965
    },
    {
      "epoch": 15.932380101280906,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0036241768082438527,
      "loss": 0.7911,
      "num_input_tokens_seen": 62107768,
      "step": 106970
    },
    {
      "epoch": 15.933124813821864,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.003622906114496166,
      "loss": 0.8156,
      "num_input_tokens_seen": 62110520,
      "step": 106975
    },
    {
      "epoch": 15.933869526362823,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0036216356129529113,
      "loss": 0.7925,
      "num_input_tokens_seen": 62113272,
      "step": 106980
    },
    {
      "epoch": 15.934614238903784,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0036203653036355637,
      "loss": 0.7902,
      "num_input_tokens_seen": 62116152,
      "step": 106985
    },
    {
      "epoch": 15.935358951444742,
      "grad_norm": 0.0390625,
      "learning_rate": 0.003619095186565574,
      "loss": 0.8196,
      "num_input_tokens_seen": 62119192,
      "step": 106990
    },
    {
      "epoch": 15.936103663985701,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0036178252617644086,
      "loss": 0.7808,
      "num_input_tokens_seen": 62122232,
      "step": 106995
    },
    {
      "epoch": 15.93684837652666,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0036165555292535117,
      "loss": 0.7802,
      "num_input_tokens_seen": 62124952,
      "step": 107000
    },
    {
      "epoch": 15.93759308906762,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0036152859890543447,
      "loss": 0.8169,
      "num_input_tokens_seen": 62127832,
      "step": 107005
    },
    {
      "epoch": 15.93833780160858,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0036140166411883505,
      "loss": 0.8013,
      "num_input_tokens_seen": 62131000,
      "step": 107010
    },
    {
      "epoch": 15.939082514149538,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0036127474856769686,
      "loss": 0.7878,
      "num_input_tokens_seen": 62133816,
      "step": 107015
    },
    {
      "epoch": 15.939827226690497,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0036114785225416484,
      "loss": 0.7969,
      "num_input_tokens_seen": 62136792,
      "step": 107020
    },
    {
      "epoch": 15.940571939231457,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.003610209751803823,
      "loss": 0.7985,
      "num_input_tokens_seen": 62139576,
      "step": 107025
    },
    {
      "epoch": 15.941316651772416,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0036089411734849296,
      "loss": 0.8247,
      "num_input_tokens_seen": 62142424,
      "step": 107030
    },
    {
      "epoch": 15.942061364313375,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0036076727876063917,
      "loss": 0.8055,
      "num_input_tokens_seen": 62145272,
      "step": 107035
    },
    {
      "epoch": 15.942806076854334,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0036064045941896486,
      "loss": 0.8059,
      "num_input_tokens_seen": 62148248,
      "step": 107040
    },
    {
      "epoch": 15.943550789395294,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.003605136593256115,
      "loss": 0.8356,
      "num_input_tokens_seen": 62150968,
      "step": 107045
    },
    {
      "epoch": 15.944295501936253,
      "grad_norm": 0.05078125,
      "learning_rate": 0.003603868784827222,
      "loss": 0.7981,
      "num_input_tokens_seen": 62153912,
      "step": 107050
    },
    {
      "epoch": 15.945040214477212,
      "grad_norm": 0.05078125,
      "learning_rate": 0.003602601168924384,
      "loss": 0.7857,
      "num_input_tokens_seen": 62157272,
      "step": 107055
    },
    {
      "epoch": 15.94578492701817,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0036013337455690106,
      "loss": 0.7833,
      "num_input_tokens_seen": 62160024,
      "step": 107060
    },
    {
      "epoch": 15.946529639559131,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0036000665147825233,
      "loss": 0.7944,
      "num_input_tokens_seen": 62162936,
      "step": 107065
    },
    {
      "epoch": 15.94727435210009,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.003598799476586321,
      "loss": 0.8137,
      "num_input_tokens_seen": 62165880,
      "step": 107070
    },
    {
      "epoch": 15.948019064641048,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0035975326310018173,
      "loss": 0.7867,
      "num_input_tokens_seen": 62168632,
      "step": 107075
    },
    {
      "epoch": 15.948763777182007,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0035962659780504128,
      "loss": 0.8048,
      "num_input_tokens_seen": 62171576,
      "step": 107080
    },
    {
      "epoch": 15.949508489722968,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0035949995177535026,
      "loss": 0.814,
      "num_input_tokens_seen": 62174488,
      "step": 107085
    },
    {
      "epoch": 15.950253202263927,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0035937332501324807,
      "loss": 0.7705,
      "num_input_tokens_seen": 62177112,
      "step": 107090
    },
    {
      "epoch": 15.950997914804885,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.003592467175208747,
      "loss": 0.8093,
      "num_input_tokens_seen": 62179928,
      "step": 107095
    },
    {
      "epoch": 15.951742627345844,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.003591201293003686,
      "loss": 0.8095,
      "num_input_tokens_seen": 62182872,
      "step": 107100
    },
    {
      "epoch": 15.952487339886805,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.003589935603538679,
      "loss": 0.7887,
      "num_input_tokens_seen": 62185848,
      "step": 107105
    },
    {
      "epoch": 15.953232052427763,
      "grad_norm": 0.044921875,
      "learning_rate": 0.003588670106835119,
      "loss": 0.804,
      "num_input_tokens_seen": 62188632,
      "step": 107110
    },
    {
      "epoch": 15.953976764968722,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.003587404802914375,
      "loss": 0.7769,
      "num_input_tokens_seen": 62191480,
      "step": 107115
    },
    {
      "epoch": 15.95472147750968,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0035861396917978303,
      "loss": 0.7737,
      "num_input_tokens_seen": 62194488,
      "step": 107120
    },
    {
      "epoch": 15.95546619005064,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.003584874773506853,
      "loss": 0.7981,
      "num_input_tokens_seen": 62197304,
      "step": 107125
    },
    {
      "epoch": 15.9562109025916,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0035836100480628173,
      "loss": 0.8007,
      "num_input_tokens_seen": 62200344,
      "step": 107130
    },
    {
      "epoch": 15.956955615132559,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0035823455154870864,
      "loss": 0.8051,
      "num_input_tokens_seen": 62203224,
      "step": 107135
    },
    {
      "epoch": 15.957700327673518,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0035810811758010194,
      "loss": 0.8,
      "num_input_tokens_seen": 62206072,
      "step": 107140
    },
    {
      "epoch": 15.958445040214476,
      "grad_norm": 0.0546875,
      "learning_rate": 0.003579817029025987,
      "loss": 0.7886,
      "num_input_tokens_seen": 62209080,
      "step": 107145
    },
    {
      "epoch": 15.959189752755437,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.003578553075183335,
      "loss": 0.7858,
      "num_input_tokens_seen": 62211832,
      "step": 107150
    },
    {
      "epoch": 15.959934465296396,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.003577289314294424,
      "loss": 0.7875,
      "num_input_tokens_seen": 62214680,
      "step": 107155
    },
    {
      "epoch": 15.960679177837354,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.003576025746380595,
      "loss": 0.813,
      "num_input_tokens_seen": 62217432,
      "step": 107160
    },
    {
      "epoch": 15.961423890378313,
      "grad_norm": 0.05078125,
      "learning_rate": 0.003574762371463205,
      "loss": 0.7955,
      "num_input_tokens_seen": 62220280,
      "step": 107165
    },
    {
      "epoch": 15.962168602919274,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.003573499189563586,
      "loss": 0.8032,
      "num_input_tokens_seen": 62223160,
      "step": 107170
    },
    {
      "epoch": 15.962913315460233,
      "grad_norm": 0.03125,
      "learning_rate": 0.0035722362007030908,
      "loss": 0.7992,
      "num_input_tokens_seen": 62226104,
      "step": 107175
    },
    {
      "epoch": 15.963658028001191,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0035709734049030515,
      "loss": 0.7998,
      "num_input_tokens_seen": 62229048,
      "step": 107180
    },
    {
      "epoch": 15.96440274054215,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0035697108021847945,
      "loss": 0.8102,
      "num_input_tokens_seen": 62231800,
      "step": 107185
    },
    {
      "epoch": 15.96514745308311,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.003568448392569662,
      "loss": 0.777,
      "num_input_tokens_seen": 62234584,
      "step": 107190
    },
    {
      "epoch": 15.96589216562407,
      "grad_norm": 0.03515625,
      "learning_rate": 0.00356718617607897,
      "loss": 0.8022,
      "num_input_tokens_seen": 62237432,
      "step": 107195
    },
    {
      "epoch": 15.966636878165028,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0035659241527340522,
      "loss": 0.8118,
      "num_input_tokens_seen": 62240344,
      "step": 107200
    },
    {
      "epoch": 15.967381590705987,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.003564662322556222,
      "loss": 0.8001,
      "num_input_tokens_seen": 62243320,
      "step": 107205
    },
    {
      "epoch": 15.968126303246947,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0035634006855668012,
      "loss": 0.776,
      "num_input_tokens_seen": 62246392,
      "step": 107210
    },
    {
      "epoch": 15.968871015787906,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0035621392417871044,
      "loss": 0.7873,
      "num_input_tokens_seen": 62249336,
      "step": 107215
    },
    {
      "epoch": 15.969615728328865,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0035608779912384386,
      "loss": 0.7993,
      "num_input_tokens_seen": 62252152,
      "step": 107220
    },
    {
      "epoch": 15.970360440869824,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0035596169339421136,
      "loss": 0.8024,
      "num_input_tokens_seen": 62254840,
      "step": 107225
    },
    {
      "epoch": 15.971105153410784,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0035583560699194287,
      "loss": 0.8006,
      "num_input_tokens_seen": 62257816,
      "step": 107230
    },
    {
      "epoch": 15.971849865951743,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0035570953991916924,
      "loss": 0.7968,
      "num_input_tokens_seen": 62260728,
      "step": 107235
    },
    {
      "epoch": 15.972594578492702,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0035558349217801976,
      "loss": 0.8106,
      "num_input_tokens_seen": 62263416,
      "step": 107240
    },
    {
      "epoch": 15.97333929103366,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.003554574637706242,
      "loss": 0.798,
      "num_input_tokens_seen": 62266552,
      "step": 107245
    },
    {
      "epoch": 15.974084003574621,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0035533145469911136,
      "loss": 0.7914,
      "num_input_tokens_seen": 62269464,
      "step": 107250
    },
    {
      "epoch": 15.97482871611558,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.003552054649656104,
      "loss": 0.794,
      "num_input_tokens_seen": 62272408,
      "step": 107255
    },
    {
      "epoch": 15.975573428656539,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.003550794945722497,
      "loss": 0.8023,
      "num_input_tokens_seen": 62275256,
      "step": 107260
    },
    {
      "epoch": 15.976318141197497,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0035495354352115685,
      "loss": 0.8038,
      "num_input_tokens_seen": 62278264,
      "step": 107265
    },
    {
      "epoch": 15.977062853738456,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.003548276118144606,
      "loss": 0.7883,
      "num_input_tokens_seen": 62281016,
      "step": 107270
    },
    {
      "epoch": 15.977807566279417,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0035470169945428735,
      "loss": 0.791,
      "num_input_tokens_seen": 62283608,
      "step": 107275
    },
    {
      "epoch": 15.978552278820375,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0035457580644276546,
      "loss": 0.7951,
      "num_input_tokens_seen": 62286584,
      "step": 107280
    },
    {
      "epoch": 15.979296991361334,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0035444993278202106,
      "loss": 0.8049,
      "num_input_tokens_seen": 62289496,
      "step": 107285
    },
    {
      "epoch": 15.980041703902295,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0035432407847418076,
      "loss": 0.7997,
      "num_input_tokens_seen": 62292536,
      "step": 107290
    },
    {
      "epoch": 15.980786416443253,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.003541982435213704,
      "loss": 0.8166,
      "num_input_tokens_seen": 62295608,
      "step": 107295
    },
    {
      "epoch": 15.981531128984212,
      "grad_norm": 0.03515625,
      "learning_rate": 0.003540724279257166,
      "loss": 0.8038,
      "num_input_tokens_seen": 62299256,
      "step": 107300
    },
    {
      "epoch": 15.982275841525171,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.003539466316893444,
      "loss": 0.8097,
      "num_input_tokens_seen": 62302328,
      "step": 107305
    },
    {
      "epoch": 15.98302055406613,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0035382085481437875,
      "loss": 0.798,
      "num_input_tokens_seen": 62305464,
      "step": 107310
    },
    {
      "epoch": 15.98376526660709,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.00353695097302945,
      "loss": 0.8084,
      "num_input_tokens_seen": 62308376,
      "step": 107315
    },
    {
      "epoch": 15.984509979148049,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0035356935915716735,
      "loss": 0.8237,
      "num_input_tokens_seen": 62311096,
      "step": 107320
    },
    {
      "epoch": 15.985254691689008,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0035344364037917055,
      "loss": 0.8274,
      "num_input_tokens_seen": 62314008,
      "step": 107325
    },
    {
      "epoch": 15.985999404229966,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0035331794097107766,
      "loss": 0.8056,
      "num_input_tokens_seen": 62317112,
      "step": 107330
    },
    {
      "epoch": 15.986744116770927,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.003531922609350132,
      "loss": 0.7974,
      "num_input_tokens_seen": 62319928,
      "step": 107335
    },
    {
      "epoch": 15.987488829311886,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0035306660027309986,
      "loss": 0.7751,
      "num_input_tokens_seen": 62322872,
      "step": 107340
    },
    {
      "epoch": 15.988233541852845,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.003529409589874603,
      "loss": 0.8207,
      "num_input_tokens_seen": 62325912,
      "step": 107345
    },
    {
      "epoch": 15.988978254393803,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0035281533708021776,
      "loss": 0.7922,
      "num_input_tokens_seen": 62328952,
      "step": 107350
    },
    {
      "epoch": 15.989722966934764,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.003526897345534943,
      "loss": 0.7641,
      "num_input_tokens_seen": 62331960,
      "step": 107355
    },
    {
      "epoch": 15.990467679475723,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0035256415140941157,
      "loss": 0.7872,
      "num_input_tokens_seen": 62334648,
      "step": 107360
    },
    {
      "epoch": 15.991212392016681,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.003524385876500909,
      "loss": 0.7899,
      "num_input_tokens_seen": 62337464,
      "step": 107365
    },
    {
      "epoch": 15.99195710455764,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0035231304327765427,
      "loss": 0.8094,
      "num_input_tokens_seen": 62340408,
      "step": 107370
    },
    {
      "epoch": 15.9927018170986,
      "grad_norm": 0.025146484375,
      "learning_rate": 0.0035218751829422207,
      "loss": 0.7914,
      "num_input_tokens_seen": 62343288,
      "step": 107375
    },
    {
      "epoch": 15.99344652963956,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0035206201270191555,
      "loss": 0.8041,
      "num_input_tokens_seen": 62346360,
      "step": 107380
    },
    {
      "epoch": 15.994191242180518,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0035193652650285455,
      "loss": 0.7904,
      "num_input_tokens_seen": 62349432,
      "step": 107385
    },
    {
      "epoch": 15.994935954721477,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.003518110596991587,
      "loss": 0.7984,
      "num_input_tokens_seen": 62352504,
      "step": 107390
    },
    {
      "epoch": 15.995680667262437,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.003516856122929485,
      "loss": 0.7709,
      "num_input_tokens_seen": 62355448,
      "step": 107395
    },
    {
      "epoch": 15.996425379803396,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0035156018428634233,
      "loss": 0.8093,
      "num_input_tokens_seen": 62358392,
      "step": 107400
    },
    {
      "epoch": 15.997170092344355,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.003514347756814599,
      "loss": 0.7902,
      "num_input_tokens_seen": 62361144,
      "step": 107405
    },
    {
      "epoch": 15.997914804885314,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0035130938648041984,
      "loss": 0.8067,
      "num_input_tokens_seen": 62363768,
      "step": 107410
    },
    {
      "epoch": 15.998659517426274,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0035118401668534,
      "loss": 0.7987,
      "num_input_tokens_seen": 62366360,
      "step": 107415
    },
    {
      "epoch": 15.999404229967233,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0035105866629833826,
      "loss": 0.7927,
      "num_input_tokens_seen": 62369464,
      "step": 107420
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.7989339828491211,
      "eval_runtime": 70.497,
      "eval_samples_per_second": 42.328,
      "eval_steps_per_second": 10.582,
      "num_input_tokens_seen": 62371472,
      "step": 107424
    },
    {
      "epoch": 16.000148942508194,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0035093333532153314,
      "loss": 0.8017,
      "num_input_tokens_seen": 62372016,
      "step": 107425
    },
    {
      "epoch": 16.00089365504915,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0035080802375704145,
      "loss": 0.8031,
      "num_input_tokens_seen": 62374608,
      "step": 107430
    },
    {
      "epoch": 16.00163836759011,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0035068273160697963,
      "loss": 0.7848,
      "num_input_tokens_seen": 62377584,
      "step": 107435
    },
    {
      "epoch": 16.002383080131068,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.003505574588734654,
      "loss": 0.7901,
      "num_input_tokens_seen": 62380464,
      "step": 107440
    },
    {
      "epoch": 16.00312779267203,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.003504322055586143,
      "loss": 0.7923,
      "num_input_tokens_seen": 62383312,
      "step": 107445
    },
    {
      "epoch": 16.00387250521299,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.003503069716645432,
      "loss": 0.8158,
      "num_input_tokens_seen": 62386320,
      "step": 107450
    },
    {
      "epoch": 16.004617217753946,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.0035018175719336675,
      "loss": 0.8035,
      "num_input_tokens_seen": 62389168,
      "step": 107455
    },
    {
      "epoch": 16.005361930294907,
      "grad_norm": 0.037109375,
      "learning_rate": 0.003500565621472014,
      "loss": 0.7946,
      "num_input_tokens_seen": 62392144,
      "step": 107460
    },
    {
      "epoch": 16.006106642835864,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.003499313865281617,
      "loss": 0.8186,
      "num_input_tokens_seen": 62395568,
      "step": 107465
    },
    {
      "epoch": 16.006851355376824,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00349806230338362,
      "loss": 0.7907,
      "num_input_tokens_seen": 62398448,
      "step": 107470
    },
    {
      "epoch": 16.007596067917785,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.0034968109357991725,
      "loss": 0.7926,
      "num_input_tokens_seen": 62401232,
      "step": 107475
    },
    {
      "epoch": 16.00834078045874,
      "grad_norm": 0.05078125,
      "learning_rate": 0.003495559762549414,
      "loss": 0.7749,
      "num_input_tokens_seen": 62404016,
      "step": 107480
    },
    {
      "epoch": 16.009085492999702,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0034943087836554814,
      "loss": 0.7966,
      "num_input_tokens_seen": 62406800,
      "step": 107485
    },
    {
      "epoch": 16.009830205540663,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.003493057999138502,
      "loss": 0.7899,
      "num_input_tokens_seen": 62409648,
      "step": 107490
    },
    {
      "epoch": 16.01057491808162,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.003491807409019618,
      "loss": 0.8006,
      "num_input_tokens_seen": 62412624,
      "step": 107495
    },
    {
      "epoch": 16.01131963062258,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.003490557013319952,
      "loss": 0.7741,
      "num_input_tokens_seen": 62415536,
      "step": 107500
    },
    {
      "epoch": 16.012064343163537,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.003489306812060623,
      "loss": 0.7899,
      "num_input_tokens_seen": 62418448,
      "step": 107505
    },
    {
      "epoch": 16.012809055704498,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.003488056805262759,
      "loss": 0.8137,
      "num_input_tokens_seen": 62421136,
      "step": 107510
    },
    {
      "epoch": 16.01355376824546,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.003486806992947472,
      "loss": 0.8119,
      "num_input_tokens_seen": 62424400,
      "step": 107515
    },
    {
      "epoch": 16.014298480786415,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.003485557375135885,
      "loss": 0.8104,
      "num_input_tokens_seen": 62427152,
      "step": 107520
    },
    {
      "epoch": 16.015043193327376,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0034843079518490972,
      "loss": 0.7886,
      "num_input_tokens_seen": 62429936,
      "step": 107525
    },
    {
      "epoch": 16.015787905868336,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.003483058723108227,
      "loss": 0.7963,
      "num_input_tokens_seen": 62432976,
      "step": 107530
    },
    {
      "epoch": 16.016532618409293,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.003481809688934371,
      "loss": 0.7901,
      "num_input_tokens_seen": 62435568,
      "step": 107535
    },
    {
      "epoch": 16.017277330950254,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0034805608493486376,
      "loss": 0.7897,
      "num_input_tokens_seen": 62438288,
      "step": 107540
    },
    {
      "epoch": 16.01802204349121,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0034793122043721213,
      "loss": 0.766,
      "num_input_tokens_seen": 62441264,
      "step": 107545
    },
    {
      "epoch": 16.01876675603217,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.003478063754025915,
      "loss": 0.8006,
      "num_input_tokens_seen": 62444432,
      "step": 107550
    },
    {
      "epoch": 16.019511468573132,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0034768154983311123,
      "loss": 0.7993,
      "num_input_tokens_seen": 62447440,
      "step": 107555
    },
    {
      "epoch": 16.02025618111409,
      "grad_norm": 0.046875,
      "learning_rate": 0.0034755674373087957,
      "loss": 0.8018,
      "num_input_tokens_seen": 62450224,
      "step": 107560
    },
    {
      "epoch": 16.02100089365505,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.003474319570980059,
      "loss": 0.7874,
      "num_input_tokens_seen": 62452912,
      "step": 107565
    },
    {
      "epoch": 16.02174560619601,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.003473071899365973,
      "loss": 0.7937,
      "num_input_tokens_seen": 62456208,
      "step": 107570
    },
    {
      "epoch": 16.022490318736967,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0034718244224876283,
      "loss": 0.7967,
      "num_input_tokens_seen": 62459344,
      "step": 107575
    },
    {
      "epoch": 16.023235031277927,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0034705771403660923,
      "loss": 0.7848,
      "num_input_tokens_seen": 62462448,
      "step": 107580
    },
    {
      "epoch": 16.023979743818884,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.003469330053022432,
      "loss": 0.7992,
      "num_input_tokens_seen": 62465264,
      "step": 107585
    },
    {
      "epoch": 16.024724456359845,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.003468083160477727,
      "loss": 0.793,
      "num_input_tokens_seen": 62468496,
      "step": 107590
    },
    {
      "epoch": 16.025469168900806,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0034668364627530317,
      "loss": 0.7856,
      "num_input_tokens_seen": 62471696,
      "step": 107595
    },
    {
      "epoch": 16.026213881441763,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.003465589959869417,
      "loss": 0.8187,
      "num_input_tokens_seen": 62474512,
      "step": 107600
    },
    {
      "epoch": 16.026958593982723,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.003464343651847933,
      "loss": 0.7887,
      "num_input_tokens_seen": 62477424,
      "step": 107605
    },
    {
      "epoch": 16.027703306523684,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0034630975387096425,
      "loss": 0.7915,
      "num_input_tokens_seen": 62480240,
      "step": 107610
    },
    {
      "epoch": 16.02844801906464,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0034618516204755925,
      "loss": 0.8206,
      "num_input_tokens_seen": 62483056,
      "step": 107615
    },
    {
      "epoch": 16.0291927316056,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.003460605897166834,
      "loss": 0.785,
      "num_input_tokens_seen": 62486032,
      "step": 107620
    },
    {
      "epoch": 16.029937444146558,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.003459360368804411,
      "loss": 0.8172,
      "num_input_tokens_seen": 62488880,
      "step": 107625
    },
    {
      "epoch": 16.03068215668752,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0034581150354093602,
      "loss": 0.7767,
      "num_input_tokens_seen": 62491952,
      "step": 107630
    },
    {
      "epoch": 16.03142686922848,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.003456869897002728,
      "loss": 0.8117,
      "num_input_tokens_seen": 62494800,
      "step": 107635
    },
    {
      "epoch": 16.032171581769436,
      "grad_norm": 0.02734375,
      "learning_rate": 0.003455624953605545,
      "loss": 0.8032,
      "num_input_tokens_seen": 62497584,
      "step": 107640
    },
    {
      "epoch": 16.032916294310397,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0034543802052388484,
      "loss": 0.7914,
      "num_input_tokens_seen": 62501136,
      "step": 107645
    },
    {
      "epoch": 16.033661006851354,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.00345313565192366,
      "loss": 0.804,
      "num_input_tokens_seen": 62504112,
      "step": 107650
    },
    {
      "epoch": 16.034405719392314,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.003451891293681013,
      "loss": 0.8069,
      "num_input_tokens_seen": 62506640,
      "step": 107655
    },
    {
      "epoch": 16.035150431933275,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.003450647130531926,
      "loss": 0.7897,
      "num_input_tokens_seen": 62509392,
      "step": 107660
    },
    {
      "epoch": 16.03589514447423,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.003449403162497413,
      "loss": 0.7867,
      "num_input_tokens_seen": 62512464,
      "step": 107665
    },
    {
      "epoch": 16.036639857015192,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.0034481593895984977,
      "loss": 0.8327,
      "num_input_tokens_seen": 62515312,
      "step": 107670
    },
    {
      "epoch": 16.037384569556153,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.003446915811856185,
      "loss": 0.8053,
      "num_input_tokens_seen": 62518192,
      "step": 107675
    },
    {
      "epoch": 16.03812928209711,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.0034456724292914937,
      "loss": 0.7906,
      "num_input_tokens_seen": 62521232,
      "step": 107680
    },
    {
      "epoch": 16.03887399463807,
      "grad_norm": 0.033203125,
      "learning_rate": 0.003444429241925421,
      "loss": 0.8085,
      "num_input_tokens_seen": 62524176,
      "step": 107685
    },
    {
      "epoch": 16.039618707179027,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.003443186249778973,
      "loss": 0.8095,
      "num_input_tokens_seen": 62527568,
      "step": 107690
    },
    {
      "epoch": 16.040363419719988,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.003441943452873143,
      "loss": 0.7997,
      "num_input_tokens_seen": 62530288,
      "step": 107695
    },
    {
      "epoch": 16.04110813226095,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0034407008512289342,
      "loss": 0.8056,
      "num_input_tokens_seen": 62533040,
      "step": 107700
    },
    {
      "epoch": 16.041852844801905,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0034394584448673377,
      "loss": 0.795,
      "num_input_tokens_seen": 62536272,
      "step": 107705
    },
    {
      "epoch": 16.042597557342866,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0034382162338093356,
      "loss": 0.7881,
      "num_input_tokens_seen": 62539056,
      "step": 107710
    },
    {
      "epoch": 16.043342269883826,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0034369742180759256,
      "loss": 0.8298,
      "num_input_tokens_seen": 62541776,
      "step": 107715
    },
    {
      "epoch": 16.044086982424783,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.003435732397688078,
      "loss": 0.7769,
      "num_input_tokens_seen": 62545008,
      "step": 107720
    },
    {
      "epoch": 16.044831694965744,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0034344907726667822,
      "loss": 0.7932,
      "num_input_tokens_seen": 62548080,
      "step": 107725
    },
    {
      "epoch": 16.0455764075067,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.003433249343033007,
      "loss": 0.8292,
      "num_input_tokens_seen": 62550704,
      "step": 107730
    },
    {
      "epoch": 16.04632112004766,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0034320081088077318,
      "loss": 0.7847,
      "num_input_tokens_seen": 62553456,
      "step": 107735
    },
    {
      "epoch": 16.047065832588622,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.003430767070011924,
      "loss": 0.784,
      "num_input_tokens_seen": 62556464,
      "step": 107740
    },
    {
      "epoch": 16.04781054512958,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0034295262266665454,
      "loss": 0.7791,
      "num_input_tokens_seen": 62559408,
      "step": 107745
    },
    {
      "epoch": 16.04855525767054,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.003428285578792559,
      "loss": 0.7785,
      "num_input_tokens_seen": 62562128,
      "step": 107750
    },
    {
      "epoch": 16.0492999702115,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0034270451264109317,
      "loss": 0.81,
      "num_input_tokens_seen": 62565008,
      "step": 107755
    },
    {
      "epoch": 16.050044682752457,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0034258048695426134,
      "loss": 0.8074,
      "num_input_tokens_seen": 62567792,
      "step": 107760
    },
    {
      "epoch": 16.050789395293418,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0034245648082085536,
      "loss": 0.787,
      "num_input_tokens_seen": 62570576,
      "step": 107765
    },
    {
      "epoch": 16.051534107834375,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.003423324942429713,
      "loss": 0.7976,
      "num_input_tokens_seen": 62573200,
      "step": 107770
    },
    {
      "epoch": 16.052278820375335,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0034220852722270257,
      "loss": 0.8014,
      "num_input_tokens_seen": 62576720,
      "step": 107775
    },
    {
      "epoch": 16.053023532916296,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.003420845797621445,
      "loss": 0.8156,
      "num_input_tokens_seen": 62579440,
      "step": 107780
    },
    {
      "epoch": 16.053768245457253,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.003419606518633906,
      "loss": 0.8107,
      "num_input_tokens_seen": 62582384,
      "step": 107785
    },
    {
      "epoch": 16.054512957998213,
      "grad_norm": 0.0546875,
      "learning_rate": 0.003418367435285341,
      "loss": 0.7955,
      "num_input_tokens_seen": 62585328,
      "step": 107790
    },
    {
      "epoch": 16.055257670539174,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0034171285475966904,
      "loss": 0.7966,
      "num_input_tokens_seen": 62588176,
      "step": 107795
    },
    {
      "epoch": 16.05600238308013,
      "grad_norm": 0.041015625,
      "learning_rate": 0.003415889855588878,
      "loss": 0.7794,
      "num_input_tokens_seen": 62591120,
      "step": 107800
    },
    {
      "epoch": 16.05674709562109,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.003414651359282836,
      "loss": 0.787,
      "num_input_tokens_seen": 62594160,
      "step": 107805
    },
    {
      "epoch": 16.057491808162048,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0034134130586994846,
      "loss": 0.8003,
      "num_input_tokens_seen": 62596784,
      "step": 107810
    },
    {
      "epoch": 16.05823652070301,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.003412174953859745,
      "loss": 0.7889,
      "num_input_tokens_seen": 62599664,
      "step": 107815
    },
    {
      "epoch": 16.05898123324397,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00341093704478453,
      "loss": 0.7858,
      "num_input_tokens_seen": 62602512,
      "step": 107820
    },
    {
      "epoch": 16.059725945784926,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.0034096993314947526,
      "loss": 0.7737,
      "num_input_tokens_seen": 62605488,
      "step": 107825
    },
    {
      "epoch": 16.060470658325887,
      "grad_norm": 0.03515625,
      "learning_rate": 0.003408461814011328,
      "loss": 0.7926,
      "num_input_tokens_seen": 62608080,
      "step": 107830
    },
    {
      "epoch": 16.061215370866844,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.003407224492355156,
      "loss": 0.8025,
      "num_input_tokens_seen": 62610768,
      "step": 107835
    },
    {
      "epoch": 16.061960083407804,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.003405987366547149,
      "loss": 0.7953,
      "num_input_tokens_seen": 62613936,
      "step": 107840
    },
    {
      "epoch": 16.062704795948765,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0034047504366081977,
      "loss": 0.7865,
      "num_input_tokens_seen": 62616976,
      "step": 107845
    },
    {
      "epoch": 16.06344950848972,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.003403513702559205,
      "loss": 0.8043,
      "num_input_tokens_seen": 62620080,
      "step": 107850
    },
    {
      "epoch": 16.064194221030682,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.003402277164421061,
      "loss": 0.7904,
      "num_input_tokens_seen": 62623024,
      "step": 107855
    },
    {
      "epoch": 16.064938933571643,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0034010408222146585,
      "loss": 0.7888,
      "num_input_tokens_seen": 62626064,
      "step": 107860
    },
    {
      "epoch": 16.0656836461126,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.003399804675960884,
      "loss": 0.7895,
      "num_input_tokens_seen": 62629104,
      "step": 107865
    },
    {
      "epoch": 16.06642835865356,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0033985687256806152,
      "loss": 0.7974,
      "num_input_tokens_seen": 62631920,
      "step": 107870
    },
    {
      "epoch": 16.067173071194517,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.00339733297139474,
      "loss": 0.7932,
      "num_input_tokens_seen": 62634864,
      "step": 107875
    },
    {
      "epoch": 16.067917783735478,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.003396097413124133,
      "loss": 0.8268,
      "num_input_tokens_seen": 62637552,
      "step": 107880
    },
    {
      "epoch": 16.06866249627644,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.003394862050889666,
      "loss": 0.7933,
      "num_input_tokens_seen": 62640560,
      "step": 107885
    },
    {
      "epoch": 16.069407208817395,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.003393626884712205,
      "loss": 0.7848,
      "num_input_tokens_seen": 62643504,
      "step": 107890
    },
    {
      "epoch": 16.070151921358356,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0033923919146126255,
      "loss": 0.8084,
      "num_input_tokens_seen": 62646640,
      "step": 107895
    },
    {
      "epoch": 16.070896633899316,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.003391157140611784,
      "loss": 0.792,
      "num_input_tokens_seen": 62649456,
      "step": 107900
    },
    {
      "epoch": 16.071641346440273,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.003389922562730548,
      "loss": 0.8078,
      "num_input_tokens_seen": 62652560,
      "step": 107905
    },
    {
      "epoch": 16.072386058981234,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0033886881809897704,
      "loss": 0.8059,
      "num_input_tokens_seen": 62655440,
      "step": 107910
    },
    {
      "epoch": 16.07313077152219,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0033874539954103005,
      "loss": 0.8094,
      "num_input_tokens_seen": 62658192,
      "step": 107915
    },
    {
      "epoch": 16.07387548406315,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0033862200060129958,
      "loss": 0.7956,
      "num_input_tokens_seen": 62660880,
      "step": 107920
    },
    {
      "epoch": 16.074620196604112,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0033849862128186967,
      "loss": 0.8027,
      "num_input_tokens_seen": 62663664,
      "step": 107925
    },
    {
      "epoch": 16.07536490914507,
      "grad_norm": 0.078125,
      "learning_rate": 0.003383752615848257,
      "loss": 0.786,
      "num_input_tokens_seen": 62666672,
      "step": 107930
    },
    {
      "epoch": 16.07610962168603,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0033825192151225036,
      "loss": 0.7936,
      "num_input_tokens_seen": 62669456,
      "step": 107935
    },
    {
      "epoch": 16.07685433422699,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.003381286010662286,
      "loss": 0.8019,
      "num_input_tokens_seen": 62672336,
      "step": 107940
    },
    {
      "epoch": 16.077599046767947,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0033800530024884323,
      "loss": 0.8082,
      "num_input_tokens_seen": 62675248,
      "step": 107945
    },
    {
      "epoch": 16.078343759308908,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.003378820190621774,
      "loss": 0.7935,
      "num_input_tokens_seen": 62678064,
      "step": 107950
    },
    {
      "epoch": 16.079088471849865,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.003377587575083136,
      "loss": 0.7932,
      "num_input_tokens_seen": 62681168,
      "step": 107955
    },
    {
      "epoch": 16.079833184390825,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.0033763551558933405,
      "loss": 0.7846,
      "num_input_tokens_seen": 62683664,
      "step": 107960
    },
    {
      "epoch": 16.080577896931786,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.003375122933073214,
      "loss": 0.8,
      "num_input_tokens_seen": 62686672,
      "step": 107965
    },
    {
      "epoch": 16.081322609472743,
      "grad_norm": 0.037109375,
      "learning_rate": 0.003373890906643568,
      "loss": 0.7901,
      "num_input_tokens_seen": 62689680,
      "step": 107970
    },
    {
      "epoch": 16.082067322013703,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0033726590766252206,
      "loss": 0.7951,
      "num_input_tokens_seen": 62692368,
      "step": 107975
    },
    {
      "epoch": 16.082812034554664,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.003371427443038979,
      "loss": 0.8108,
      "num_input_tokens_seen": 62695472,
      "step": 107980
    },
    {
      "epoch": 16.08355674709562,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0033701960059056547,
      "loss": 0.7975,
      "num_input_tokens_seen": 62698288,
      "step": 107985
    },
    {
      "epoch": 16.08430145963658,
      "grad_norm": 0.056640625,
      "learning_rate": 0.003368964765246049,
      "loss": 0.8107,
      "num_input_tokens_seen": 62701040,
      "step": 107990
    },
    {
      "epoch": 16.085046172177538,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0033677337210809575,
      "loss": 0.8335,
      "num_input_tokens_seen": 62703984,
      "step": 107995
    },
    {
      "epoch": 16.0857908847185,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0033665028734311867,
      "loss": 0.7953,
      "num_input_tokens_seen": 62706736,
      "step": 108000
    },
    {
      "epoch": 16.08653559725946,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.003365272222317523,
      "loss": 0.811,
      "num_input_tokens_seen": 62709296,
      "step": 108005
    },
    {
      "epoch": 16.087280309800416,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.003364041767760764,
      "loss": 0.8042,
      "num_input_tokens_seen": 62712272,
      "step": 108010
    },
    {
      "epoch": 16.088025022341377,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0033628115097816924,
      "loss": 0.7952,
      "num_input_tokens_seen": 62715184,
      "step": 108015
    },
    {
      "epoch": 16.088769734882334,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.003361581448401093,
      "loss": 0.7976,
      "num_input_tokens_seen": 62718288,
      "step": 108020
    },
    {
      "epoch": 16.089514447423294,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.003360351583639746,
      "loss": 0.7811,
      "num_input_tokens_seen": 62721168,
      "step": 108025
    },
    {
      "epoch": 16.090259159964255,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0033591219155184255,
      "loss": 0.7734,
      "num_input_tokens_seen": 62723728,
      "step": 108030
    },
    {
      "epoch": 16.091003872505212,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0033578924440579136,
      "loss": 0.793,
      "num_input_tokens_seen": 62726768,
      "step": 108035
    },
    {
      "epoch": 16.091748585046172,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0033566631692789717,
      "loss": 0.7855,
      "num_input_tokens_seen": 62729616,
      "step": 108040
    },
    {
      "epoch": 16.092493297587133,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.0033554340912023755,
      "loss": 0.794,
      "num_input_tokens_seen": 62732432,
      "step": 108045
    },
    {
      "epoch": 16.09323801012809,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0033542052098488803,
      "loss": 0.8142,
      "num_input_tokens_seen": 62735440,
      "step": 108050
    },
    {
      "epoch": 16.09398272266905,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0033529765252392585,
      "loss": 0.8062,
      "num_input_tokens_seen": 62738288,
      "step": 108055
    },
    {
      "epoch": 16.094727435210007,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0033517480373942542,
      "loss": 0.7981,
      "num_input_tokens_seen": 62741008,
      "step": 108060
    },
    {
      "epoch": 16.095472147750968,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.003350519746334633,
      "loss": 0.8055,
      "num_input_tokens_seen": 62744048,
      "step": 108065
    },
    {
      "epoch": 16.09621686029193,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0033492916520811416,
      "loss": 0.8165,
      "num_input_tokens_seen": 62747376,
      "step": 108070
    },
    {
      "epoch": 16.096961572832885,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0033480637546545247,
      "loss": 0.7839,
      "num_input_tokens_seen": 62750160,
      "step": 108075
    },
    {
      "epoch": 16.097706285373846,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0033468360540755263,
      "loss": 0.8018,
      "num_input_tokens_seen": 62753104,
      "step": 108080
    },
    {
      "epoch": 16.098450997914806,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.003345608550364891,
      "loss": 0.8147,
      "num_input_tokens_seen": 62755984,
      "step": 108085
    },
    {
      "epoch": 16.099195710455763,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.003344381243543356,
      "loss": 0.8228,
      "num_input_tokens_seen": 62759024,
      "step": 108090
    },
    {
      "epoch": 16.099940422996724,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0033431541336316492,
      "loss": 0.7724,
      "num_input_tokens_seen": 62762000,
      "step": 108095
    },
    {
      "epoch": 16.10068513553768,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0033419272206505093,
      "loss": 0.8037,
      "num_input_tokens_seen": 62764688,
      "step": 108100
    },
    {
      "epoch": 16.10142984807864,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0033407005046206598,
      "loss": 0.7849,
      "num_input_tokens_seen": 62767440,
      "step": 108105
    },
    {
      "epoch": 16.102174560619602,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0033394739855628222,
      "loss": 0.8086,
      "num_input_tokens_seen": 62770224,
      "step": 108110
    },
    {
      "epoch": 16.10291927316056,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0033382476634977237,
      "loss": 0.7985,
      "num_input_tokens_seen": 62773168,
      "step": 108115
    },
    {
      "epoch": 16.10366398570152,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0033370215384460754,
      "loss": 0.7794,
      "num_input_tokens_seen": 62775888,
      "step": 108120
    },
    {
      "epoch": 16.10440869824248,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0033357956104285978,
      "loss": 0.795,
      "num_input_tokens_seen": 62778672,
      "step": 108125
    },
    {
      "epoch": 16.105153410783437,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.0033345698794659943,
      "loss": 0.8014,
      "num_input_tokens_seen": 62781552,
      "step": 108130
    },
    {
      "epoch": 16.105898123324398,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.00333334434557898,
      "loss": 0.7779,
      "num_input_tokens_seen": 62784368,
      "step": 108135
    },
    {
      "epoch": 16.106642835865355,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0033321190087882567,
      "loss": 0.7933,
      "num_input_tokens_seen": 62787120,
      "step": 108140
    },
    {
      "epoch": 16.107387548406315,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0033308938691145244,
      "loss": 0.7901,
      "num_input_tokens_seen": 62789712,
      "step": 108145
    },
    {
      "epoch": 16.108132260947276,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0033296689265784785,
      "loss": 0.7812,
      "num_input_tokens_seen": 62792624,
      "step": 108150
    },
    {
      "epoch": 16.108876973488233,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.003328444181200812,
      "loss": 0.7863,
      "num_input_tokens_seen": 62795440,
      "step": 108155
    },
    {
      "epoch": 16.109621686029193,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0033272196330022233,
      "loss": 0.8039,
      "num_input_tokens_seen": 62798256,
      "step": 108160
    },
    {
      "epoch": 16.11036639857015,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.00332599528200339,
      "loss": 0.7866,
      "num_input_tokens_seen": 62801232,
      "step": 108165
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0033247711282250046,
      "loss": 0.8036,
      "num_input_tokens_seen": 62804112,
      "step": 108170
    },
    {
      "epoch": 16.11185582365207,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0033235471716877426,
      "loss": 0.8039,
      "num_input_tokens_seen": 62806928,
      "step": 108175
    },
    {
      "epoch": 16.11260053619303,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0033223234124122878,
      "loss": 0.8099,
      "num_input_tokens_seen": 62810032,
      "step": 108180
    },
    {
      "epoch": 16.11334524873399,
      "grad_norm": 0.2353515625,
      "learning_rate": 0.00332109985041931,
      "loss": 0.8279,
      "num_input_tokens_seen": 62813136,
      "step": 108185
    },
    {
      "epoch": 16.11408996127495,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0033198764857294743,
      "loss": 0.7845,
      "num_input_tokens_seen": 62816176,
      "step": 108190
    },
    {
      "epoch": 16.114834673815906,
      "grad_norm": 0.02392578125,
      "learning_rate": 0.0033186533183634606,
      "loss": 0.8057,
      "num_input_tokens_seen": 62818960,
      "step": 108195
    },
    {
      "epoch": 16.115579386356867,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0033174303483419214,
      "loss": 0.8078,
      "num_input_tokens_seen": 62821712,
      "step": 108200
    },
    {
      "epoch": 16.116324098897824,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0033162075756855277,
      "loss": 0.784,
      "num_input_tokens_seen": 62824496,
      "step": 108205
    },
    {
      "epoch": 16.117068811438784,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.0033149850004149323,
      "loss": 0.8022,
      "num_input_tokens_seen": 62827408,
      "step": 108210
    },
    {
      "epoch": 16.117813523979745,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.003313762622550789,
      "loss": 0.7879,
      "num_input_tokens_seen": 62830000,
      "step": 108215
    },
    {
      "epoch": 16.118558236520702,
      "grad_norm": 0.0390625,
      "learning_rate": 0.003312540442113744,
      "loss": 0.779,
      "num_input_tokens_seen": 62833040,
      "step": 108220
    },
    {
      "epoch": 16.119302949061662,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.0033113184591244543,
      "loss": 0.8147,
      "num_input_tokens_seen": 62836240,
      "step": 108225
    },
    {
      "epoch": 16.120047661602623,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.003310096673603559,
      "loss": 0.8573,
      "num_input_tokens_seen": 62838896,
      "step": 108230
    },
    {
      "epoch": 16.12079237414358,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0033088750855716957,
      "loss": 0.7897,
      "num_input_tokens_seen": 62842000,
      "step": 108235
    },
    {
      "epoch": 16.12153708668454,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0033076536950495095,
      "loss": 0.817,
      "num_input_tokens_seen": 62845072,
      "step": 108240
    },
    {
      "epoch": 16.122281799225497,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0033064325020576266,
      "loss": 0.794,
      "num_input_tokens_seen": 62847920,
      "step": 108245
    },
    {
      "epoch": 16.123026511766458,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.0033052115066166845,
      "loss": 0.8003,
      "num_input_tokens_seen": 62850768,
      "step": 108250
    },
    {
      "epoch": 16.12377122430742,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.0033039907087473053,
      "loss": 0.8202,
      "num_input_tokens_seen": 62853456,
      "step": 108255
    },
    {
      "epoch": 16.124515936848375,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.00330277010847012,
      "loss": 0.7958,
      "num_input_tokens_seen": 62856208,
      "step": 108260
    },
    {
      "epoch": 16.125260649389336,
      "grad_norm": 0.037109375,
      "learning_rate": 0.003301549705805746,
      "loss": 0.7855,
      "num_input_tokens_seen": 62858928,
      "step": 108265
    },
    {
      "epoch": 16.126005361930297,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.003300329500774793,
      "loss": 0.7988,
      "num_input_tokens_seen": 62861744,
      "step": 108270
    },
    {
      "epoch": 16.126750074471254,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0032991094933978882,
      "loss": 0.8182,
      "num_input_tokens_seen": 62864784,
      "step": 108275
    },
    {
      "epoch": 16.127494787012214,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0032978896836956366,
      "loss": 0.7918,
      "num_input_tokens_seen": 62867408,
      "step": 108280
    },
    {
      "epoch": 16.12823949955317,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0032966700716886435,
      "loss": 0.7799,
      "num_input_tokens_seen": 62870256,
      "step": 108285
    },
    {
      "epoch": 16.12898421209413,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.003295450657397513,
      "loss": 0.791,
      "num_input_tokens_seen": 62873424,
      "step": 108290
    },
    {
      "epoch": 16.129728924635092,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.003294231440842849,
      "loss": 0.8151,
      "num_input_tokens_seen": 62876368,
      "step": 108295
    },
    {
      "epoch": 16.13047363717605,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.003293012422045246,
      "loss": 0.7973,
      "num_input_tokens_seen": 62879312,
      "step": 108300
    },
    {
      "epoch": 16.13121834971701,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0032917936010253023,
      "loss": 0.8065,
      "num_input_tokens_seen": 62882000,
      "step": 108305
    },
    {
      "epoch": 16.13196306225797,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0032905749778036064,
      "loss": 0.8058,
      "num_input_tokens_seen": 62884848,
      "step": 108310
    },
    {
      "epoch": 16.132707774798927,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0032893565524007417,
      "loss": 0.8048,
      "num_input_tokens_seen": 62887664,
      "step": 108315
    },
    {
      "epoch": 16.133452487339888,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0032881383248373,
      "loss": 0.778,
      "num_input_tokens_seen": 62890704,
      "step": 108320
    },
    {
      "epoch": 16.134197199880845,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0032869202951338523,
      "loss": 0.7728,
      "num_input_tokens_seen": 62893584,
      "step": 108325
    },
    {
      "epoch": 16.134941912421805,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0032857024633109874,
      "loss": 0.7905,
      "num_input_tokens_seen": 62896432,
      "step": 108330
    },
    {
      "epoch": 16.135686624962766,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0032844848293892685,
      "loss": 0.7839,
      "num_input_tokens_seen": 62899824,
      "step": 108335
    },
    {
      "epoch": 16.136431337503723,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.003283267393389275,
      "loss": 0.7929,
      "num_input_tokens_seen": 62902512,
      "step": 108340
    },
    {
      "epoch": 16.137176050044683,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0032820501553315726,
      "loss": 0.798,
      "num_input_tokens_seen": 62905776,
      "step": 108345
    },
    {
      "epoch": 16.13792076258564,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0032808331152367216,
      "loss": 0.7805,
      "num_input_tokens_seen": 62908560,
      "step": 108350
    },
    {
      "epoch": 16.1386654751266,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0032796162731252845,
      "loss": 0.795,
      "num_input_tokens_seen": 62911504,
      "step": 108355
    },
    {
      "epoch": 16.13941018766756,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0032783996290178144,
      "loss": 0.8073,
      "num_input_tokens_seen": 62914320,
      "step": 108360
    },
    {
      "epoch": 16.14015490020852,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.003277183182934873,
      "loss": 0.804,
      "num_input_tokens_seen": 62917168,
      "step": 108365
    },
    {
      "epoch": 16.14089961274948,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.003275966934897003,
      "loss": 0.7964,
      "num_input_tokens_seen": 62919952,
      "step": 108370
    },
    {
      "epoch": 16.14164432529044,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0032747508849247605,
      "loss": 0.8041,
      "num_input_tokens_seen": 62922736,
      "step": 108375
    },
    {
      "epoch": 16.142389037831396,
      "grad_norm": 0.0625,
      "learning_rate": 0.003273535033038681,
      "loss": 0.7852,
      "num_input_tokens_seen": 62925552,
      "step": 108380
    },
    {
      "epoch": 16.143133750372357,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.003272319379259312,
      "loss": 0.7931,
      "num_input_tokens_seen": 62928208,
      "step": 108385
    },
    {
      "epoch": 16.143878462913314,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0032711039236071873,
      "loss": 0.7859,
      "num_input_tokens_seen": 62931184,
      "step": 108390
    },
    {
      "epoch": 16.144623175454274,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.003269888666102838,
      "loss": 0.7854,
      "num_input_tokens_seen": 62933936,
      "step": 108395
    },
    {
      "epoch": 16.145367887995235,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.003268673606766802,
      "loss": 0.8046,
      "num_input_tokens_seen": 62937264,
      "step": 108400
    },
    {
      "epoch": 16.146112600536192,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.003267458745619603,
      "loss": 0.7911,
      "num_input_tokens_seen": 62940240,
      "step": 108405
    },
    {
      "epoch": 16.146857313077152,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.003266244082681758,
      "loss": 0.7943,
      "num_input_tokens_seen": 62943088,
      "step": 108410
    },
    {
      "epoch": 16.147602025618113,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0032650296179737985,
      "loss": 0.8012,
      "num_input_tokens_seen": 62945968,
      "step": 108415
    },
    {
      "epoch": 16.14834673815907,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.003263815351516237,
      "loss": 0.794,
      "num_input_tokens_seen": 62948624,
      "step": 108420
    },
    {
      "epoch": 16.14909145070003,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.0032626012833295867,
      "loss": 0.8064,
      "num_input_tokens_seen": 62951728,
      "step": 108425
    },
    {
      "epoch": 16.149836163240987,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0032613874134343533,
      "loss": 0.8068,
      "num_input_tokens_seen": 62954704,
      "step": 108430
    },
    {
      "epoch": 16.150580875781948,
      "grad_norm": 0.0625,
      "learning_rate": 0.003260173741851055,
      "loss": 0.8066,
      "num_input_tokens_seen": 62957392,
      "step": 108435
    },
    {
      "epoch": 16.15132558832291,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0032589602686001837,
      "loss": 0.8011,
      "num_input_tokens_seen": 62960144,
      "step": 108440
    },
    {
      "epoch": 16.152070300863866,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.00325774699370225,
      "loss": 0.798,
      "num_input_tokens_seen": 62963120,
      "step": 108445
    },
    {
      "epoch": 16.152815013404826,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.003256533917177744,
      "loss": 0.7965,
      "num_input_tokens_seen": 62966064,
      "step": 108450
    },
    {
      "epoch": 16.153559725945787,
      "grad_norm": 0.091796875,
      "learning_rate": 0.003255321039047164,
      "loss": 0.8214,
      "num_input_tokens_seen": 62969168,
      "step": 108455
    },
    {
      "epoch": 16.154304438486744,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.003254108359330996,
      "loss": 0.7907,
      "num_input_tokens_seen": 62971952,
      "step": 108460
    },
    {
      "epoch": 16.155049151027704,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0032528958780497328,
      "loss": 0.8034,
      "num_input_tokens_seen": 62974704,
      "step": 108465
    },
    {
      "epoch": 16.15579386356866,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0032516835952238537,
      "loss": 0.7904,
      "num_input_tokens_seen": 62977680,
      "step": 108470
    },
    {
      "epoch": 16.15653857610962,
      "grad_norm": 0.03515625,
      "learning_rate": 0.00325047151087384,
      "loss": 0.8011,
      "num_input_tokens_seen": 62980368,
      "step": 108475
    },
    {
      "epoch": 16.157283288650582,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0032492596250201693,
      "loss": 0.8097,
      "num_input_tokens_seen": 62983696,
      "step": 108480
    },
    {
      "epoch": 16.15802800119154,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.00324804793768331,
      "loss": 0.7891,
      "num_input_tokens_seen": 62986544,
      "step": 108485
    },
    {
      "epoch": 16.1587727137325,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0032468364488837404,
      "loss": 0.7966,
      "num_input_tokens_seen": 62989328,
      "step": 108490
    },
    {
      "epoch": 16.15951742627346,
      "grad_norm": 0.0625,
      "learning_rate": 0.0032456251586419193,
      "loss": 0.7979,
      "num_input_tokens_seen": 62992464,
      "step": 108495
    },
    {
      "epoch": 16.160262138814417,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.003244414066978319,
      "loss": 0.7899,
      "num_input_tokens_seen": 62995632,
      "step": 108500
    },
    {
      "epoch": 16.161006851355378,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0032432031739133907,
      "loss": 0.7992,
      "num_input_tokens_seen": 62998544,
      "step": 108505
    },
    {
      "epoch": 16.161751563896335,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.003241992479467599,
      "loss": 0.8083,
      "num_input_tokens_seen": 63001584,
      "step": 108510
    },
    {
      "epoch": 16.162496276437295,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0032407819836613953,
      "loss": 0.804,
      "num_input_tokens_seen": 63004208,
      "step": 108515
    },
    {
      "epoch": 16.163240988978256,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.003239571686515225,
      "loss": 0.7981,
      "num_input_tokens_seen": 63007056,
      "step": 108520
    },
    {
      "epoch": 16.163985701519213,
      "grad_norm": 0.0208740234375,
      "learning_rate": 0.0032383615880495417,
      "loss": 0.815,
      "num_input_tokens_seen": 63010192,
      "step": 108525
    },
    {
      "epoch": 16.164730414060173,
      "grad_norm": 0.04296875,
      "learning_rate": 0.003237151688284781,
      "loss": 0.8179,
      "num_input_tokens_seen": 63013008,
      "step": 108530
    },
    {
      "epoch": 16.16547512660113,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.003235941987241391,
      "loss": 0.7986,
      "num_input_tokens_seen": 63016112,
      "step": 108535
    },
    {
      "epoch": 16.16621983914209,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0032347324849398046,
      "loss": 0.8112,
      "num_input_tokens_seen": 63019088,
      "step": 108540
    },
    {
      "epoch": 16.16696455168305,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.003233523181400456,
      "loss": 0.7891,
      "num_input_tokens_seen": 63022160,
      "step": 108545
    },
    {
      "epoch": 16.16770926422401,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0032323140766437747,
      "loss": 0.7749,
      "num_input_tokens_seen": 63024976,
      "step": 108550
    },
    {
      "epoch": 16.16845397676497,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.003231105170690182,
      "loss": 0.7974,
      "num_input_tokens_seen": 63027696,
      "step": 108555
    },
    {
      "epoch": 16.16919868930593,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0032298964635601108,
      "loss": 0.8024,
      "num_input_tokens_seen": 63030480,
      "step": 108560
    },
    {
      "epoch": 16.169943401846886,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.003228687955273972,
      "loss": 0.818,
      "num_input_tokens_seen": 63033392,
      "step": 108565
    },
    {
      "epoch": 16.170688114387847,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0032274796458521904,
      "loss": 0.7901,
      "num_input_tokens_seen": 63036496,
      "step": 108570
    },
    {
      "epoch": 16.171432826928804,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0032262715353151704,
      "loss": 0.7935,
      "num_input_tokens_seen": 63039472,
      "step": 108575
    },
    {
      "epoch": 16.172177539469764,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0032250636236833313,
      "loss": 0.7932,
      "num_input_tokens_seen": 63042544,
      "step": 108580
    },
    {
      "epoch": 16.172922252010725,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.0032238559109770697,
      "loss": 0.7952,
      "num_input_tokens_seen": 63045520,
      "step": 108585
    },
    {
      "epoch": 16.173666964551682,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0032226483972167996,
      "loss": 0.8098,
      "num_input_tokens_seen": 63048464,
      "step": 108590
    },
    {
      "epoch": 16.174411677092642,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.003221441082422913,
      "loss": 0.8116,
      "num_input_tokens_seen": 63051440,
      "step": 108595
    },
    {
      "epoch": 16.175156389633603,
      "grad_norm": 0.078125,
      "learning_rate": 0.0032202339666158046,
      "loss": 0.8018,
      "num_input_tokens_seen": 63054192,
      "step": 108600
    },
    {
      "epoch": 16.17590110217456,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0032190270498158756,
      "loss": 0.7972,
      "num_input_tokens_seen": 63057296,
      "step": 108605
    },
    {
      "epoch": 16.17664581471552,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.003217820332043511,
      "loss": 0.7902,
      "num_input_tokens_seen": 63059728,
      "step": 108610
    },
    {
      "epoch": 16.177390527256478,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.003216613813319096,
      "loss": 0.7921,
      "num_input_tokens_seen": 63062640,
      "step": 108615
    },
    {
      "epoch": 16.178135239797438,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0032154074936630115,
      "loss": 0.7985,
      "num_input_tokens_seen": 63065392,
      "step": 108620
    },
    {
      "epoch": 16.1788799523384,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.003214201373095644,
      "loss": 0.8091,
      "num_input_tokens_seen": 63068272,
      "step": 108625
    },
    {
      "epoch": 16.179624664879356,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0032129954516373657,
      "loss": 0.7999,
      "num_input_tokens_seen": 63070928,
      "step": 108630
    },
    {
      "epoch": 16.180369377420316,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0032117897293085437,
      "loss": 0.799,
      "num_input_tokens_seen": 63073904,
      "step": 108635
    },
    {
      "epoch": 16.181114089961277,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0032105842061295603,
      "loss": 0.7817,
      "num_input_tokens_seen": 63077136,
      "step": 108640
    },
    {
      "epoch": 16.181858802502234,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.0032093788821207685,
      "loss": 0.7905,
      "num_input_tokens_seen": 63079952,
      "step": 108645
    },
    {
      "epoch": 16.182603515043194,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.003208173757302541,
      "loss": 0.8059,
      "num_input_tokens_seen": 63082768,
      "step": 108650
    },
    {
      "epoch": 16.18334822758415,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0032069688316952293,
      "loss": 0.8129,
      "num_input_tokens_seen": 63085776,
      "step": 108655
    },
    {
      "epoch": 16.18409294012511,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0032057641053191994,
      "loss": 0.7834,
      "num_input_tokens_seen": 63088432,
      "step": 108660
    },
    {
      "epoch": 16.184837652666072,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.003204559578194791,
      "loss": 0.8042,
      "num_input_tokens_seen": 63091440,
      "step": 108665
    },
    {
      "epoch": 16.18558236520703,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.003203355250342365,
      "loss": 0.8094,
      "num_input_tokens_seen": 63094480,
      "step": 108670
    },
    {
      "epoch": 16.18632707774799,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0032021511217822634,
      "loss": 0.7943,
      "num_input_tokens_seen": 63097488,
      "step": 108675
    },
    {
      "epoch": 16.187071790288947,
      "grad_norm": 0.048828125,
      "learning_rate": 0.003200947192534826,
      "loss": 0.8038,
      "num_input_tokens_seen": 63100528,
      "step": 108680
    },
    {
      "epoch": 16.187816502829907,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.003199743462620396,
      "loss": 0.7923,
      "num_input_tokens_seen": 63103760,
      "step": 108685
    },
    {
      "epoch": 16.188561215370868,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0031985399320593015,
      "loss": 0.8031,
      "num_input_tokens_seen": 63106544,
      "step": 108690
    },
    {
      "epoch": 16.189305927911825,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.003197336600871884,
      "loss": 0.7934,
      "num_input_tokens_seen": 63109296,
      "step": 108695
    },
    {
      "epoch": 16.190050640452785,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.003196133469078467,
      "loss": 0.8033,
      "num_input_tokens_seen": 63112016,
      "step": 108700
    },
    {
      "epoch": 16.190795352993746,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.003194930536699381,
      "loss": 0.7796,
      "num_input_tokens_seen": 63114768,
      "step": 108705
    },
    {
      "epoch": 16.191540065534703,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.003193727803754944,
      "loss": 0.7855,
      "num_input_tokens_seen": 63117424,
      "step": 108710
    },
    {
      "epoch": 16.192284778075663,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.003192525270265475,
      "loss": 0.7917,
      "num_input_tokens_seen": 63120432,
      "step": 108715
    },
    {
      "epoch": 16.19302949061662,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0031913229362512934,
      "loss": 0.8166,
      "num_input_tokens_seen": 63123248,
      "step": 108720
    },
    {
      "epoch": 16.19377420315758,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0031901208017327052,
      "loss": 0.799,
      "num_input_tokens_seen": 63126192,
      "step": 108725
    },
    {
      "epoch": 16.19451891569854,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.003188918866730026,
      "loss": 0.8248,
      "num_input_tokens_seen": 63128912,
      "step": 108730
    },
    {
      "epoch": 16.1952636282395,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.003187717131263559,
      "loss": 0.7964,
      "num_input_tokens_seen": 63131696,
      "step": 108735
    },
    {
      "epoch": 16.19600834078046,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0031865155953536024,
      "loss": 0.8162,
      "num_input_tokens_seen": 63134704,
      "step": 108740
    },
    {
      "epoch": 16.19675305332142,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0031853142590204616,
      "loss": 0.79,
      "num_input_tokens_seen": 63137648,
      "step": 108745
    },
    {
      "epoch": 16.197497765862376,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.003184113122284429,
      "loss": 0.7797,
      "num_input_tokens_seen": 63140880,
      "step": 108750
    },
    {
      "epoch": 16.198242478403337,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0031829121851657963,
      "loss": 0.7889,
      "num_input_tokens_seen": 63143568,
      "step": 108755
    },
    {
      "epoch": 16.198987190944294,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0031817114476848474,
      "loss": 0.7763,
      "num_input_tokens_seen": 63146544,
      "step": 108760
    },
    {
      "epoch": 16.199731903485254,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0031805109098618748,
      "loss": 0.7837,
      "num_input_tokens_seen": 63149648,
      "step": 108765
    },
    {
      "epoch": 16.200476616026215,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.003179310571717155,
      "loss": 0.7911,
      "num_input_tokens_seen": 63152560,
      "step": 108770
    },
    {
      "epoch": 16.201221328567172,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0031781104332709725,
      "loss": 0.8134,
      "num_input_tokens_seen": 63155344,
      "step": 108775
    },
    {
      "epoch": 16.201966041108133,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0031769104945435947,
      "loss": 0.824,
      "num_input_tokens_seen": 63158352,
      "step": 108780
    },
    {
      "epoch": 16.202710753649093,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0031757107555553028,
      "loss": 0.7922,
      "num_input_tokens_seen": 63161008,
      "step": 108785
    },
    {
      "epoch": 16.20345546619005,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.0031745112163263593,
      "loss": 0.8089,
      "num_input_tokens_seen": 63163888,
      "step": 108790
    },
    {
      "epoch": 16.20420017873101,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0031733118768770257,
      "loss": 0.7916,
      "num_input_tokens_seen": 63166704,
      "step": 108795
    },
    {
      "epoch": 16.204944891271968,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.003172112737227572,
      "loss": 0.793,
      "num_input_tokens_seen": 63169680,
      "step": 108800
    },
    {
      "epoch": 16.205689603812928,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.003170913797398252,
      "loss": 0.8311,
      "num_input_tokens_seen": 63172592,
      "step": 108805
    },
    {
      "epoch": 16.20643431635389,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0031697150574093213,
      "loss": 0.7917,
      "num_input_tokens_seen": 63175344,
      "step": 108810
    },
    {
      "epoch": 16.207179028894846,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0031685165172810267,
      "loss": 0.7878,
      "num_input_tokens_seen": 63178320,
      "step": 108815
    },
    {
      "epoch": 16.207923741435806,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.003167318177033626,
      "loss": 0.7982,
      "num_input_tokens_seen": 63181200,
      "step": 108820
    },
    {
      "epoch": 16.208668453976767,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0031661200366873525,
      "loss": 0.8064,
      "num_input_tokens_seen": 63184176,
      "step": 108825
    },
    {
      "epoch": 16.209413166517724,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0031649220962624586,
      "loss": 0.7864,
      "num_input_tokens_seen": 63187248,
      "step": 108830
    },
    {
      "epoch": 16.210157879058684,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.003163724355779176,
      "loss": 0.7646,
      "num_input_tokens_seen": 63190256,
      "step": 108835
    },
    {
      "epoch": 16.21090259159964,
      "grad_norm": 0.0223388671875,
      "learning_rate": 0.003162526815257739,
      "loss": 0.8021,
      "num_input_tokens_seen": 63192944,
      "step": 108840
    },
    {
      "epoch": 16.2116473041406,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0031613294747183827,
      "loss": 0.8016,
      "num_input_tokens_seen": 63195664,
      "step": 108845
    },
    {
      "epoch": 16.212392016681562,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.003160132334181328,
      "loss": 0.7848,
      "num_input_tokens_seen": 63198608,
      "step": 108850
    },
    {
      "epoch": 16.21313672922252,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0031589353936668082,
      "loss": 0.8066,
      "num_input_tokens_seen": 63201520,
      "step": 108855
    },
    {
      "epoch": 16.21388144176348,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0031577386531950378,
      "loss": 0.8017,
      "num_input_tokens_seen": 63204464,
      "step": 108860
    },
    {
      "epoch": 16.214626154304437,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00315654211278624,
      "loss": 0.7818,
      "num_input_tokens_seen": 63207696,
      "step": 108865
    },
    {
      "epoch": 16.215370866845397,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.003155345772460626,
      "loss": 0.8066,
      "num_input_tokens_seen": 63210768,
      "step": 108870
    },
    {
      "epoch": 16.216115579386358,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.003154149632238407,
      "loss": 0.821,
      "num_input_tokens_seen": 63213968,
      "step": 108875
    },
    {
      "epoch": 16.216860291927315,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.003152953692139789,
      "loss": 0.7807,
      "num_input_tokens_seen": 63216656,
      "step": 108880
    },
    {
      "epoch": 16.217605004468275,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.003151757952184975,
      "loss": 0.7927,
      "num_input_tokens_seen": 63219344,
      "step": 108885
    },
    {
      "epoch": 16.218349717009236,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.003150562412394172,
      "loss": 0.7886,
      "num_input_tokens_seen": 63222448,
      "step": 108890
    },
    {
      "epoch": 16.219094429550193,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0031493670727875704,
      "loss": 0.8222,
      "num_input_tokens_seen": 63225584,
      "step": 108895
    },
    {
      "epoch": 16.219839142091153,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.003148171933385372,
      "loss": 0.8073,
      "num_input_tokens_seen": 63228656,
      "step": 108900
    },
    {
      "epoch": 16.22058385463211,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0031469769942077566,
      "loss": 0.7992,
      "num_input_tokens_seen": 63231472,
      "step": 108905
    },
    {
      "epoch": 16.22132856717307,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.003145782255274924,
      "loss": 0.8082,
      "num_input_tokens_seen": 63234416,
      "step": 108910
    },
    {
      "epoch": 16.22207327971403,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0031445877166070527,
      "loss": 0.7801,
      "num_input_tokens_seen": 63237168,
      "step": 108915
    },
    {
      "epoch": 16.22281799225499,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0031433933782243164,
      "loss": 0.7696,
      "num_input_tokens_seen": 63239920,
      "step": 108920
    },
    {
      "epoch": 16.22356270479595,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0031421992401469046,
      "loss": 0.8097,
      "num_input_tokens_seen": 63242800,
      "step": 108925
    },
    {
      "epoch": 16.22430741733691,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0031410053023949802,
      "loss": 0.8009,
      "num_input_tokens_seen": 63246288,
      "step": 108930
    },
    {
      "epoch": 16.225052129877866,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0031398115649887226,
      "loss": 0.7934,
      "num_input_tokens_seen": 63249200,
      "step": 108935
    },
    {
      "epoch": 16.225796842418827,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.003138618027948294,
      "loss": 0.8011,
      "num_input_tokens_seen": 63251984,
      "step": 108940
    },
    {
      "epoch": 16.226541554959784,
      "grad_norm": 0.078125,
      "learning_rate": 0.003137424691293858,
      "loss": 0.8048,
      "num_input_tokens_seen": 63254864,
      "step": 108945
    },
    {
      "epoch": 16.227286267500745,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.003136231555045575,
      "loss": 0.7876,
      "num_input_tokens_seen": 63257616,
      "step": 108950
    },
    {
      "epoch": 16.228030980041705,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0031350386192236,
      "loss": 0.8115,
      "num_input_tokens_seen": 63260336,
      "step": 108955
    },
    {
      "epoch": 16.228775692582662,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0031338458838480908,
      "loss": 0.7999,
      "num_input_tokens_seen": 63263120,
      "step": 108960
    },
    {
      "epoch": 16.229520405123623,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0031326533489391913,
      "loss": 0.7947,
      "num_input_tokens_seen": 63265840,
      "step": 108965
    },
    {
      "epoch": 16.230265117664583,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.003131461014517058,
      "loss": 0.806,
      "num_input_tokens_seen": 63268400,
      "step": 108970
    },
    {
      "epoch": 16.23100983020554,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.003130268880601822,
      "loss": 0.7937,
      "num_input_tokens_seen": 63271280,
      "step": 108975
    },
    {
      "epoch": 16.2317545427465,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.003129076947213634,
      "loss": 0.8124,
      "num_input_tokens_seen": 63273936,
      "step": 108980
    },
    {
      "epoch": 16.232499255287458,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.003127885214372622,
      "loss": 0.8055,
      "num_input_tokens_seen": 63276656,
      "step": 108985
    },
    {
      "epoch": 16.233243967828418,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0031266936820989275,
      "loss": 0.8,
      "num_input_tokens_seen": 63279536,
      "step": 108990
    },
    {
      "epoch": 16.23398868036938,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.003125502350412675,
      "loss": 0.8375,
      "num_input_tokens_seen": 63282320,
      "step": 108995
    },
    {
      "epoch": 16.234733392910336,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.003124311219333988,
      "loss": 0.7958,
      "num_input_tokens_seen": 63285264,
      "step": 109000
    },
    {
      "epoch": 16.235478105451296,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0031231202888829966,
      "loss": 0.8037,
      "num_input_tokens_seen": 63287856,
      "step": 109005
    },
    {
      "epoch": 16.236222817992257,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0031219295590798174,
      "loss": 0.7907,
      "num_input_tokens_seen": 63290320,
      "step": 109010
    },
    {
      "epoch": 16.236967530533214,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.003120739029944565,
      "loss": 0.7898,
      "num_input_tokens_seen": 63293136,
      "step": 109015
    },
    {
      "epoch": 16.237712243074174,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.003119548701497349,
      "loss": 0.787,
      "num_input_tokens_seen": 63296304,
      "step": 109020
    },
    {
      "epoch": 16.23845695561513,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0031183585737582886,
      "loss": 0.786,
      "num_input_tokens_seen": 63299344,
      "step": 109025
    },
    {
      "epoch": 16.239201668156092,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0031171686467474833,
      "loss": 0.784,
      "num_input_tokens_seen": 63302384,
      "step": 109030
    },
    {
      "epoch": 16.239946380697052,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.003115978920485033,
      "loss": 0.7871,
      "num_input_tokens_seen": 63305264,
      "step": 109035
    },
    {
      "epoch": 16.24069109323801,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0031147893949910444,
      "loss": 0.7995,
      "num_input_tokens_seen": 63308240,
      "step": 109040
    },
    {
      "epoch": 16.24143580577897,
      "grad_norm": 0.024658203125,
      "learning_rate": 0.003113600070285605,
      "loss": 0.8143,
      "num_input_tokens_seen": 63311280,
      "step": 109045
    },
    {
      "epoch": 16.242180518319927,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.003112410946388815,
      "loss": 0.8015,
      "num_input_tokens_seen": 63314128,
      "step": 109050
    },
    {
      "epoch": 16.242925230860887,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.003111222023320757,
      "loss": 0.7954,
      "num_input_tokens_seen": 63316848,
      "step": 109055
    },
    {
      "epoch": 16.243669943401848,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.0031100333011015234,
      "loss": 0.786,
      "num_input_tokens_seen": 63319696,
      "step": 109060
    },
    {
      "epoch": 16.244414655942805,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0031088447797511915,
      "loss": 0.8045,
      "num_input_tokens_seen": 63322512,
      "step": 109065
    },
    {
      "epoch": 16.245159368483765,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0031076564592898395,
      "loss": 0.7903,
      "num_input_tokens_seen": 63325328,
      "step": 109070
    },
    {
      "epoch": 16.245904081024726,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0031064683397375474,
      "loss": 0.8161,
      "num_input_tokens_seen": 63328080,
      "step": 109075
    },
    {
      "epoch": 16.246648793565683,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.0031052804211143853,
      "loss": 0.7972,
      "num_input_tokens_seen": 63331184,
      "step": 109080
    },
    {
      "epoch": 16.247393506106643,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.003104092703440422,
      "loss": 0.7999,
      "num_input_tokens_seen": 63333936,
      "step": 109085
    },
    {
      "epoch": 16.2481382186476,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0031029051867357163,
      "loss": 0.7954,
      "num_input_tokens_seen": 63336848,
      "step": 109090
    },
    {
      "epoch": 16.24888293118856,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0031017178710203413,
      "loss": 0.7891,
      "num_input_tokens_seen": 63339888,
      "step": 109095
    },
    {
      "epoch": 16.24962764372952,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.003100530756314345,
      "loss": 0.7948,
      "num_input_tokens_seen": 63342960,
      "step": 109100
    },
    {
      "epoch": 16.25037235627048,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0030993438426377903,
      "loss": 0.7971,
      "num_input_tokens_seen": 63346224,
      "step": 109105
    },
    {
      "epoch": 16.25111706881144,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0030981571300107247,
      "loss": 0.7982,
      "num_input_tokens_seen": 63349136,
      "step": 109110
    },
    {
      "epoch": 16.2518617813524,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0030969706184532015,
      "loss": 0.7919,
      "num_input_tokens_seen": 63352240,
      "step": 109115
    },
    {
      "epoch": 16.252606493893357,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0030957843079852596,
      "loss": 0.7915,
      "num_input_tokens_seen": 63354928,
      "step": 109120
    },
    {
      "epoch": 16.253351206434317,
      "grad_norm": 0.048828125,
      "learning_rate": 0.003094598198626942,
      "loss": 0.8208,
      "num_input_tokens_seen": 63357744,
      "step": 109125
    },
    {
      "epoch": 16.254095918975274,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.003093412290398289,
      "loss": 0.7868,
      "num_input_tokens_seen": 63360464,
      "step": 109130
    },
    {
      "epoch": 16.254840631516235,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.003092226583319336,
      "loss": 0.8369,
      "num_input_tokens_seen": 63363600,
      "step": 109135
    },
    {
      "epoch": 16.255585344057195,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0030910410774101127,
      "loss": 0.7954,
      "num_input_tokens_seen": 63366512,
      "step": 109140
    },
    {
      "epoch": 16.256330056598152,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.0030898557726906405,
      "loss": 0.8446,
      "num_input_tokens_seen": 63369488,
      "step": 109145
    },
    {
      "epoch": 16.257074769139113,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0030886706691809554,
      "loss": 0.814,
      "num_input_tokens_seen": 63372368,
      "step": 109150
    },
    {
      "epoch": 16.257819481680073,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0030874857669010735,
      "loss": 0.786,
      "num_input_tokens_seen": 63375216,
      "step": 109155
    },
    {
      "epoch": 16.25856419422103,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.0030863010658710084,
      "loss": 0.801,
      "num_input_tokens_seen": 63378064,
      "step": 109160
    },
    {
      "epoch": 16.25930890676199,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0030851165661107804,
      "loss": 0.8064,
      "num_input_tokens_seen": 63380848,
      "step": 109165
    },
    {
      "epoch": 16.260053619302948,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0030839322676403957,
      "loss": 0.7946,
      "num_input_tokens_seen": 63383696,
      "step": 109170
    },
    {
      "epoch": 16.260798331843908,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0030827481704798682,
      "loss": 0.8039,
      "num_input_tokens_seen": 63386544,
      "step": 109175
    },
    {
      "epoch": 16.26154304438487,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.003081564274649195,
      "loss": 0.7808,
      "num_input_tokens_seen": 63389520,
      "step": 109180
    },
    {
      "epoch": 16.262287756925826,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.003080380580168382,
      "loss": 0.808,
      "num_input_tokens_seen": 63392240,
      "step": 109185
    },
    {
      "epoch": 16.263032469466786,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.003079197087057423,
      "loss": 0.8132,
      "num_input_tokens_seen": 63395280,
      "step": 109190
    },
    {
      "epoch": 16.263777182007743,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.0030780137953363156,
      "loss": 0.7956,
      "num_input_tokens_seen": 63398288,
      "step": 109195
    },
    {
      "epoch": 16.264521894548704,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0030768307050250488,
      "loss": 0.8133,
      "num_input_tokens_seen": 63401040,
      "step": 109200
    },
    {
      "epoch": 16.265266607089664,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0030756478161436097,
      "loss": 0.7923,
      "num_input_tokens_seen": 63403696,
      "step": 109205
    },
    {
      "epoch": 16.26601131963062,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0030744651287119805,
      "loss": 0.8381,
      "num_input_tokens_seen": 63406640,
      "step": 109210
    },
    {
      "epoch": 16.266756032171582,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0030732826427501375,
      "loss": 0.8199,
      "num_input_tokens_seen": 63409680,
      "step": 109215
    },
    {
      "epoch": 16.267500744712542,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0030721003582780673,
      "loss": 0.7812,
      "num_input_tokens_seen": 63412496,
      "step": 109220
    },
    {
      "epoch": 16.2682454572535,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0030709182753157365,
      "loss": 0.8139,
      "num_input_tokens_seen": 63415632,
      "step": 109225
    },
    {
      "epoch": 16.26899016979446,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0030697363938831184,
      "loss": 0.8151,
      "num_input_tokens_seen": 63418512,
      "step": 109230
    },
    {
      "epoch": 16.269734882335417,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0030685547140001804,
      "loss": 0.7954,
      "num_input_tokens_seen": 63421328,
      "step": 109235
    },
    {
      "epoch": 16.270479594876377,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.003067373235686879,
      "loss": 0.7914,
      "num_input_tokens_seen": 63424336,
      "step": 109240
    },
    {
      "epoch": 16.271224307417338,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0030661919589631834,
      "loss": 0.7951,
      "num_input_tokens_seen": 63427312,
      "step": 109245
    },
    {
      "epoch": 16.271969019958295,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0030650108838490425,
      "loss": 0.8023,
      "num_input_tokens_seen": 63430288,
      "step": 109250
    },
    {
      "epoch": 16.272713732499255,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.003063830010364414,
      "loss": 0.797,
      "num_input_tokens_seen": 63433136,
      "step": 109255
    },
    {
      "epoch": 16.273458445040216,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.003062649338529245,
      "loss": 0.7884,
      "num_input_tokens_seen": 63435696,
      "step": 109260
    },
    {
      "epoch": 16.274203157581173,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.003061468868363485,
      "loss": 0.8007,
      "num_input_tokens_seen": 63439088,
      "step": 109265
    },
    {
      "epoch": 16.274947870122134,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0030602885998870758,
      "loss": 0.7905,
      "num_input_tokens_seen": 63441968,
      "step": 109270
    },
    {
      "epoch": 16.27569258266309,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.003059108533119954,
      "loss": 0.8017,
      "num_input_tokens_seen": 63444816,
      "step": 109275
    },
    {
      "epoch": 16.27643729520405,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.0030579286680820585,
      "loss": 0.8216,
      "num_input_tokens_seen": 63447504,
      "step": 109280
    },
    {
      "epoch": 16.27718200774501,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.0030567490047933178,
      "loss": 0.8244,
      "num_input_tokens_seen": 63450480,
      "step": 109285
    },
    {
      "epoch": 16.27792672028597,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0030555695432736662,
      "loss": 0.8339,
      "num_input_tokens_seen": 63453680,
      "step": 109290
    },
    {
      "epoch": 16.27867143282693,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.003054390283543025,
      "loss": 0.7929,
      "num_input_tokens_seen": 63456624,
      "step": 109295
    },
    {
      "epoch": 16.27941614536789,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.003053211225621322,
      "loss": 0.8079,
      "num_input_tokens_seen": 63459440,
      "step": 109300
    },
    {
      "epoch": 16.280160857908847,
      "grad_norm": 0.0625,
      "learning_rate": 0.0030520323695284696,
      "loss": 0.7949,
      "num_input_tokens_seen": 63462128,
      "step": 109305
    },
    {
      "epoch": 16.280905570449807,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0030508537152843913,
      "loss": 0.8297,
      "num_input_tokens_seen": 63464784,
      "step": 109310
    },
    {
      "epoch": 16.281650282990764,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0030496752629089944,
      "loss": 0.7975,
      "num_input_tokens_seen": 63467472,
      "step": 109315
    },
    {
      "epoch": 16.282394995531725,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0030484970124221838,
      "loss": 0.8182,
      "num_input_tokens_seen": 63470288,
      "step": 109320
    },
    {
      "epoch": 16.283139708072685,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.003047318963843873,
      "loss": 0.804,
      "num_input_tokens_seen": 63472848,
      "step": 109325
    },
    {
      "epoch": 16.283884420613642,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.003046141117193956,
      "loss": 0.7837,
      "num_input_tokens_seen": 63475984,
      "step": 109330
    },
    {
      "epoch": 16.284629133154603,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0030449634724923407,
      "loss": 0.8113,
      "num_input_tokens_seen": 63478672,
      "step": 109335
    },
    {
      "epoch": 16.285373845695563,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0030437860297589173,
      "loss": 0.8092,
      "num_input_tokens_seen": 63481424,
      "step": 109340
    },
    {
      "epoch": 16.28611855823652,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0030426087890135754,
      "loss": 0.7936,
      "num_input_tokens_seen": 63484048,
      "step": 109345
    },
    {
      "epoch": 16.28686327077748,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.003041431750276202,
      "loss": 0.8018,
      "num_input_tokens_seen": 63487152,
      "step": 109350
    },
    {
      "epoch": 16.287607983318438,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.00304025491356669,
      "loss": 0.8213,
      "num_input_tokens_seen": 63490064,
      "step": 109355
    },
    {
      "epoch": 16.2883526958594,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0030390782789049145,
      "loss": 0.7894,
      "num_input_tokens_seen": 63493008,
      "step": 109360
    },
    {
      "epoch": 16.28909740840036,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.00303790184631075,
      "loss": 0.7981,
      "num_input_tokens_seen": 63495824,
      "step": 109365
    },
    {
      "epoch": 16.289842120941316,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0030367256158040827,
      "loss": 0.7896,
      "num_input_tokens_seen": 63498384,
      "step": 109370
    },
    {
      "epoch": 16.290586833482276,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.003035549587404771,
      "loss": 0.8058,
      "num_input_tokens_seen": 63501264,
      "step": 109375
    },
    {
      "epoch": 16.291331546023233,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.003034373761132694,
      "loss": 0.7932,
      "num_input_tokens_seen": 63504016,
      "step": 109380
    },
    {
      "epoch": 16.292076258564194,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0030331981370077064,
      "loss": 0.7822,
      "num_input_tokens_seen": 63506800,
      "step": 109385
    },
    {
      "epoch": 16.292820971105154,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0030320227150496784,
      "loss": 0.797,
      "num_input_tokens_seen": 63509872,
      "step": 109390
    },
    {
      "epoch": 16.29356568364611,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.003030847495278463,
      "loss": 0.7953,
      "num_input_tokens_seen": 63512880,
      "step": 109395
    },
    {
      "epoch": 16.294310396187072,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0030296724777139103,
      "loss": 0.8159,
      "num_input_tokens_seen": 63515824,
      "step": 109400
    },
    {
      "epoch": 16.295055108728032,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.0030284976623758783,
      "loss": 0.7861,
      "num_input_tokens_seen": 63518864,
      "step": 109405
    },
    {
      "epoch": 16.29579982126899,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.003027323049284211,
      "loss": 0.8263,
      "num_input_tokens_seen": 63521872,
      "step": 109410
    },
    {
      "epoch": 16.29654453380995,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0030261486384587527,
      "loss": 0.796,
      "num_input_tokens_seen": 63525072,
      "step": 109415
    },
    {
      "epoch": 16.297289246350907,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0030249744299193403,
      "loss": 0.7918,
      "num_input_tokens_seen": 63528048,
      "step": 109420
    },
    {
      "epoch": 16.298033958891867,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0030238004236858166,
      "loss": 0.7936,
      "num_input_tokens_seen": 63530960,
      "step": 109425
    },
    {
      "epoch": 16.298778671432828,
      "grad_norm": 0.0233154296875,
      "learning_rate": 0.0030226266197780093,
      "loss": 0.7968,
      "num_input_tokens_seen": 63533968,
      "step": 109430
    },
    {
      "epoch": 16.299523383973785,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0030214530182157556,
      "loss": 0.7786,
      "num_input_tokens_seen": 63536464,
      "step": 109435
    },
    {
      "epoch": 16.300268096514746,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0030202796190188794,
      "loss": 0.8063,
      "num_input_tokens_seen": 63539440,
      "step": 109440
    },
    {
      "epoch": 16.301012809055706,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.003019106422207199,
      "loss": 0.7894,
      "num_input_tokens_seen": 63542128,
      "step": 109445
    },
    {
      "epoch": 16.301757521596663,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0030179334278005424,
      "loss": 0.7885,
      "num_input_tokens_seen": 63545040,
      "step": 109450
    },
    {
      "epoch": 16.302502234137624,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.00301676063581872,
      "loss": 0.8024,
      "num_input_tokens_seen": 63547824,
      "step": 109455
    },
    {
      "epoch": 16.30324694667858,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.00301558804628155,
      "loss": 0.8175,
      "num_input_tokens_seen": 63550928,
      "step": 109460
    },
    {
      "epoch": 16.30399165921954,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0030144156592088397,
      "loss": 0.7969,
      "num_input_tokens_seen": 63554224,
      "step": 109465
    },
    {
      "epoch": 16.3047363717605,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0030132434746203943,
      "loss": 0.7996,
      "num_input_tokens_seen": 63557520,
      "step": 109470
    },
    {
      "epoch": 16.30548108430146,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.003012071492536019,
      "loss": 0.7992,
      "num_input_tokens_seen": 63560208,
      "step": 109475
    },
    {
      "epoch": 16.30622579684242,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.003010899712975508,
      "loss": 0.7822,
      "num_input_tokens_seen": 63563120,
      "step": 109480
    },
    {
      "epoch": 16.30697050938338,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.003009728135958664,
      "loss": 0.8135,
      "num_input_tokens_seen": 63565968,
      "step": 109485
    },
    {
      "epoch": 16.307715221924337,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.003008556761505274,
      "loss": 0.822,
      "num_input_tokens_seen": 63568752,
      "step": 109490
    },
    {
      "epoch": 16.308459934465297,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.003007385589635133,
      "loss": 0.799,
      "num_input_tokens_seen": 63571664,
      "step": 109495
    },
    {
      "epoch": 16.309204647006254,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.003006214620368021,
      "loss": 0.802,
      "num_input_tokens_seen": 63574416,
      "step": 109500
    },
    {
      "epoch": 16.309949359547215,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.003005043853723726,
      "loss": 0.7955,
      "num_input_tokens_seen": 63577520,
      "step": 109505
    },
    {
      "epoch": 16.310694072088175,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0030038732897220205,
      "loss": 0.7869,
      "num_input_tokens_seen": 63580400,
      "step": 109510
    },
    {
      "epoch": 16.311438784629132,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0030027029283826866,
      "loss": 0.7882,
      "num_input_tokens_seen": 63583056,
      "step": 109515
    },
    {
      "epoch": 16.312183497170093,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0030015327697254937,
      "loss": 0.7938,
      "num_input_tokens_seen": 63586544,
      "step": 109520
    },
    {
      "epoch": 16.312928209711053,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0030003628137702065,
      "loss": 0.7888,
      "num_input_tokens_seen": 63589232,
      "step": 109525
    },
    {
      "epoch": 16.31367292225201,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.002999193060536597,
      "loss": 0.8008,
      "num_input_tokens_seen": 63592048,
      "step": 109530
    },
    {
      "epoch": 16.31441763479297,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.002998023510044424,
      "loss": 0.8055,
      "num_input_tokens_seen": 63595184,
      "step": 109535
    },
    {
      "epoch": 16.315162347333928,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.002996854162313446,
      "loss": 0.8188,
      "num_input_tokens_seen": 63598160,
      "step": 109540
    },
    {
      "epoch": 16.31590705987489,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.002995685017363411,
      "loss": 0.7973,
      "num_input_tokens_seen": 63601168,
      "step": 109545
    },
    {
      "epoch": 16.31665177241585,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0029945160752140832,
      "loss": 0.7806,
      "num_input_tokens_seen": 63604176,
      "step": 109550
    },
    {
      "epoch": 16.317396484956806,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.002993347335885205,
      "loss": 0.7872,
      "num_input_tokens_seen": 63606672,
      "step": 109555
    },
    {
      "epoch": 16.318141197497766,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0029921787993965154,
      "loss": 0.806,
      "num_input_tokens_seen": 63609904,
      "step": 109560
    },
    {
      "epoch": 16.318885910038723,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0029910104657677643,
      "loss": 0.8162,
      "num_input_tokens_seen": 63612560,
      "step": 109565
    },
    {
      "epoch": 16.319630622579684,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.002989842335018683,
      "loss": 0.7903,
      "num_input_tokens_seen": 63615664,
      "step": 109570
    },
    {
      "epoch": 16.320375335120644,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.0029886744071690122,
      "loss": 0.7943,
      "num_input_tokens_seen": 63618928,
      "step": 109575
    },
    {
      "epoch": 16.3211200476616,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0029875066822384767,
      "loss": 0.7975,
      "num_input_tokens_seen": 63621936,
      "step": 109580
    },
    {
      "epoch": 16.321864760202562,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0029863391602468095,
      "loss": 0.7841,
      "num_input_tokens_seen": 63624592,
      "step": 109585
    },
    {
      "epoch": 16.322609472743522,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0029851718412137293,
      "loss": 0.8147,
      "num_input_tokens_seen": 63627536,
      "step": 109590
    },
    {
      "epoch": 16.32335418528448,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0029840047251589634,
      "loss": 0.781,
      "num_input_tokens_seen": 63630544,
      "step": 109595
    },
    {
      "epoch": 16.32409889782544,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.002982837812102224,
      "loss": 0.8041,
      "num_input_tokens_seen": 63633328,
      "step": 109600
    },
    {
      "epoch": 16.324843610366397,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.002981671102063227,
      "loss": 0.7919,
      "num_input_tokens_seen": 63636144,
      "step": 109605
    },
    {
      "epoch": 16.325588322907358,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.002980504595061681,
      "loss": 0.8046,
      "num_input_tokens_seen": 63638704,
      "step": 109610
    },
    {
      "epoch": 16.326333035448318,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0029793382911172906,
      "loss": 0.8021,
      "num_input_tokens_seen": 63641744,
      "step": 109615
    },
    {
      "epoch": 16.327077747989275,
      "grad_norm": 0.0234375,
      "learning_rate": 0.0029781721902497654,
      "loss": 0.7896,
      "num_input_tokens_seen": 63645008,
      "step": 109620
    },
    {
      "epoch": 16.327822460530236,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0029770062924787998,
      "loss": 0.7929,
      "num_input_tokens_seen": 63647696,
      "step": 109625
    },
    {
      "epoch": 16.328567173071196,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.002975840597824096,
      "loss": 0.8095,
      "num_input_tokens_seen": 63650416,
      "step": 109630
    },
    {
      "epoch": 16.329311885612153,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.002974675106305346,
      "loss": 0.8071,
      "num_input_tokens_seen": 63653296,
      "step": 109635
    },
    {
      "epoch": 16.330056598153114,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.002973509817942235,
      "loss": 0.808,
      "num_input_tokens_seen": 63656144,
      "step": 109640
    },
    {
      "epoch": 16.33080131069407,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.002972344732754453,
      "loss": 0.7816,
      "num_input_tokens_seen": 63659248,
      "step": 109645
    },
    {
      "epoch": 16.33154602323503,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0029711798507616807,
      "loss": 0.8178,
      "num_input_tokens_seen": 63662032,
      "step": 109650
    },
    {
      "epoch": 16.33229073577599,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0029700151719836035,
      "loss": 0.7989,
      "num_input_tokens_seen": 63664944,
      "step": 109655
    },
    {
      "epoch": 16.33303544831695,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.002968850696439889,
      "loss": 0.8076,
      "num_input_tokens_seen": 63668016,
      "step": 109660
    },
    {
      "epoch": 16.33378016085791,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0029676864241502164,
      "loss": 0.7783,
      "num_input_tokens_seen": 63671152,
      "step": 109665
    },
    {
      "epoch": 16.33452487339887,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.002966522355134255,
      "loss": 0.8111,
      "num_input_tokens_seen": 63673872,
      "step": 109670
    },
    {
      "epoch": 16.335269585939827,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0029653584894116674,
      "loss": 0.7769,
      "num_input_tokens_seen": 63676656,
      "step": 109675
    },
    {
      "epoch": 16.336014298480787,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.002964194827002116,
      "loss": 0.7967,
      "num_input_tokens_seen": 63679792,
      "step": 109680
    },
    {
      "epoch": 16.336759011021744,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0029630313679252555,
      "loss": 0.7868,
      "num_input_tokens_seen": 63682736,
      "step": 109685
    },
    {
      "epoch": 16.337503723562705,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0029618681122007516,
      "loss": 0.8248,
      "num_input_tokens_seen": 63685776,
      "step": 109690
    },
    {
      "epoch": 16.338248436103665,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.002960705059848247,
      "loss": 0.7909,
      "num_input_tokens_seen": 63688912,
      "step": 109695
    },
    {
      "epoch": 16.338993148644622,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0029595422108873973,
      "loss": 0.8043,
      "num_input_tokens_seen": 63692048,
      "step": 109700
    },
    {
      "epoch": 16.339737861185583,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.002958379565337841,
      "loss": 0.8022,
      "num_input_tokens_seen": 63694864,
      "step": 109705
    },
    {
      "epoch": 16.34048257372654,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0029572171232192276,
      "loss": 0.7943,
      "num_input_tokens_seen": 63697904,
      "step": 109710
    },
    {
      "epoch": 16.3412272862675,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.002956054884551188,
      "loss": 0.8167,
      "num_input_tokens_seen": 63700720,
      "step": 109715
    },
    {
      "epoch": 16.34197199880846,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.002954892849353362,
      "loss": 0.8041,
      "num_input_tokens_seen": 63703504,
      "step": 109720
    },
    {
      "epoch": 16.342716711349418,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.002953731017645381,
      "loss": 0.7948,
      "num_input_tokens_seen": 63706192,
      "step": 109725
    },
    {
      "epoch": 16.34346142389038,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.002952569389446866,
      "loss": 0.8004,
      "num_input_tokens_seen": 63708944,
      "step": 109730
    },
    {
      "epoch": 16.34420613643134,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0029514079647774516,
      "loss": 0.792,
      "num_input_tokens_seen": 63711664,
      "step": 109735
    },
    {
      "epoch": 16.344950848972296,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.002950246743656755,
      "loss": 0.8065,
      "num_input_tokens_seen": 63714608,
      "step": 109740
    },
    {
      "epoch": 16.345695561513256,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0029490857261043916,
      "loss": 0.7874,
      "num_input_tokens_seen": 63717648,
      "step": 109745
    },
    {
      "epoch": 16.346440274054213,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.002947924912139972,
      "loss": 0.8046,
      "num_input_tokens_seen": 63720528,
      "step": 109750
    },
    {
      "epoch": 16.347184986595174,
      "grad_norm": 0.130859375,
      "learning_rate": 0.002946764301783116,
      "loss": 0.7904,
      "num_input_tokens_seen": 63723472,
      "step": 109755
    },
    {
      "epoch": 16.347929699136134,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.002945603895053429,
      "loss": 0.7909,
      "num_input_tokens_seen": 63726160,
      "step": 109760
    },
    {
      "epoch": 16.34867441167709,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.002944443691970506,
      "loss": 0.7885,
      "num_input_tokens_seen": 63729200,
      "step": 109765
    },
    {
      "epoch": 16.349419124218052,
      "grad_norm": 0.052734375,
      "learning_rate": 0.002943283692553959,
      "loss": 0.7875,
      "num_input_tokens_seen": 63731888,
      "step": 109770
    },
    {
      "epoch": 16.350163836759013,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0029421238968233765,
      "loss": 0.7928,
      "num_input_tokens_seen": 63734608,
      "step": 109775
    },
    {
      "epoch": 16.35090854929997,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0029409643047983596,
      "loss": 0.7934,
      "num_input_tokens_seen": 63737616,
      "step": 109780
    },
    {
      "epoch": 16.35165326184093,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.002939804916498491,
      "loss": 0.7882,
      "num_input_tokens_seen": 63740336,
      "step": 109785
    },
    {
      "epoch": 16.352397974381887,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0029386457319433637,
      "loss": 0.7843,
      "num_input_tokens_seen": 63742960,
      "step": 109790
    },
    {
      "epoch": 16.353142686922848,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0029374867511525586,
      "loss": 0.7824,
      "num_input_tokens_seen": 63745776,
      "step": 109795
    },
    {
      "epoch": 16.353887399463808,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0029363279741456564,
      "loss": 0.8044,
      "num_input_tokens_seen": 63748528,
      "step": 109800
    },
    {
      "epoch": 16.354632112004765,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.00293516940094223,
      "loss": 0.7785,
      "num_input_tokens_seen": 63751696,
      "step": 109805
    },
    {
      "epoch": 16.355376824545726,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0029340110315618523,
      "loss": 0.8156,
      "num_input_tokens_seen": 63754576,
      "step": 109810
    },
    {
      "epoch": 16.356121537086686,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.002932852866024097,
      "loss": 0.7889,
      "num_input_tokens_seen": 63757456,
      "step": 109815
    },
    {
      "epoch": 16.356866249627643,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0029316949043485256,
      "loss": 0.7976,
      "num_input_tokens_seen": 63760400,
      "step": 109820
    },
    {
      "epoch": 16.357610962168604,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.002930537146554706,
      "loss": 0.7922,
      "num_input_tokens_seen": 63763248,
      "step": 109825
    },
    {
      "epoch": 16.35835567470956,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0029293795926621916,
      "loss": 0.7906,
      "num_input_tokens_seen": 63766256,
      "step": 109830
    },
    {
      "epoch": 16.35910038725052,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0029282222426905446,
      "loss": 0.8106,
      "num_input_tokens_seen": 63768944,
      "step": 109835
    },
    {
      "epoch": 16.35984509979148,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.002927065096659312,
      "loss": 0.7873,
      "num_input_tokens_seen": 63771664,
      "step": 109840
    },
    {
      "epoch": 16.36058981233244,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0029259081545880417,
      "loss": 0.7967,
      "num_input_tokens_seen": 63774448,
      "step": 109845
    },
    {
      "epoch": 16.3613345248734,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.002924751416496286,
      "loss": 0.7879,
      "num_input_tokens_seen": 63777264,
      "step": 109850
    },
    {
      "epoch": 16.36207923741436,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0029235948824035767,
      "loss": 0.8091,
      "num_input_tokens_seen": 63780112,
      "step": 109855
    },
    {
      "epoch": 16.362823949955317,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.002922438552329464,
      "loss": 0.7795,
      "num_input_tokens_seen": 63782832,
      "step": 109860
    },
    {
      "epoch": 16.363568662496277,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.002921282426293475,
      "loss": 0.7757,
      "num_input_tokens_seen": 63785936,
      "step": 109865
    },
    {
      "epoch": 16.364313375037234,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0029201265043151447,
      "loss": 0.8183,
      "num_input_tokens_seen": 63788848,
      "step": 109870
    },
    {
      "epoch": 16.365058087578195,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.0029189707864139946,
      "loss": 0.7921,
      "num_input_tokens_seen": 63791856,
      "step": 109875
    },
    {
      "epoch": 16.365802800119155,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.002917815272609559,
      "loss": 0.8077,
      "num_input_tokens_seen": 63794672,
      "step": 109880
    },
    {
      "epoch": 16.366547512660112,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.002916659962921354,
      "loss": 0.8032,
      "num_input_tokens_seen": 63797648,
      "step": 109885
    },
    {
      "epoch": 16.367292225201073,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0029155048573688957,
      "loss": 0.8005,
      "num_input_tokens_seen": 63800848,
      "step": 109890
    },
    {
      "epoch": 16.36803693774203,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0029143499559717034,
      "loss": 0.7979,
      "num_input_tokens_seen": 63803952,
      "step": 109895
    },
    {
      "epoch": 16.36878165028299,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0029131952587492815,
      "loss": 0.8018,
      "num_input_tokens_seen": 63806672,
      "step": 109900
    },
    {
      "epoch": 16.36952636282395,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0029120407657211443,
      "loss": 0.7815,
      "num_input_tokens_seen": 63809616,
      "step": 109905
    },
    {
      "epoch": 16.370271075364908,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0029108864769067898,
      "loss": 0.7866,
      "num_input_tokens_seen": 63812336,
      "step": 109910
    },
    {
      "epoch": 16.37101578790587,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.002909732392325724,
      "loss": 0.8105,
      "num_input_tokens_seen": 63815568,
      "step": 109915
    },
    {
      "epoch": 16.37176050044683,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0029085785119974425,
      "loss": 0.7842,
      "num_input_tokens_seen": 63818224,
      "step": 109920
    },
    {
      "epoch": 16.372505212987786,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0029074248359414337,
      "loss": 0.8032,
      "num_input_tokens_seen": 63821072,
      "step": 109925
    },
    {
      "epoch": 16.373249925528746,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.002906271364177197,
      "loss": 0.8014,
      "num_input_tokens_seen": 63824688,
      "step": 109930
    },
    {
      "epoch": 16.373994638069703,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0029051180967242128,
      "loss": 0.786,
      "num_input_tokens_seen": 63827952,
      "step": 109935
    },
    {
      "epoch": 16.374739350610664,
      "grad_norm": 0.0250244140625,
      "learning_rate": 0.002903965033601966,
      "loss": 0.8011,
      "num_input_tokens_seen": 63830832,
      "step": 109940
    },
    {
      "epoch": 16.375484063151625,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.002902812174829933,
      "loss": 0.7939,
      "num_input_tokens_seen": 63833776,
      "step": 109945
    },
    {
      "epoch": 16.37622877569258,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0029016595204275956,
      "loss": 0.8043,
      "num_input_tokens_seen": 63836592,
      "step": 109950
    },
    {
      "epoch": 16.376973488233542,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0029005070704144226,
      "loss": 0.7919,
      "num_input_tokens_seen": 63839504,
      "step": 109955
    },
    {
      "epoch": 16.377718200774503,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.0028993548248098877,
      "loss": 0.7987,
      "num_input_tokens_seen": 63842320,
      "step": 109960
    },
    {
      "epoch": 16.37846291331546,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.002898202783633454,
      "loss": 0.8095,
      "num_input_tokens_seen": 63845040,
      "step": 109965
    },
    {
      "epoch": 16.37920762585642,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0028970509469045808,
      "loss": 0.7784,
      "num_input_tokens_seen": 63847920,
      "step": 109970
    },
    {
      "epoch": 16.379952338397377,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.002895899314642736,
      "loss": 0.8032,
      "num_input_tokens_seen": 63850864,
      "step": 109975
    },
    {
      "epoch": 16.380697050938338,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.002894747886867365,
      "loss": 0.8077,
      "num_input_tokens_seen": 63854128,
      "step": 109980
    },
    {
      "epoch": 16.381441763479298,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0028935966635979305,
      "loss": 0.7918,
      "num_input_tokens_seen": 63857232,
      "step": 109985
    },
    {
      "epoch": 16.382186476020255,
      "grad_norm": 0.0546875,
      "learning_rate": 0.002892445644853872,
      "loss": 0.7978,
      "num_input_tokens_seen": 63859888,
      "step": 109990
    },
    {
      "epoch": 16.382931188561216,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0028912948306546414,
      "loss": 0.8073,
      "num_input_tokens_seen": 63862832,
      "step": 109995
    },
    {
      "epoch": 16.383675901102176,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.002890144221019678,
      "loss": 0.8026,
      "num_input_tokens_seen": 63865616,
      "step": 110000
    },
    {
      "epoch": 16.384420613643133,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.002888993815968421,
      "loss": 0.7954,
      "num_input_tokens_seen": 63868496,
      "step": 110005
    },
    {
      "epoch": 16.385165326184094,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.002887843615520304,
      "loss": 0.8006,
      "num_input_tokens_seen": 63871792,
      "step": 110010
    },
    {
      "epoch": 16.38591003872505,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0028866936196947557,
      "loss": 0.7937,
      "num_input_tokens_seen": 63874352,
      "step": 110015
    },
    {
      "epoch": 16.38665475126601,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0028855438285112106,
      "loss": 0.7847,
      "num_input_tokens_seen": 63877264,
      "step": 110020
    },
    {
      "epoch": 16.38739946380697,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0028843942419890848,
      "loss": 0.7868,
      "num_input_tokens_seen": 63880368,
      "step": 110025
    },
    {
      "epoch": 16.38814417634793,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00288324486014781,
      "loss": 0.805,
      "num_input_tokens_seen": 63883088,
      "step": 110030
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.002882095683006795,
      "loss": 0.798,
      "num_input_tokens_seen": 63885744,
      "step": 110035
    },
    {
      "epoch": 16.38963360142985,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0028809467105854595,
      "loss": 0.7991,
      "num_input_tokens_seen": 63888624,
      "step": 110040
    },
    {
      "epoch": 16.390378313970807,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0028797979429032133,
      "loss": 0.7934,
      "num_input_tokens_seen": 63891664,
      "step": 110045
    },
    {
      "epoch": 16.391123026511767,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0028786493799794572,
      "loss": 0.7959,
      "num_input_tokens_seen": 63894416,
      "step": 110050
    },
    {
      "epoch": 16.391867739052724,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0028775010218336054,
      "loss": 0.7929,
      "num_input_tokens_seen": 63897520,
      "step": 110055
    },
    {
      "epoch": 16.392612451593685,
      "grad_norm": 0.02734375,
      "learning_rate": 0.0028763528684850495,
      "loss": 0.7935,
      "num_input_tokens_seen": 63900336,
      "step": 110060
    },
    {
      "epoch": 16.393357164134645,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0028752049199531923,
      "loss": 0.7935,
      "num_input_tokens_seen": 63903280,
      "step": 110065
    },
    {
      "epoch": 16.394101876675602,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.002874057176257427,
      "loss": 0.8018,
      "num_input_tokens_seen": 63906224,
      "step": 110070
    },
    {
      "epoch": 16.394846589216563,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0028729096374171385,
      "loss": 0.783,
      "num_input_tokens_seen": 63908944,
      "step": 110075
    },
    {
      "epoch": 16.39559130175752,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.002871762303451719,
      "loss": 0.8077,
      "num_input_tokens_seen": 63911728,
      "step": 110080
    },
    {
      "epoch": 16.39633601429848,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.002870615174380543,
      "loss": 0.7644,
      "num_input_tokens_seen": 63914480,
      "step": 110085
    },
    {
      "epoch": 16.39708072683944,
      "grad_norm": 0.05078125,
      "learning_rate": 0.002869468250223001,
      "loss": 0.7913,
      "num_input_tokens_seen": 63917360,
      "step": 110090
    },
    {
      "epoch": 16.397825439380398,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0028683215309984597,
      "loss": 0.7846,
      "num_input_tokens_seen": 63920368,
      "step": 110095
    },
    {
      "epoch": 16.39857015192136,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0028671750167263005,
      "loss": 0.8084,
      "num_input_tokens_seen": 63923312,
      "step": 110100
    },
    {
      "epoch": 16.39931486446232,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.002866028707425884,
      "loss": 0.8131,
      "num_input_tokens_seen": 63926032,
      "step": 110105
    },
    {
      "epoch": 16.400059577003276,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0028648826031165827,
      "loss": 0.7871,
      "num_input_tokens_seen": 63928944,
      "step": 110110
    },
    {
      "epoch": 16.400804289544237,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.002863736703817754,
      "loss": 0.7928,
      "num_input_tokens_seen": 63931824,
      "step": 110115
    },
    {
      "epoch": 16.401549002085194,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0028625910095487614,
      "loss": 0.8018,
      "num_input_tokens_seen": 63934672,
      "step": 110120
    },
    {
      "epoch": 16.402293714626154,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0028614455203289585,
      "loss": 0.7899,
      "num_input_tokens_seen": 63937488,
      "step": 110125
    },
    {
      "epoch": 16.403038427167115,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0028603002361776973,
      "loss": 0.7786,
      "num_input_tokens_seen": 63940624,
      "step": 110130
    },
    {
      "epoch": 16.40378313970807,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0028591551571143245,
      "loss": 0.7982,
      "num_input_tokens_seen": 63943792,
      "step": 110135
    },
    {
      "epoch": 16.404527852249032,
      "grad_norm": 0.078125,
      "learning_rate": 0.0028580102831581816,
      "loss": 0.8012,
      "num_input_tokens_seen": 63946544,
      "step": 110140
    },
    {
      "epoch": 16.405272564789993,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.002856865614328619,
      "loss": 0.7958,
      "num_input_tokens_seen": 63949648,
      "step": 110145
    },
    {
      "epoch": 16.40601727733095,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.002855721150644965,
      "loss": 0.7781,
      "num_input_tokens_seen": 63952720,
      "step": 110150
    },
    {
      "epoch": 16.40676198987191,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0028545768921265646,
      "loss": 0.8229,
      "num_input_tokens_seen": 63955984,
      "step": 110155
    },
    {
      "epoch": 16.407506702412867,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.002853432838792743,
      "loss": 0.8022,
      "num_input_tokens_seen": 63959184,
      "step": 110160
    },
    {
      "epoch": 16.408251414953828,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0028522889906628258,
      "loss": 0.7962,
      "num_input_tokens_seen": 63961904,
      "step": 110165
    },
    {
      "epoch": 16.408996127494788,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0028511453477561414,
      "loss": 0.7843,
      "num_input_tokens_seen": 63964496,
      "step": 110170
    },
    {
      "epoch": 16.409740840035745,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0028500019100920052,
      "loss": 0.8008,
      "num_input_tokens_seen": 63967312,
      "step": 110175
    },
    {
      "epoch": 16.410485552576706,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0028488586776897434,
      "loss": 0.7935,
      "num_input_tokens_seen": 63970192,
      "step": 110180
    },
    {
      "epoch": 16.411230265117666,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.00284771565056866,
      "loss": 0.7787,
      "num_input_tokens_seen": 63973264,
      "step": 110185
    },
    {
      "epoch": 16.411974977658623,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0028465728287480735,
      "loss": 0.7872,
      "num_input_tokens_seen": 63976144,
      "step": 110190
    },
    {
      "epoch": 16.412719690199584,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.002845430212247286,
      "loss": 0.8212,
      "num_input_tokens_seen": 63978864,
      "step": 110195
    },
    {
      "epoch": 16.41346440274054,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0028442878010856033,
      "loss": 0.8145,
      "num_input_tokens_seen": 63981744,
      "step": 110200
    },
    {
      "epoch": 16.4142091152815,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0028431455952823217,
      "loss": 0.8009,
      "num_input_tokens_seen": 63984880,
      "step": 110205
    },
    {
      "epoch": 16.414953827822462,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0028420035948567373,
      "loss": 0.8073,
      "num_input_tokens_seen": 63987760,
      "step": 110210
    },
    {
      "epoch": 16.41569854036342,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.0028408617998281487,
      "loss": 0.8077,
      "num_input_tokens_seen": 63990576,
      "step": 110215
    },
    {
      "epoch": 16.41644325290438,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0028397202102158374,
      "loss": 0.8144,
      "num_input_tokens_seen": 63993616,
      "step": 110220
    },
    {
      "epoch": 16.417187965445336,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0028385788260390983,
      "loss": 0.7948,
      "num_input_tokens_seen": 63996304,
      "step": 110225
    },
    {
      "epoch": 16.417932677986297,
      "grad_norm": 0.078125,
      "learning_rate": 0.0028374376473172067,
      "loss": 0.7942,
      "num_input_tokens_seen": 63999152,
      "step": 110230
    },
    {
      "epoch": 16.418677390527257,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.002836296674069447,
      "loss": 0.7931,
      "num_input_tokens_seen": 64002032,
      "step": 110235
    },
    {
      "epoch": 16.419422103068214,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.002835155906315094,
      "loss": 0.7912,
      "num_input_tokens_seen": 64004880,
      "step": 110240
    },
    {
      "epoch": 16.420166815609175,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0028340153440734126,
      "loss": 0.7945,
      "num_input_tokens_seen": 64007984,
      "step": 110245
    },
    {
      "epoch": 16.420911528150135,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.0028328749873636817,
      "loss": 0.8336,
      "num_input_tokens_seen": 64011056,
      "step": 110250
    },
    {
      "epoch": 16.421656240691092,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.002831734836205158,
      "loss": 0.7941,
      "num_input_tokens_seen": 64014064,
      "step": 110255
    },
    {
      "epoch": 16.422400953232053,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00283059489061711,
      "loss": 0.8061,
      "num_input_tokens_seen": 64017168,
      "step": 110260
    },
    {
      "epoch": 16.42314566577301,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.002829455150618794,
      "loss": 0.7902,
      "num_input_tokens_seen": 64020016,
      "step": 110265
    },
    {
      "epoch": 16.42389037831397,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.002828315616229463,
      "loss": 0.8408,
      "num_input_tokens_seen": 64022896,
      "step": 110270
    },
    {
      "epoch": 16.42463509085493,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.002827176287468365,
      "loss": 0.7922,
      "num_input_tokens_seen": 64026032,
      "step": 110275
    },
    {
      "epoch": 16.425379803395888,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0028260371643547572,
      "loss": 0.796,
      "num_input_tokens_seen": 64029136,
      "step": 110280
    },
    {
      "epoch": 16.42612451593685,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0028248982469078763,
      "loss": 0.787,
      "num_input_tokens_seen": 64031824,
      "step": 110285
    },
    {
      "epoch": 16.42686922847781,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0028237595351469622,
      "loss": 0.7823,
      "num_input_tokens_seen": 64034896,
      "step": 110290
    },
    {
      "epoch": 16.427613941018766,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0028226210290912593,
      "loss": 0.7762,
      "num_input_tokens_seen": 64037808,
      "step": 110295
    },
    {
      "epoch": 16.428358653559727,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.002821482728759994,
      "loss": 0.7933,
      "num_input_tokens_seen": 64040848,
      "step": 110300
    },
    {
      "epoch": 16.429103366100684,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.002820344634172404,
      "loss": 0.7831,
      "num_input_tokens_seen": 64043696,
      "step": 110305
    },
    {
      "epoch": 16.429848078641644,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0028192067453477075,
      "loss": 0.8119,
      "num_input_tokens_seen": 64046416,
      "step": 110310
    },
    {
      "epoch": 16.430592791182605,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0028180690623051388,
      "loss": 0.8071,
      "num_input_tokens_seen": 64049520,
      "step": 110315
    },
    {
      "epoch": 16.43133750372356,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0028169315850639066,
      "loss": 0.7904,
      "num_input_tokens_seen": 64052592,
      "step": 110320
    },
    {
      "epoch": 16.432082216264522,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0028157943136432376,
      "loss": 0.7873,
      "num_input_tokens_seen": 64055408,
      "step": 110325
    },
    {
      "epoch": 16.432826928805483,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.002814657248062341,
      "loss": 0.8459,
      "num_input_tokens_seen": 64058096,
      "step": 110330
    },
    {
      "epoch": 16.43357164134644,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.002813520388340424,
      "loss": 0.7739,
      "num_input_tokens_seen": 64060784,
      "step": 110335
    },
    {
      "epoch": 16.4343163538874,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.002812383734496695,
      "loss": 0.7932,
      "num_input_tokens_seen": 64063568,
      "step": 110340
    },
    {
      "epoch": 16.435061066428357,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.002811247286550352,
      "loss": 0.8152,
      "num_input_tokens_seen": 64066224,
      "step": 110345
    },
    {
      "epoch": 16.435805778969318,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.002810111044520602,
      "loss": 0.7733,
      "num_input_tokens_seen": 64069264,
      "step": 110350
    },
    {
      "epoch": 16.43655049151028,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0028089750084266327,
      "loss": 0.7841,
      "num_input_tokens_seen": 64071856,
      "step": 110355
    },
    {
      "epoch": 16.437295204051235,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.002807839178287645,
      "loss": 0.7977,
      "num_input_tokens_seen": 64074704,
      "step": 110360
    },
    {
      "epoch": 16.438039916592196,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.002806703554122821,
      "loss": 0.7911,
      "num_input_tokens_seen": 64077712,
      "step": 110365
    },
    {
      "epoch": 16.438784629133156,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.002805568135951345,
      "loss": 0.7817,
      "num_input_tokens_seen": 64080656,
      "step": 110370
    },
    {
      "epoch": 16.439529341674113,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.002804432923792406,
      "loss": 0.797,
      "num_input_tokens_seen": 64083600,
      "step": 110375
    },
    {
      "epoch": 16.440274054215074,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.002803297917665172,
      "loss": 0.813,
      "num_input_tokens_seen": 64086672,
      "step": 110380
    },
    {
      "epoch": 16.44101876675603,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0028021631175888287,
      "loss": 0.7934,
      "num_input_tokens_seen": 64089552,
      "step": 110385
    },
    {
      "epoch": 16.44176347929699,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0028010285235825372,
      "loss": 0.7873,
      "num_input_tokens_seen": 64092304,
      "step": 110390
    },
    {
      "epoch": 16.442508191837952,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.002799894135665476,
      "loss": 0.8069,
      "num_input_tokens_seen": 64095216,
      "step": 110395
    },
    {
      "epoch": 16.44325290437891,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0027987599538568018,
      "loss": 0.798,
      "num_input_tokens_seen": 64098192,
      "step": 110400
    },
    {
      "epoch": 16.44399761691987,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0027976259781756795,
      "loss": 0.8019,
      "num_input_tokens_seen": 64101040,
      "step": 110405
    },
    {
      "epoch": 16.44474232946083,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0027964922086412616,
      "loss": 0.7876,
      "num_input_tokens_seen": 64104112,
      "step": 110410
    },
    {
      "epoch": 16.445487042001787,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0027953586452727037,
      "loss": 0.7915,
      "num_input_tokens_seen": 64107088,
      "step": 110415
    },
    {
      "epoch": 16.446231754542747,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0027942252880891593,
      "loss": 0.8029,
      "num_input_tokens_seen": 64110000,
      "step": 110420
    },
    {
      "epoch": 16.446976467083704,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0027930921371097706,
      "loss": 0.794,
      "num_input_tokens_seen": 64112752,
      "step": 110425
    },
    {
      "epoch": 16.447721179624665,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.002791959192353687,
      "loss": 0.7892,
      "num_input_tokens_seen": 64115568,
      "step": 110430
    },
    {
      "epoch": 16.448465892165625,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0027908264538400412,
      "loss": 0.8045,
      "num_input_tokens_seen": 64118320,
      "step": 110435
    },
    {
      "epoch": 16.449210604706582,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0027896939215879767,
      "loss": 0.8233,
      "num_input_tokens_seen": 64121072,
      "step": 110440
    },
    {
      "epoch": 16.449955317247543,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0027885615956166248,
      "loss": 0.7991,
      "num_input_tokens_seen": 64124016,
      "step": 110445
    },
    {
      "epoch": 16.4507000297885,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0027874294759451085,
      "loss": 0.7826,
      "num_input_tokens_seen": 64127280,
      "step": 110450
    },
    {
      "epoch": 16.45144474232946,
      "grad_norm": 0.0238037109375,
      "learning_rate": 0.0027862975625925638,
      "loss": 0.7974,
      "num_input_tokens_seen": 64130192,
      "step": 110455
    },
    {
      "epoch": 16.45218945487042,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0027851658555781075,
      "loss": 0.7829,
      "num_input_tokens_seen": 64133360,
      "step": 110460
    },
    {
      "epoch": 16.452934167411378,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0027840343549208595,
      "loss": 0.7807,
      "num_input_tokens_seen": 64136560,
      "step": 110465
    },
    {
      "epoch": 16.45367887995234,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0027829030606399325,
      "loss": 0.7989,
      "num_input_tokens_seen": 64139760,
      "step": 110470
    },
    {
      "epoch": 16.4544235924933,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0027817719727544444,
      "loss": 0.7907,
      "num_input_tokens_seen": 64142672,
      "step": 110475
    },
    {
      "epoch": 16.455168305034256,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0027806410912834967,
      "loss": 0.7988,
      "num_input_tokens_seen": 64145488,
      "step": 110480
    },
    {
      "epoch": 16.455913017575217,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0027795104162462022,
      "loss": 0.8119,
      "num_input_tokens_seen": 64148464,
      "step": 110485
    },
    {
      "epoch": 16.456657730116174,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0027783799476616597,
      "loss": 0.8087,
      "num_input_tokens_seen": 64151376,
      "step": 110490
    },
    {
      "epoch": 16.457402442657134,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.0027772496855489634,
      "loss": 0.8159,
      "num_input_tokens_seen": 64154352,
      "step": 110495
    },
    {
      "epoch": 16.458147155198095,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0027761196299272133,
      "loss": 0.7858,
      "num_input_tokens_seen": 64157232,
      "step": 110500
    },
    {
      "epoch": 16.45889186773905,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0027749897808154954,
      "loss": 0.7978,
      "num_input_tokens_seen": 64160400,
      "step": 110505
    },
    {
      "epoch": 16.459636580280012,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.002773860138232903,
      "loss": 0.8042,
      "num_input_tokens_seen": 64163088,
      "step": 110510
    },
    {
      "epoch": 16.460381292820973,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0027727307021985143,
      "loss": 0.7892,
      "num_input_tokens_seen": 64166000,
      "step": 110515
    },
    {
      "epoch": 16.46112600536193,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.002771601472731417,
      "loss": 0.7954,
      "num_input_tokens_seen": 64169040,
      "step": 110520
    },
    {
      "epoch": 16.46187071790289,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0027704724498506844,
      "loss": 0.838,
      "num_input_tokens_seen": 64171696,
      "step": 110525
    },
    {
      "epoch": 16.462615430443847,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0027693436335753897,
      "loss": 0.816,
      "num_input_tokens_seen": 64174672,
      "step": 110530
    },
    {
      "epoch": 16.463360142984808,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0027682150239246037,
      "loss": 0.7871,
      "num_input_tokens_seen": 64177488,
      "step": 110535
    },
    {
      "epoch": 16.46410485552577,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0027670866209173906,
      "loss": 0.808,
      "num_input_tokens_seen": 64180592,
      "step": 110540
    },
    {
      "epoch": 16.464849568066725,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.002765958424572819,
      "loss": 0.7959,
      "num_input_tokens_seen": 64183376,
      "step": 110545
    },
    {
      "epoch": 16.465594280607686,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.002764830434909941,
      "loss": 0.7891,
      "num_input_tokens_seen": 64186160,
      "step": 110550
    },
    {
      "epoch": 16.466338993148646,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.002763702651947823,
      "loss": 0.7801,
      "num_input_tokens_seen": 64188976,
      "step": 110555
    },
    {
      "epoch": 16.467083705689603,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0027625750757055076,
      "loss": 0.7718,
      "num_input_tokens_seen": 64192368,
      "step": 110560
    },
    {
      "epoch": 16.467828418230564,
      "grad_norm": 0.0390625,
      "learning_rate": 0.002761447706202052,
      "loss": 0.7946,
      "num_input_tokens_seen": 64194960,
      "step": 110565
    },
    {
      "epoch": 16.46857313077152,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0027603205434564987,
      "loss": 0.8086,
      "num_input_tokens_seen": 64197968,
      "step": 110570
    },
    {
      "epoch": 16.46931784331248,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.0027591935874878874,
      "loss": 0.7985,
      "num_input_tokens_seen": 64200720,
      "step": 110575
    },
    {
      "epoch": 16.470062555853442,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.002758066838315262,
      "loss": 0.8074,
      "num_input_tokens_seen": 64203568,
      "step": 110580
    },
    {
      "epoch": 16.4708072683944,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0027569402959576527,
      "loss": 0.799,
      "num_input_tokens_seen": 64206448,
      "step": 110585
    },
    {
      "epoch": 16.47155198093536,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.002755813960434097,
      "loss": 0.794,
      "num_input_tokens_seen": 64209488,
      "step": 110590
    },
    {
      "epoch": 16.472296693476316,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.002754687831763619,
      "loss": 0.8067,
      "num_input_tokens_seen": 64212432,
      "step": 110595
    },
    {
      "epoch": 16.473041406017277,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.002753561909965246,
      "loss": 0.8031,
      "num_input_tokens_seen": 64215568,
      "step": 110600
    },
    {
      "epoch": 16.473786118558237,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0027524361950579967,
      "loss": 0.7965,
      "num_input_tokens_seen": 64218512,
      "step": 110605
    },
    {
      "epoch": 16.474530831099194,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0027513106870608863,
      "loss": 0.7992,
      "num_input_tokens_seen": 64221520,
      "step": 110610
    },
    {
      "epoch": 16.475275543640155,
      "grad_norm": 0.06640625,
      "learning_rate": 0.002750185385992937,
      "loss": 0.8127,
      "num_input_tokens_seen": 64224176,
      "step": 110615
    },
    {
      "epoch": 16.476020256181116,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0027490602918731526,
      "loss": 0.8087,
      "num_input_tokens_seen": 64226896,
      "step": 110620
    },
    {
      "epoch": 16.476764968722073,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0027479354047205465,
      "loss": 0.7784,
      "num_input_tokens_seen": 64229904,
      "step": 110625
    },
    {
      "epoch": 16.477509681263033,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.002746810724554114,
      "loss": 0.7731,
      "num_input_tokens_seen": 64232816,
      "step": 110630
    },
    {
      "epoch": 16.47825439380399,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.002745686251392866,
      "loss": 0.8007,
      "num_input_tokens_seen": 64236208,
      "step": 110635
    },
    {
      "epoch": 16.47899910634495,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.0027445619852557905,
      "loss": 0.8,
      "num_input_tokens_seen": 64239344,
      "step": 110640
    },
    {
      "epoch": 16.47974381888591,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0027434379261618877,
      "loss": 0.7919,
      "num_input_tokens_seen": 64242096,
      "step": 110645
    },
    {
      "epoch": 16.480488531426868,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0027423140741301443,
      "loss": 0.7932,
      "num_input_tokens_seen": 64244880,
      "step": 110650
    },
    {
      "epoch": 16.48123324396783,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0027411904291795414,
      "loss": 0.7819,
      "num_input_tokens_seen": 64247536,
      "step": 110655
    },
    {
      "epoch": 16.48197795650879,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.002740066991329072,
      "loss": 0.8019,
      "num_input_tokens_seen": 64250736,
      "step": 110660
    },
    {
      "epoch": 16.482722669049746,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0027389437605977096,
      "loss": 0.8037,
      "num_input_tokens_seen": 64253648,
      "step": 110665
    },
    {
      "epoch": 16.483467381590707,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.002737820737004432,
      "loss": 0.7746,
      "num_input_tokens_seen": 64256368,
      "step": 110670
    },
    {
      "epoch": 16.484212094131664,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0027366979205682043,
      "loss": 0.7995,
      "num_input_tokens_seen": 64259280,
      "step": 110675
    },
    {
      "epoch": 16.484956806672624,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.002735575311308008,
      "loss": 0.7961,
      "num_input_tokens_seen": 64262160,
      "step": 110680
    },
    {
      "epoch": 16.485701519213585,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0027344529092428,
      "loss": 0.7906,
      "num_input_tokens_seen": 64265040,
      "step": 110685
    },
    {
      "epoch": 16.48644623175454,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0027333307143915407,
      "loss": 0.8007,
      "num_input_tokens_seen": 64267952,
      "step": 110690
    },
    {
      "epoch": 16.487190944295502,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0027322087267731957,
      "loss": 0.7835,
      "num_input_tokens_seen": 64271184,
      "step": 110695
    },
    {
      "epoch": 16.487935656836463,
      "grad_norm": 0.171875,
      "learning_rate": 0.0027310869464067114,
      "loss": 0.8009,
      "num_input_tokens_seen": 64273904,
      "step": 110700
    },
    {
      "epoch": 16.48868036937742,
      "grad_norm": 0.30859375,
      "learning_rate": 0.002729965373311048,
      "loss": 0.8144,
      "num_input_tokens_seen": 64276880,
      "step": 110705
    },
    {
      "epoch": 16.48942508191838,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0027288440075051443,
      "loss": 0.7879,
      "num_input_tokens_seen": 64279984,
      "step": 110710
    },
    {
      "epoch": 16.490169794459337,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.002727722849007953,
      "loss": 0.7796,
      "num_input_tokens_seen": 64283056,
      "step": 110715
    },
    {
      "epoch": 16.490914507000298,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.002726601897838408,
      "loss": 0.7917,
      "num_input_tokens_seen": 64285680,
      "step": 110720
    },
    {
      "epoch": 16.49165921954126,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0027254811540154533,
      "loss": 0.8083,
      "num_input_tokens_seen": 64288272,
      "step": 110725
    },
    {
      "epoch": 16.492403932082215,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.00272436061755802,
      "loss": 0.8047,
      "num_input_tokens_seen": 64290928,
      "step": 110730
    },
    {
      "epoch": 16.493148644623176,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0027232402884850375,
      "loss": 0.7992,
      "num_input_tokens_seen": 64293872,
      "step": 110735
    },
    {
      "epoch": 16.493893357164133,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.0027221201668154314,
      "loss": 0.7898,
      "num_input_tokens_seen": 64296944,
      "step": 110740
    },
    {
      "epoch": 16.494638069705093,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0027210002525681247,
      "loss": 0.798,
      "num_input_tokens_seen": 64299856,
      "step": 110745
    },
    {
      "epoch": 16.495382782246054,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0027198805457620406,
      "loss": 0.8002,
      "num_input_tokens_seen": 64302768,
      "step": 110750
    },
    {
      "epoch": 16.49612749478701,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.002718761046416092,
      "loss": 0.8183,
      "num_input_tokens_seen": 64305936,
      "step": 110755
    },
    {
      "epoch": 16.49687220732797,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.002717641754549196,
      "loss": 0.7981,
      "num_input_tokens_seen": 64308880,
      "step": 110760
    },
    {
      "epoch": 16.497616919868932,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0027165226701802584,
      "loss": 0.7807,
      "num_input_tokens_seen": 64311952,
      "step": 110765
    },
    {
      "epoch": 16.49836163240989,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0027154037933281847,
      "loss": 0.7744,
      "num_input_tokens_seen": 64314672,
      "step": 110770
    },
    {
      "epoch": 16.49910634495085,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0027142851240118796,
      "loss": 0.8289,
      "num_input_tokens_seen": 64317552,
      "step": 110775
    },
    {
      "epoch": 16.499851057491806,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.002713166662250238,
      "loss": 0.8034,
      "num_input_tokens_seen": 64320464,
      "step": 110780
    },
    {
      "epoch": 16.500595770032767,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.002712048408062161,
      "loss": 0.7822,
      "num_input_tokens_seen": 64323472,
      "step": 110785
    },
    {
      "epoch": 16.501340482573728,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0027109303614665397,
      "loss": 0.808,
      "num_input_tokens_seen": 64326224,
      "step": 110790
    },
    {
      "epoch": 16.502085195114685,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.002709812522482258,
      "loss": 0.7863,
      "num_input_tokens_seen": 64329104,
      "step": 110795
    },
    {
      "epoch": 16.502829907655645,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0027086948911281988,
      "loss": 0.8171,
      "num_input_tokens_seen": 64332176,
      "step": 110800
    },
    {
      "epoch": 16.503574620196606,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0027075774674232512,
      "loss": 0.8545,
      "num_input_tokens_seen": 64335184,
      "step": 110805
    },
    {
      "epoch": 16.504319332737563,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0027064602513862876,
      "loss": 0.8064,
      "num_input_tokens_seen": 64338032,
      "step": 110810
    },
    {
      "epoch": 16.505064045278523,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0027053432430361804,
      "loss": 0.7831,
      "num_input_tokens_seen": 64340784,
      "step": 110815
    },
    {
      "epoch": 16.50580875781948,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0027042264423918078,
      "loss": 0.7943,
      "num_input_tokens_seen": 64343664,
      "step": 110820
    },
    {
      "epoch": 16.50655347036044,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.0027031098494720287,
      "loss": 0.7973,
      "num_input_tokens_seen": 64346672,
      "step": 110825
    },
    {
      "epoch": 16.5072981829014,
      "grad_norm": 0.078125,
      "learning_rate": 0.0027019934642957145,
      "loss": 0.7954,
      "num_input_tokens_seen": 64349616,
      "step": 110830
    },
    {
      "epoch": 16.508042895442358,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.002700877286881717,
      "loss": 0.7778,
      "num_input_tokens_seen": 64352816,
      "step": 110835
    },
    {
      "epoch": 16.50878760798332,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.002699761317248902,
      "loss": 0.799,
      "num_input_tokens_seen": 64355856,
      "step": 110840
    },
    {
      "epoch": 16.50953232052428,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0026986455554161176,
      "loss": 0.8527,
      "num_input_tokens_seen": 64359120,
      "step": 110845
    },
    {
      "epoch": 16.510277033065236,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.00269753000140221,
      "loss": 0.8119,
      "num_input_tokens_seen": 64361616,
      "step": 110850
    },
    {
      "epoch": 16.511021745606197,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.002696414655226033,
      "loss": 0.8127,
      "num_input_tokens_seen": 64364624,
      "step": 110855
    },
    {
      "epoch": 16.511766458147154,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0026952995169064265,
      "loss": 0.8157,
      "num_input_tokens_seen": 64367472,
      "step": 110860
    },
    {
      "epoch": 16.512511170688114,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.00269418458646223,
      "loss": 0.8028,
      "num_input_tokens_seen": 64370224,
      "step": 110865
    },
    {
      "epoch": 16.513255883229075,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0026930698639122718,
      "loss": 0.7843,
      "num_input_tokens_seen": 64373008,
      "step": 110870
    },
    {
      "epoch": 16.51400059577003,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.002691955349275394,
      "loss": 0.769,
      "num_input_tokens_seen": 64375664,
      "step": 110875
    },
    {
      "epoch": 16.514745308310992,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0026908410425704175,
      "loss": 0.7911,
      "num_input_tokens_seen": 64378352,
      "step": 110880
    },
    {
      "epoch": 16.515490020851953,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.002689726943816176,
      "loss": 0.8035,
      "num_input_tokens_seen": 64381200,
      "step": 110885
    },
    {
      "epoch": 16.51623473339291,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.002688613053031485,
      "loss": 0.7754,
      "num_input_tokens_seen": 64384304,
      "step": 110890
    },
    {
      "epoch": 16.51697944593387,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.002687499370235162,
      "loss": 0.7814,
      "num_input_tokens_seen": 64387024,
      "step": 110895
    },
    {
      "epoch": 16.517724158474827,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.002686385895446025,
      "loss": 0.7863,
      "num_input_tokens_seen": 64389904,
      "step": 110900
    },
    {
      "epoch": 16.518468871015788,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00268527262868288,
      "loss": 0.7908,
      "num_input_tokens_seen": 64392880,
      "step": 110905
    },
    {
      "epoch": 16.51921358355675,
      "grad_norm": 0.060546875,
      "learning_rate": 0.002684159569964541,
      "loss": 0.81,
      "num_input_tokens_seen": 64395632,
      "step": 110910
    },
    {
      "epoch": 16.519958296097705,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0026830467193098067,
      "loss": 0.8073,
      "num_input_tokens_seen": 64398608,
      "step": 110915
    },
    {
      "epoch": 16.520703008638666,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.0026819340767374822,
      "loss": 0.7877,
      "num_input_tokens_seen": 64401264,
      "step": 110920
    },
    {
      "epoch": 16.521447721179626,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.002680821642266362,
      "loss": 0.7872,
      "num_input_tokens_seen": 64404304,
      "step": 110925
    },
    {
      "epoch": 16.522192433720583,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.00267970941591524,
      "loss": 0.8071,
      "num_input_tokens_seen": 64407248,
      "step": 110930
    },
    {
      "epoch": 16.522937146261544,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.002678597397702904,
      "loss": 0.7996,
      "num_input_tokens_seen": 64410256,
      "step": 110935
    },
    {
      "epoch": 16.5236818588025,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0026774855876481414,
      "loss": 0.7935,
      "num_input_tokens_seen": 64413200,
      "step": 110940
    },
    {
      "epoch": 16.52442657134346,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0026763739857697377,
      "loss": 0.8123,
      "num_input_tokens_seen": 64415984,
      "step": 110945
    },
    {
      "epoch": 16.525171283884422,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0026752625920864652,
      "loss": 0.8018,
      "num_input_tokens_seen": 64418800,
      "step": 110950
    },
    {
      "epoch": 16.52591599642538,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0026741514066171107,
      "loss": 0.8208,
      "num_input_tokens_seen": 64421552,
      "step": 110955
    },
    {
      "epoch": 16.52666070896634,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0026730404293804372,
      "loss": 0.7803,
      "num_input_tokens_seen": 64424400,
      "step": 110960
    },
    {
      "epoch": 16.527405421507297,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00267192966039522,
      "loss": 0.7806,
      "num_input_tokens_seen": 64427216,
      "step": 110965
    },
    {
      "epoch": 16.528150134048257,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.00267081909968022,
      "loss": 0.7984,
      "num_input_tokens_seen": 64430032,
      "step": 110970
    },
    {
      "epoch": 16.528894846589218,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0026697087472541997,
      "loss": 0.7954,
      "num_input_tokens_seen": 64432976,
      "step": 110975
    },
    {
      "epoch": 16.529639559130175,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0026685986031359204,
      "loss": 0.7821,
      "num_input_tokens_seen": 64435664,
      "step": 110980
    },
    {
      "epoch": 16.530384271671135,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0026674886673441305,
      "loss": 0.8071,
      "num_input_tokens_seen": 64438352,
      "step": 110985
    },
    {
      "epoch": 16.531128984212096,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.00266637893989759,
      "loss": 0.7885,
      "num_input_tokens_seen": 64441360,
      "step": 110990
    },
    {
      "epoch": 16.531873696753053,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0026652694208150404,
      "loss": 0.7995,
      "num_input_tokens_seen": 64444528,
      "step": 110995
    },
    {
      "epoch": 16.532618409294013,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.002664160110115229,
      "loss": 0.7829,
      "num_input_tokens_seen": 64447440,
      "step": 111000
    },
    {
      "epoch": 16.53336312183497,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0026630510078168917,
      "loss": 0.7777,
      "num_input_tokens_seen": 64450352,
      "step": 111005
    },
    {
      "epoch": 16.53410783437593,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.0026619421139387704,
      "loss": 0.8003,
      "num_input_tokens_seen": 64453072,
      "step": 111010
    },
    {
      "epoch": 16.53485254691689,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0026608334284995987,
      "loss": 0.7947,
      "num_input_tokens_seen": 64456240,
      "step": 111015
    },
    {
      "epoch": 16.535597259457848,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0026597249515181013,
      "loss": 0.8248,
      "num_input_tokens_seen": 64459088,
      "step": 111020
    },
    {
      "epoch": 16.53634197199881,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0026586166830130137,
      "loss": 0.7724,
      "num_input_tokens_seen": 64461776,
      "step": 111025
    },
    {
      "epoch": 16.53708668453977,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0026575086230030487,
      "loss": 0.8075,
      "num_input_tokens_seen": 64464624,
      "step": 111030
    },
    {
      "epoch": 16.537831397080726,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.002656400771506935,
      "loss": 0.7975,
      "num_input_tokens_seen": 64467280,
      "step": 111035
    },
    {
      "epoch": 16.538576109621687,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.002655293128543382,
      "loss": 0.7903,
      "num_input_tokens_seen": 64470032,
      "step": 111040
    },
    {
      "epoch": 16.539320822162644,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0026541856941311075,
      "loss": 0.7937,
      "num_input_tokens_seen": 64473104,
      "step": 111045
    },
    {
      "epoch": 16.540065534703604,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0026530784682888193,
      "loss": 0.7905,
      "num_input_tokens_seen": 64475824,
      "step": 111050
    },
    {
      "epoch": 16.540810247244565,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.002651971451035217,
      "loss": 0.7876,
      "num_input_tokens_seen": 64478768,
      "step": 111055
    },
    {
      "epoch": 16.541554959785522,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0026508646423890115,
      "loss": 0.7997,
      "num_input_tokens_seen": 64481616,
      "step": 111060
    },
    {
      "epoch": 16.542299672326482,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.002649758042368895,
      "loss": 0.8173,
      "num_input_tokens_seen": 64484528,
      "step": 111065
    },
    {
      "epoch": 16.543044384867443,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0026486516509935656,
      "loss": 0.7939,
      "num_input_tokens_seen": 64487664,
      "step": 111070
    },
    {
      "epoch": 16.5437890974084,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0026475454682817093,
      "loss": 0.7919,
      "num_input_tokens_seen": 64490544,
      "step": 111075
    },
    {
      "epoch": 16.54453380994936,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0026464394942520213,
      "loss": 0.8127,
      "num_input_tokens_seen": 64493872,
      "step": 111080
    },
    {
      "epoch": 16.545278522490317,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0026453337289231798,
      "loss": 0.7889,
      "num_input_tokens_seen": 64496880,
      "step": 111085
    },
    {
      "epoch": 16.546023235031278,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.00264422817231387,
      "loss": 0.7949,
      "num_input_tokens_seen": 64499536,
      "step": 111090
    },
    {
      "epoch": 16.54676794757224,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0026431228244427683,
      "loss": 0.7787,
      "num_input_tokens_seen": 64502416,
      "step": 111095
    },
    {
      "epoch": 16.547512660113195,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.002642017685328545,
      "loss": 0.798,
      "num_input_tokens_seen": 64505264,
      "step": 111100
    },
    {
      "epoch": 16.548257372654156,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0026409127549898758,
      "loss": 0.8035,
      "num_input_tokens_seen": 64508272,
      "step": 111105
    },
    {
      "epoch": 16.549002085195113,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0026398080334454196,
      "loss": 0.7987,
      "num_input_tokens_seen": 64511120,
      "step": 111110
    },
    {
      "epoch": 16.549746797736073,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0026387035207138493,
      "loss": 0.7907,
      "num_input_tokens_seen": 64514608,
      "step": 111115
    },
    {
      "epoch": 16.550491510277034,
      "grad_norm": 0.0625,
      "learning_rate": 0.002637599216813819,
      "loss": 0.8157,
      "num_input_tokens_seen": 64517424,
      "step": 111120
    },
    {
      "epoch": 16.55123622281799,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.002636495121763986,
      "loss": 0.8003,
      "num_input_tokens_seen": 64520304,
      "step": 111125
    },
    {
      "epoch": 16.55198093535895,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0026353912355829966,
      "loss": 0.7871,
      "num_input_tokens_seen": 64522832,
      "step": 111130
    },
    {
      "epoch": 16.552725647899912,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0026342875582895113,
      "loss": 0.8112,
      "num_input_tokens_seen": 64525744,
      "step": 111135
    },
    {
      "epoch": 16.55347036044087,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0026331840899021684,
      "loss": 0.7934,
      "num_input_tokens_seen": 64528688,
      "step": 111140
    },
    {
      "epoch": 16.55421507298183,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.002632080830439608,
      "loss": 0.823,
      "num_input_tokens_seen": 64531632,
      "step": 111145
    },
    {
      "epoch": 16.554959785522787,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.002630977779920475,
      "loss": 0.8059,
      "num_input_tokens_seen": 64534384,
      "step": 111150
    },
    {
      "epoch": 16.555704498063747,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.002629874938363398,
      "loss": 0.8349,
      "num_input_tokens_seen": 64537008,
      "step": 111155
    },
    {
      "epoch": 16.556449210604708,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.002628772305787016,
      "loss": 0.7932,
      "num_input_tokens_seen": 64539440,
      "step": 111160
    },
    {
      "epoch": 16.557193923145665,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.002627669882209948,
      "loss": 0.786,
      "num_input_tokens_seen": 64542224,
      "step": 111165
    },
    {
      "epoch": 16.557938635686625,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0026265676676508246,
      "loss": 0.8126,
      "num_input_tokens_seen": 64545168,
      "step": 111170
    },
    {
      "epoch": 16.558683348227586,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.002625465662128268,
      "loss": 0.7975,
      "num_input_tokens_seen": 64548272,
      "step": 111175
    },
    {
      "epoch": 16.559428060768543,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0026243638656608856,
      "loss": 0.7874,
      "num_input_tokens_seen": 64550864,
      "step": 111180
    },
    {
      "epoch": 16.560172773309503,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.002623262278267302,
      "loss": 0.7912,
      "num_input_tokens_seen": 64553904,
      "step": 111185
    },
    {
      "epoch": 16.56091748585046,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0026221608999661234,
      "loss": 0.7904,
      "num_input_tokens_seen": 64557072,
      "step": 111190
    },
    {
      "epoch": 16.56166219839142,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0026210597307759565,
      "loss": 0.8017,
      "num_input_tokens_seen": 64560176,
      "step": 111195
    },
    {
      "epoch": 16.56240691093238,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0026199587707154,
      "loss": 0.8044,
      "num_input_tokens_seen": 64562736,
      "step": 111200
    },
    {
      "epoch": 16.56315162347334,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.002618858019803062,
      "loss": 0.8091,
      "num_input_tokens_seen": 64565328,
      "step": 111205
    },
    {
      "epoch": 16.5638963360143,
      "grad_norm": 0.05078125,
      "learning_rate": 0.002617757478057533,
      "loss": 0.8012,
      "num_input_tokens_seen": 64568176,
      "step": 111210
    },
    {
      "epoch": 16.56464104855526,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.002616657145497403,
      "loss": 0.7868,
      "num_input_tokens_seen": 64571056,
      "step": 111215
    },
    {
      "epoch": 16.565385761096216,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0026155570221412685,
      "loss": 0.7919,
      "num_input_tokens_seen": 64573904,
      "step": 111220
    },
    {
      "epoch": 16.566130473637177,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.002614457108007707,
      "loss": 0.8202,
      "num_input_tokens_seen": 64576656,
      "step": 111225
    },
    {
      "epoch": 16.566875186178134,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.00261335740311531,
      "loss": 0.799,
      "num_input_tokens_seen": 64579376,
      "step": 111230
    },
    {
      "epoch": 16.567619898719094,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0026122579074826457,
      "loss": 0.7997,
      "num_input_tokens_seen": 64582448,
      "step": 111235
    },
    {
      "epoch": 16.568364611260055,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.002611158621128296,
      "loss": 0.7913,
      "num_input_tokens_seen": 64585424,
      "step": 111240
    },
    {
      "epoch": 16.569109323801012,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.002610059544070828,
      "loss": 0.8286,
      "num_input_tokens_seen": 64588240,
      "step": 111245
    },
    {
      "epoch": 16.569854036341972,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0026089606763288146,
      "loss": 0.81,
      "num_input_tokens_seen": 64591024,
      "step": 111250
    },
    {
      "epoch": 16.57059874888293,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0026078620179208183,
      "loss": 0.7902,
      "num_input_tokens_seen": 64593872,
      "step": 111255
    },
    {
      "epoch": 16.57134346142389,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0026067635688653956,
      "loss": 0.7953,
      "num_input_tokens_seen": 64596528,
      "step": 111260
    },
    {
      "epoch": 16.57208817396485,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0026056653291811082,
      "loss": 0.8009,
      "num_input_tokens_seen": 64599216,
      "step": 111265
    },
    {
      "epoch": 16.572832886505807,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.002604567298886505,
      "loss": 0.8202,
      "num_input_tokens_seen": 64601808,
      "step": 111270
    },
    {
      "epoch": 16.573577599046768,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0026034694780001423,
      "loss": 0.7984,
      "num_input_tokens_seen": 64604560,
      "step": 111275
    },
    {
      "epoch": 16.57432231158773,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0026023718665405585,
      "loss": 0.7738,
      "num_input_tokens_seen": 64607536,
      "step": 111280
    },
    {
      "epoch": 16.575067024128685,
      "grad_norm": 0.07421875,
      "learning_rate": 0.002601274464526306,
      "loss": 0.7917,
      "num_input_tokens_seen": 64610224,
      "step": 111285
    },
    {
      "epoch": 16.575811736669646,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00260017727197592,
      "loss": 0.7945,
      "num_input_tokens_seen": 64613168,
      "step": 111290
    },
    {
      "epoch": 16.576556449210603,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0025990802889079337,
      "loss": 0.8234,
      "num_input_tokens_seen": 64616176,
      "step": 111295
    },
    {
      "epoch": 16.577301161751564,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.002597983515340884,
      "loss": 0.7982,
      "num_input_tokens_seen": 64619056,
      "step": 111300
    },
    {
      "epoch": 16.578045874292524,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0025968869512932964,
      "loss": 0.8115,
      "num_input_tokens_seen": 64622064,
      "step": 111305
    },
    {
      "epoch": 16.57879058683348,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.0025957905967837003,
      "loss": 0.8104,
      "num_input_tokens_seen": 64625232,
      "step": 111310
    },
    {
      "epoch": 16.57953529937444,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0025946944518306114,
      "loss": 0.7873,
      "num_input_tokens_seen": 64628048,
      "step": 111315
    },
    {
      "epoch": 16.580280011915402,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0025935985164525547,
      "loss": 0.8116,
      "num_input_tokens_seen": 64630896,
      "step": 111320
    },
    {
      "epoch": 16.58102472445636,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0025925027906680423,
      "loss": 0.7993,
      "num_input_tokens_seen": 64633456,
      "step": 111325
    },
    {
      "epoch": 16.58176943699732,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0025914072744955847,
      "loss": 0.7883,
      "num_input_tokens_seen": 64636432,
      "step": 111330
    },
    {
      "epoch": 16.582514149538277,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0025903119679536895,
      "loss": 0.8027,
      "num_input_tokens_seen": 64639280,
      "step": 111335
    },
    {
      "epoch": 16.583258862079237,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.0025892168710608576,
      "loss": 0.7916,
      "num_input_tokens_seen": 64642416,
      "step": 111340
    },
    {
      "epoch": 16.584003574620198,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0025881219838355973,
      "loss": 0.7934,
      "num_input_tokens_seen": 64646512,
      "step": 111345
    },
    {
      "epoch": 16.584748287161155,
      "grad_norm": 0.0625,
      "learning_rate": 0.002587027306296397,
      "loss": 0.7926,
      "num_input_tokens_seen": 64649232,
      "step": 111350
    },
    {
      "epoch": 16.585492999702115,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0025859328384617575,
      "loss": 0.7786,
      "num_input_tokens_seen": 64652208,
      "step": 111355
    },
    {
      "epoch": 16.586237712243076,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0025848385803501615,
      "loss": 0.8246,
      "num_input_tokens_seen": 64654896,
      "step": 111360
    },
    {
      "epoch": 16.586982424784033,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0025837445319801047,
      "loss": 0.7996,
      "num_input_tokens_seen": 64657776,
      "step": 111365
    },
    {
      "epoch": 16.587727137324993,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.002582650693370064,
      "loss": 0.784,
      "num_input_tokens_seen": 64660784,
      "step": 111370
    },
    {
      "epoch": 16.58847184986595,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0025815570645385147,
      "loss": 0.7804,
      "num_input_tokens_seen": 64663760,
      "step": 111375
    },
    {
      "epoch": 16.58921656240691,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.002580463645503942,
      "loss": 0.8005,
      "num_input_tokens_seen": 64666544,
      "step": 111380
    },
    {
      "epoch": 16.58996127494787,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0025793704362848105,
      "loss": 0.7773,
      "num_input_tokens_seen": 64669552,
      "step": 111385
    },
    {
      "epoch": 16.59070598748883,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0025782774368995945,
      "loss": 0.8039,
      "num_input_tokens_seen": 64672528,
      "step": 111390
    },
    {
      "epoch": 16.59145070002979,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.002577184647366756,
      "loss": 0.8117,
      "num_input_tokens_seen": 64675440,
      "step": 111395
    },
    {
      "epoch": 16.59219541257075,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0025760920677047575,
      "loss": 0.7954,
      "num_input_tokens_seen": 64678448,
      "step": 111400
    },
    {
      "epoch": 16.592940125111706,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.002574999697932052,
      "loss": 0.8007,
      "num_input_tokens_seen": 64681104,
      "step": 111405
    },
    {
      "epoch": 16.593684837652667,
      "grad_norm": 0.0703125,
      "learning_rate": 0.002573907538067103,
      "loss": 0.8172,
      "num_input_tokens_seen": 64683920,
      "step": 111410
    },
    {
      "epoch": 16.594429550193624,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0025728155881283563,
      "loss": 0.8014,
      "num_input_tokens_seen": 64686576,
      "step": 111415
    },
    {
      "epoch": 16.595174262734584,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0025717238481342553,
      "loss": 0.7914,
      "num_input_tokens_seen": 64689424,
      "step": 111420
    },
    {
      "epoch": 16.595918975275545,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.0025706323181032523,
      "loss": 0.785,
      "num_input_tokens_seen": 64692432,
      "step": 111425
    },
    {
      "epoch": 16.596663687816502,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.002569540998053782,
      "loss": 0.7873,
      "num_input_tokens_seen": 64695280,
      "step": 111430
    },
    {
      "epoch": 16.597408400357462,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.002568449888004284,
      "loss": 0.7955,
      "num_input_tokens_seen": 64698128,
      "step": 111435
    },
    {
      "epoch": 16.598153112898423,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.002567358987973188,
      "loss": 0.7935,
      "num_input_tokens_seen": 64700944,
      "step": 111440
    },
    {
      "epoch": 16.59889782543938,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0025662682979789293,
      "loss": 0.7909,
      "num_input_tokens_seen": 64703920,
      "step": 111445
    },
    {
      "epoch": 16.59964253798034,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0025651778180399297,
      "loss": 0.7796,
      "num_input_tokens_seen": 64706704,
      "step": 111450
    },
    {
      "epoch": 16.600387250521297,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.002564087548174614,
      "loss": 0.8074,
      "num_input_tokens_seen": 64709584,
      "step": 111455
    },
    {
      "epoch": 16.601131963062258,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0025629974884013966,
      "loss": 0.7708,
      "num_input_tokens_seen": 64712336,
      "step": 111460
    },
    {
      "epoch": 16.60187667560322,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.0025619076387386986,
      "loss": 0.7988,
      "num_input_tokens_seen": 64715120,
      "step": 111465
    },
    {
      "epoch": 16.602621388144176,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0025608179992049297,
      "loss": 0.8031,
      "num_input_tokens_seen": 64717936,
      "step": 111470
    },
    {
      "epoch": 16.603366100685136,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.002559728569818495,
      "loss": 0.7968,
      "num_input_tokens_seen": 64721040,
      "step": 111475
    },
    {
      "epoch": 16.604110813226093,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.0025586393505978053,
      "loss": 0.7845,
      "num_input_tokens_seen": 64723856,
      "step": 111480
    },
    {
      "epoch": 16.604855525767054,
      "grad_norm": 0.06640625,
      "learning_rate": 0.002557550341561256,
      "loss": 0.7794,
      "num_input_tokens_seen": 64726608,
      "step": 111485
    },
    {
      "epoch": 16.605600238308014,
      "grad_norm": 0.0546875,
      "learning_rate": 0.002556461542727252,
      "loss": 0.8086,
      "num_input_tokens_seen": 64729616,
      "step": 111490
    },
    {
      "epoch": 16.60634495084897,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0025553729541141815,
      "loss": 0.7795,
      "num_input_tokens_seen": 64732272,
      "step": 111495
    },
    {
      "epoch": 16.60708966338993,
      "grad_norm": 0.083984375,
      "learning_rate": 0.002554284575740434,
      "loss": 0.801,
      "num_input_tokens_seen": 64735056,
      "step": 111500
    },
    {
      "epoch": 16.607834375930892,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0025531964076244027,
      "loss": 0.7787,
      "num_input_tokens_seen": 64737776,
      "step": 111505
    },
    {
      "epoch": 16.60857908847185,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.002552108449784465,
      "loss": 0.7863,
      "num_input_tokens_seen": 64740944,
      "step": 111510
    },
    {
      "epoch": 16.60932380101281,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.002551020702239005,
      "loss": 0.8019,
      "num_input_tokens_seen": 64744176,
      "step": 111515
    },
    {
      "epoch": 16.610068513553767,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0025499331650063995,
      "loss": 0.8009,
      "num_input_tokens_seen": 64746960,
      "step": 111520
    },
    {
      "epoch": 16.610813226094727,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0025488458381050195,
      "loss": 0.8187,
      "num_input_tokens_seen": 64749968,
      "step": 111525
    },
    {
      "epoch": 16.611557938635688,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0025477587215532306,
      "loss": 0.8018,
      "num_input_tokens_seen": 64753168,
      "step": 111530
    },
    {
      "epoch": 16.612302651176645,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0025466718153694043,
      "loss": 0.8005,
      "num_input_tokens_seen": 64756240,
      "step": 111535
    },
    {
      "epoch": 16.613047363717605,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0025455851195719035,
      "loss": 0.7644,
      "num_input_tokens_seen": 64759056,
      "step": 111540
    },
    {
      "epoch": 16.613792076258566,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0025444986341790782,
      "loss": 0.7907,
      "num_input_tokens_seen": 64761904,
      "step": 111545
    },
    {
      "epoch": 16.614536788799523,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.002543412359209293,
      "loss": 0.7803,
      "num_input_tokens_seen": 64764784,
      "step": 111550
    },
    {
      "epoch": 16.615281501340483,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0025423262946808927,
      "loss": 0.7893,
      "num_input_tokens_seen": 64767792,
      "step": 111555
    },
    {
      "epoch": 16.61602621388144,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0025412404406122318,
      "loss": 0.8068,
      "num_input_tokens_seen": 64770992,
      "step": 111560
    },
    {
      "epoch": 16.6167709264224,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0025401547970216477,
      "loss": 0.7756,
      "num_input_tokens_seen": 64773840,
      "step": 111565
    },
    {
      "epoch": 16.61751563896336,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0025390693639274885,
      "loss": 0.7893,
      "num_input_tokens_seen": 64776976,
      "step": 111570
    },
    {
      "epoch": 16.61826035150432,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.002537984141348087,
      "loss": 0.7844,
      "num_input_tokens_seen": 64779760,
      "step": 111575
    },
    {
      "epoch": 16.61900506404528,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0025368991293017746,
      "loss": 0.8328,
      "num_input_tokens_seen": 64782832,
      "step": 111580
    },
    {
      "epoch": 16.61974977658624,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0025358143278068866,
      "loss": 0.7766,
      "num_input_tokens_seen": 64785840,
      "step": 111585
    },
    {
      "epoch": 16.620494489127196,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.002534729736881749,
      "loss": 0.7963,
      "num_input_tokens_seen": 64788752,
      "step": 111590
    },
    {
      "epoch": 16.621239201668157,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0025336453565446833,
      "loss": 0.831,
      "num_input_tokens_seen": 64791664,
      "step": 111595
    },
    {
      "epoch": 16.621983914209114,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0025325611868140044,
      "loss": 0.7895,
      "num_input_tokens_seen": 64794416,
      "step": 111600
    },
    {
      "epoch": 16.622728626750074,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.002531477227708037,
      "loss": 0.7922,
      "num_input_tokens_seen": 64797616,
      "step": 111605
    },
    {
      "epoch": 16.623473339291035,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.002530393479245086,
      "loss": 0.8312,
      "num_input_tokens_seen": 64800944,
      "step": 111610
    },
    {
      "epoch": 16.624218051831992,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.0025293099414434683,
      "loss": 0.7912,
      "num_input_tokens_seen": 64803792,
      "step": 111615
    },
    {
      "epoch": 16.624962764372953,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.0025282266143214827,
      "loss": 0.7825,
      "num_input_tokens_seen": 64806480,
      "step": 111620
    },
    {
      "epoch": 16.62570747691391,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0025271434978974294,
      "loss": 0.7952,
      "num_input_tokens_seen": 64809392,
      "step": 111625
    },
    {
      "epoch": 16.62645218945487,
      "grad_norm": 0.1015625,
      "learning_rate": 0.002526060592189614,
      "loss": 0.7902,
      "num_input_tokens_seen": 64813744,
      "step": 111630
    },
    {
      "epoch": 16.62719690199583,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.0025249778972163232,
      "loss": 0.8071,
      "num_input_tokens_seen": 64816304,
      "step": 111635
    },
    {
      "epoch": 16.627941614536788,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0025238954129958562,
      "loss": 0.7942,
      "num_input_tokens_seen": 64818928,
      "step": 111640
    },
    {
      "epoch": 16.628686327077748,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.002522813139546491,
      "loss": 0.7871,
      "num_input_tokens_seen": 64821808,
      "step": 111645
    },
    {
      "epoch": 16.62943103961871,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0025217310768865207,
      "loss": 0.7913,
      "num_input_tokens_seen": 64824688,
      "step": 111650
    },
    {
      "epoch": 16.630175752159666,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0025206492250342215,
      "loss": 0.8246,
      "num_input_tokens_seen": 64828016,
      "step": 111655
    },
    {
      "epoch": 16.630920464700626,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0025195675840078706,
      "loss": 0.7979,
      "num_input_tokens_seen": 64830768,
      "step": 111660
    },
    {
      "epoch": 16.631665177241583,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0025184861538257402,
      "loss": 0.7952,
      "num_input_tokens_seen": 64833968,
      "step": 111665
    },
    {
      "epoch": 16.632409889782544,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.002517404934506097,
      "loss": 0.7833,
      "num_input_tokens_seen": 64836624,
      "step": 111670
    },
    {
      "epoch": 16.633154602323504,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0025163239260672155,
      "loss": 0.8013,
      "num_input_tokens_seen": 64839312,
      "step": 111675
    },
    {
      "epoch": 16.63389931486446,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0025152431285273484,
      "loss": 0.8074,
      "num_input_tokens_seen": 64842192,
      "step": 111680
    },
    {
      "epoch": 16.63464402740542,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0025141625419047653,
      "loss": 0.797,
      "num_input_tokens_seen": 64845104,
      "step": 111685
    },
    {
      "epoch": 16.635388739946382,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.002513082166217709,
      "loss": 0.7843,
      "num_input_tokens_seen": 64848304,
      "step": 111690
    },
    {
      "epoch": 16.63613345248734,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0025120020014844447,
      "loss": 0.7898,
      "num_input_tokens_seen": 64851344,
      "step": 111695
    },
    {
      "epoch": 16.6368781650283,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0025109220477232148,
      "loss": 0.7845,
      "num_input_tokens_seen": 64854224,
      "step": 111700
    },
    {
      "epoch": 16.637622877569257,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0025098423049522583,
      "loss": 0.7958,
      "num_input_tokens_seen": 64857072,
      "step": 111705
    },
    {
      "epoch": 16.638367590110217,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.002508762773189826,
      "loss": 0.7983,
      "num_input_tokens_seen": 64860112,
      "step": 111710
    },
    {
      "epoch": 16.639112302651178,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0025076834524541463,
      "loss": 0.7814,
      "num_input_tokens_seen": 64863056,
      "step": 111715
    },
    {
      "epoch": 16.639857015192135,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.002506604342763463,
      "loss": 0.8015,
      "num_input_tokens_seen": 64865936,
      "step": 111720
    },
    {
      "epoch": 16.640601727733095,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0025055254441360003,
      "loss": 0.8026,
      "num_input_tokens_seen": 64868464,
      "step": 111725
    },
    {
      "epoch": 16.641346440274056,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.002504446756589988,
      "loss": 0.7906,
      "num_input_tokens_seen": 64871472,
      "step": 111730
    },
    {
      "epoch": 16.642091152815013,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0025033682801436467,
      "loss": 0.761,
      "num_input_tokens_seen": 64874160,
      "step": 111735
    },
    {
      "epoch": 16.642835865355973,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0025022900148151937,
      "loss": 0.7944,
      "num_input_tokens_seen": 64877232,
      "step": 111740
    },
    {
      "epoch": 16.64358057789693,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0025012119606228544,
      "loss": 0.7734,
      "num_input_tokens_seen": 64880016,
      "step": 111745
    },
    {
      "epoch": 16.64432529043789,
      "grad_norm": 0.0859375,
      "learning_rate": 0.002500134117584829,
      "loss": 0.7962,
      "num_input_tokens_seen": 64882768,
      "step": 111750
    },
    {
      "epoch": 16.64507000297885,
      "grad_norm": 0.203125,
      "learning_rate": 0.0024990564857193397,
      "loss": 0.8273,
      "num_input_tokens_seen": 64885936,
      "step": 111755
    },
    {
      "epoch": 16.64581471551981,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0024979790650445804,
      "loss": 0.7768,
      "num_input_tokens_seen": 64888656,
      "step": 111760
    },
    {
      "epoch": 16.64655942806077,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.002496901855578763,
      "loss": 0.8051,
      "num_input_tokens_seen": 64891568,
      "step": 111765
    },
    {
      "epoch": 16.647304140601726,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0024958248573400763,
      "loss": 0.8004,
      "num_input_tokens_seen": 64894384,
      "step": 111770
    },
    {
      "epoch": 16.648048853142686,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0024947480703467243,
      "loss": 0.79,
      "num_input_tokens_seen": 64897296,
      "step": 111775
    },
    {
      "epoch": 16.648793565683647,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.002493671494616894,
      "loss": 0.8053,
      "num_input_tokens_seen": 64899984,
      "step": 111780
    },
    {
      "epoch": 16.649538278224604,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0024925951301687727,
      "loss": 0.7853,
      "num_input_tokens_seen": 64902672,
      "step": 111785
    },
    {
      "epoch": 16.650282990765565,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.002491518977020542,
      "loss": 0.8044,
      "num_input_tokens_seen": 64905328,
      "step": 111790
    },
    {
      "epoch": 16.651027703306525,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0024904430351903877,
      "loss": 0.8155,
      "num_input_tokens_seen": 64908080,
      "step": 111795
    },
    {
      "epoch": 16.651772415847482,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.002489367304696485,
      "loss": 0.7751,
      "num_input_tokens_seen": 64910960,
      "step": 111800
    },
    {
      "epoch": 16.652517128388443,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.002488291785557002,
      "loss": 0.7941,
      "num_input_tokens_seen": 64913776,
      "step": 111805
    },
    {
      "epoch": 16.6532618409294,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.002487216477790116,
      "loss": 0.7719,
      "num_input_tokens_seen": 64916656,
      "step": 111810
    },
    {
      "epoch": 16.65400655347036,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.002486141381413992,
      "loss": 0.8136,
      "num_input_tokens_seen": 64919440,
      "step": 111815
    },
    {
      "epoch": 16.65475126601132,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0024850664964467866,
      "loss": 0.7677,
      "num_input_tokens_seen": 64922288,
      "step": 111820
    },
    {
      "epoch": 16.655495978552278,
      "grad_norm": 0.123046875,
      "learning_rate": 0.002483991822906665,
      "loss": 0.8099,
      "num_input_tokens_seen": 64925488,
      "step": 111825
    },
    {
      "epoch": 16.656240691093238,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0024829173608117805,
      "loss": 0.7849,
      "num_input_tokens_seen": 64928496,
      "step": 111830
    },
    {
      "epoch": 16.6569854036342,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0024818431101802873,
      "loss": 0.8284,
      "num_input_tokens_seen": 64931312,
      "step": 111835
    },
    {
      "epoch": 16.657730116175156,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0024807690710303264,
      "loss": 0.7708,
      "num_input_tokens_seen": 64934448,
      "step": 111840
    },
    {
      "epoch": 16.658474828716116,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.002479695243380055,
      "loss": 0.8078,
      "num_input_tokens_seen": 64937776,
      "step": 111845
    },
    {
      "epoch": 16.659219541257073,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0024786216272476053,
      "loss": 0.7796,
      "num_input_tokens_seen": 64940912,
      "step": 111850
    },
    {
      "epoch": 16.659964253798034,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.0024775482226511176,
      "loss": 0.7596,
      "num_input_tokens_seen": 64943696,
      "step": 111855
    },
    {
      "epoch": 16.660708966338994,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0024764750296087257,
      "loss": 0.8412,
      "num_input_tokens_seen": 64946608,
      "step": 111860
    },
    {
      "epoch": 16.66145367887995,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.002475402048138555,
      "loss": 0.7859,
      "num_input_tokens_seen": 64949712,
      "step": 111865
    },
    {
      "epoch": 16.66219839142091,
      "grad_norm": 0.09375,
      "learning_rate": 0.0024743292782587415,
      "loss": 0.8104,
      "num_input_tokens_seen": 64952336,
      "step": 111870
    },
    {
      "epoch": 16.662943103961872,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0024732567199874022,
      "loss": 0.7981,
      "num_input_tokens_seen": 64955568,
      "step": 111875
    },
    {
      "epoch": 16.66368781650283,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.0024721843733426607,
      "loss": 0.8314,
      "num_input_tokens_seen": 64958416,
      "step": 111880
    },
    {
      "epoch": 16.66443252904379,
      "grad_norm": 0.037109375,
      "learning_rate": 0.002471112238342629,
      "loss": 0.7956,
      "num_input_tokens_seen": 64961168,
      "step": 111885
    },
    {
      "epoch": 16.665177241584747,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.002470040315005425,
      "loss": 0.7944,
      "num_input_tokens_seen": 64963792,
      "step": 111890
    },
    {
      "epoch": 16.665921954125707,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.002468968603349155,
      "loss": 0.807,
      "num_input_tokens_seen": 64966608,
      "step": 111895
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0024678971033919223,
      "loss": 0.7966,
      "num_input_tokens_seen": 64969296,
      "step": 111900
    },
    {
      "epoch": 16.667411379207625,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.002466825815151833,
      "loss": 0.8142,
      "num_input_tokens_seen": 64972240,
      "step": 111905
    },
    {
      "epoch": 16.668156091748585,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0024657547386469814,
      "loss": 0.7954,
      "num_input_tokens_seen": 64975024,
      "step": 111910
    },
    {
      "epoch": 16.668900804289546,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.002464683873895466,
      "loss": 0.7894,
      "num_input_tokens_seen": 64978000,
      "step": 111915
    },
    {
      "epoch": 16.669645516830503,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0024636132209153766,
      "loss": 0.8142,
      "num_input_tokens_seen": 64981104,
      "step": 111920
    },
    {
      "epoch": 16.670390229371463,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.002462542779724801,
      "loss": 0.7922,
      "num_input_tokens_seen": 64984048,
      "step": 111925
    },
    {
      "epoch": 16.67113494191242,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0024614725503418187,
      "loss": 0.7878,
      "num_input_tokens_seen": 64986992,
      "step": 111930
    },
    {
      "epoch": 16.67187965445338,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0024604025327845183,
      "loss": 0.8118,
      "num_input_tokens_seen": 64989744,
      "step": 111935
    },
    {
      "epoch": 16.67262436699434,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0024593327270709723,
      "loss": 0.7819,
      "num_input_tokens_seen": 64992496,
      "step": 111940
    },
    {
      "epoch": 16.6733690795353,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0024582631332192495,
      "loss": 0.7998,
      "num_input_tokens_seen": 64995952,
      "step": 111945
    },
    {
      "epoch": 16.67411379207626,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00245719375124743,
      "loss": 0.8095,
      "num_input_tokens_seen": 64998960,
      "step": 111950
    },
    {
      "epoch": 16.67485850461722,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.002456124581173568,
      "loss": 0.8305,
      "num_input_tokens_seen": 65001936,
      "step": 111955
    },
    {
      "epoch": 16.675603217158177,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0024550556230157378,
      "loss": 0.8103,
      "num_input_tokens_seen": 65005392,
      "step": 111960
    },
    {
      "epoch": 16.676347929699137,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0024539868767919876,
      "loss": 0.7823,
      "num_input_tokens_seen": 65008144,
      "step": 111965
    },
    {
      "epoch": 16.677092642240094,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0024529183425203815,
      "loss": 0.7991,
      "num_input_tokens_seen": 65011152,
      "step": 111970
    },
    {
      "epoch": 16.677837354781055,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.0024518500202189686,
      "loss": 0.8163,
      "num_input_tokens_seen": 65013808,
      "step": 111975
    },
    {
      "epoch": 16.678582067322015,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0024507819099057926,
      "loss": 0.8178,
      "num_input_tokens_seen": 65016944,
      "step": 111980
    },
    {
      "epoch": 16.679326779862972,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.002449714011598906,
      "loss": 0.8019,
      "num_input_tokens_seen": 65019696,
      "step": 111985
    },
    {
      "epoch": 16.680071492403933,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.002448646325316344,
      "loss": 0.8138,
      "num_input_tokens_seen": 65022544,
      "step": 111990
    },
    {
      "epoch": 16.68081620494489,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.002447578851076147,
      "loss": 0.8138,
      "num_input_tokens_seen": 65025488,
      "step": 111995
    },
    {
      "epoch": 16.68156091748585,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.002446511588896343,
      "loss": 0.7962,
      "num_input_tokens_seen": 65028336,
      "step": 112000
    },
    {
      "epoch": 16.68230563002681,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.002445444538794972,
      "loss": 0.8013,
      "num_input_tokens_seen": 65031184,
      "step": 112005
    },
    {
      "epoch": 16.683050342567768,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0024443777007900494,
      "loss": 0.7915,
      "num_input_tokens_seen": 65033904,
      "step": 112010
    },
    {
      "epoch": 16.683795055108728,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.00244331107489961,
      "loss": 0.7877,
      "num_input_tokens_seen": 65037040,
      "step": 112015
    },
    {
      "epoch": 16.68453976764969,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0024422446611416684,
      "loss": 0.7973,
      "num_input_tokens_seen": 65040016,
      "step": 112020
    },
    {
      "epoch": 16.685284480190646,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.002441178459534236,
      "loss": 0.7787,
      "num_input_tokens_seen": 65042704,
      "step": 112025
    },
    {
      "epoch": 16.686029192731606,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.002440112470095333,
      "loss": 0.7931,
      "num_input_tokens_seen": 65045360,
      "step": 112030
    },
    {
      "epoch": 16.686773905272563,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.00243904669284296,
      "loss": 0.8128,
      "num_input_tokens_seen": 65048464,
      "step": 112035
    },
    {
      "epoch": 16.687518617813524,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0024379811277951305,
      "loss": 0.7922,
      "num_input_tokens_seen": 65051376,
      "step": 112040
    },
    {
      "epoch": 16.688263330354484,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.002436915774969839,
      "loss": 0.7944,
      "num_input_tokens_seen": 65054256,
      "step": 112045
    },
    {
      "epoch": 16.68900804289544,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0024358506343850916,
      "loss": 0.8107,
      "num_input_tokens_seen": 65057008,
      "step": 112050
    },
    {
      "epoch": 16.689752755436402,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0024347857060588777,
      "loss": 0.789,
      "num_input_tokens_seen": 65060048,
      "step": 112055
    },
    {
      "epoch": 16.690497467977362,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0024337209900091887,
      "loss": 0.795,
      "num_input_tokens_seen": 65063056,
      "step": 112060
    },
    {
      "epoch": 16.69124218051832,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0024326564862540115,
      "loss": 0.7924,
      "num_input_tokens_seen": 65066000,
      "step": 112065
    },
    {
      "epoch": 16.69198689305928,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.0024315921948113257,
      "loss": 0.7844,
      "num_input_tokens_seen": 65068816,
      "step": 112070
    },
    {
      "epoch": 16.692731605600237,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.002430528115699121,
      "loss": 0.8045,
      "num_input_tokens_seen": 65071760,
      "step": 112075
    },
    {
      "epoch": 16.693476318141197,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0024294642489353633,
      "loss": 0.7961,
      "num_input_tokens_seen": 65074768,
      "step": 112080
    },
    {
      "epoch": 16.694221030682158,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0024284005945380354,
      "loss": 0.7988,
      "num_input_tokens_seen": 65077808,
      "step": 112085
    },
    {
      "epoch": 16.694965743223115,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0024273371525250992,
      "loss": 0.8037,
      "num_input_tokens_seen": 65080720,
      "step": 112090
    },
    {
      "epoch": 16.695710455764075,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.002426273922914527,
      "loss": 0.7885,
      "num_input_tokens_seen": 65083440,
      "step": 112095
    },
    {
      "epoch": 16.696455168305036,
      "grad_norm": 0.0390625,
      "learning_rate": 0.002425210905724276,
      "loss": 0.7936,
      "num_input_tokens_seen": 65086608,
      "step": 112100
    },
    {
      "epoch": 16.697199880845993,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0024241481009723035,
      "loss": 0.8024,
      "num_input_tokens_seen": 65089232,
      "step": 112105
    },
    {
      "epoch": 16.697944593386953,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0024230855086765724,
      "loss": 0.7928,
      "num_input_tokens_seen": 65091792,
      "step": 112110
    },
    {
      "epoch": 16.69868930592791,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0024220231288550275,
      "loss": 0.7962,
      "num_input_tokens_seen": 65094832,
      "step": 112115
    },
    {
      "epoch": 16.69943401846887,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.002420960961525615,
      "loss": 0.792,
      "num_input_tokens_seen": 65097808,
      "step": 112120
    },
    {
      "epoch": 16.70017873100983,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0024198990067062854,
      "loss": 0.7721,
      "num_input_tokens_seen": 65100784,
      "step": 112125
    },
    {
      "epoch": 16.70092344355079,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0024188372644149774,
      "loss": 0.8109,
      "num_input_tokens_seen": 65103696,
      "step": 112130
    },
    {
      "epoch": 16.70166815609175,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.002417775734669622,
      "loss": 0.7864,
      "num_input_tokens_seen": 65106448,
      "step": 112135
    },
    {
      "epoch": 16.702412868632706,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0024167144174881616,
      "loss": 0.79,
      "num_input_tokens_seen": 65109264,
      "step": 112140
    },
    {
      "epoch": 16.703157581173667,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.002415653312888523,
      "loss": 0.7879,
      "num_input_tokens_seen": 65112240,
      "step": 112145
    },
    {
      "epoch": 16.703902293714627,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0024145924208886257,
      "loss": 0.8142,
      "num_input_tokens_seen": 65115408,
      "step": 112150
    },
    {
      "epoch": 16.704647006255584,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.002413531741506404,
      "loss": 0.8148,
      "num_input_tokens_seen": 65118320,
      "step": 112155
    },
    {
      "epoch": 16.705391718796545,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0024124712747597677,
      "loss": 0.7854,
      "num_input_tokens_seen": 65121296,
      "step": 112160
    },
    {
      "epoch": 16.706136431337505,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0024114110206666377,
      "loss": 0.7948,
      "num_input_tokens_seen": 65124080,
      "step": 112165
    },
    {
      "epoch": 16.706881143878462,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0024103509792449216,
      "loss": 0.7809,
      "num_input_tokens_seen": 65126736,
      "step": 112170
    },
    {
      "epoch": 16.707625856419423,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0024092911505125335,
      "loss": 0.7973,
      "num_input_tokens_seen": 65129648,
      "step": 112175
    },
    {
      "epoch": 16.70837056896038,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.002408231534487374,
      "loss": 0.7751,
      "num_input_tokens_seen": 65132848,
      "step": 112180
    },
    {
      "epoch": 16.70911528150134,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.0024071721311873465,
      "loss": 0.796,
      "num_input_tokens_seen": 65135504,
      "step": 112185
    },
    {
      "epoch": 16.7098599940423,
      "grad_norm": 0.029296875,
      "learning_rate": 0.002406112940630346,
      "loss": 0.7938,
      "num_input_tokens_seen": 65138288,
      "step": 112190
    },
    {
      "epoch": 16.710604706583258,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.002405053962834264,
      "loss": 0.7853,
      "num_input_tokens_seen": 65140944,
      "step": 112195
    },
    {
      "epoch": 16.71134941912422,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.002403995197816999,
      "loss": 0.7985,
      "num_input_tokens_seen": 65144048,
      "step": 112200
    },
    {
      "epoch": 16.71209413166518,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0024029366455964284,
      "loss": 0.8001,
      "num_input_tokens_seen": 65146800,
      "step": 112205
    },
    {
      "epoch": 16.712838844206136,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.002401878306190444,
      "loss": 0.7977,
      "num_input_tokens_seen": 65149712,
      "step": 112210
    },
    {
      "epoch": 16.713583556747096,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0024008201796169183,
      "loss": 0.7883,
      "num_input_tokens_seen": 65152528,
      "step": 112215
    },
    {
      "epoch": 16.714328269288053,
      "grad_norm": 0.044921875,
      "learning_rate": 0.002399762265893733,
      "loss": 0.8075,
      "num_input_tokens_seen": 65155248,
      "step": 112220
    },
    {
      "epoch": 16.715072981829014,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.002398704565038759,
      "loss": 0.7847,
      "num_input_tokens_seen": 65158128,
      "step": 112225
    },
    {
      "epoch": 16.715817694369974,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.002397647077069859,
      "loss": 0.7875,
      "num_input_tokens_seen": 65160912,
      "step": 112230
    },
    {
      "epoch": 16.71656240691093,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.002396589802004908,
      "loss": 0.8097,
      "num_input_tokens_seen": 65163760,
      "step": 112235
    },
    {
      "epoch": 16.717307119451892,
      "grad_norm": 0.357421875,
      "learning_rate": 0.0023955327398617575,
      "loss": 0.8527,
      "num_input_tokens_seen": 65166512,
      "step": 112240
    },
    {
      "epoch": 16.718051831992852,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0023944758906582764,
      "loss": 0.7897,
      "num_input_tokens_seen": 65169776,
      "step": 112245
    },
    {
      "epoch": 16.71879654453381,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.002393419254412313,
      "loss": 0.7758,
      "num_input_tokens_seen": 65172752,
      "step": 112250
    },
    {
      "epoch": 16.71954125707477,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.002392362831141718,
      "loss": 0.7976,
      "num_input_tokens_seen": 65176240,
      "step": 112255
    },
    {
      "epoch": 16.720285969615727,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.00239130662086434,
      "loss": 0.8076,
      "num_input_tokens_seen": 65179408,
      "step": 112260
    },
    {
      "epoch": 16.721030682156687,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.0023902506235980166,
      "loss": 0.8097,
      "num_input_tokens_seen": 65182096,
      "step": 112265
    },
    {
      "epoch": 16.721775394697648,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.002389194839360597,
      "loss": 0.7718,
      "num_input_tokens_seen": 65185072,
      "step": 112270
    },
    {
      "epoch": 16.722520107238605,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0023881392681699113,
      "loss": 0.8042,
      "num_input_tokens_seen": 65187984,
      "step": 112275
    },
    {
      "epoch": 16.723264819779565,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0023870839100437967,
      "loss": 0.7993,
      "num_input_tokens_seen": 65190960,
      "step": 112280
    },
    {
      "epoch": 16.724009532320522,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0023860287650000773,
      "loss": 0.7941,
      "num_input_tokens_seen": 65194000,
      "step": 112285
    },
    {
      "epoch": 16.724754244861483,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.002384973833056585,
      "loss": 0.8068,
      "num_input_tokens_seen": 65197136,
      "step": 112290
    },
    {
      "epoch": 16.725498957402444,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0023839191142311362,
      "loss": 0.8054,
      "num_input_tokens_seen": 65200016,
      "step": 112295
    },
    {
      "epoch": 16.7262436699434,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.002382864608541553,
      "loss": 0.7873,
      "num_input_tokens_seen": 65202736,
      "step": 112300
    },
    {
      "epoch": 16.72698838248436,
      "grad_norm": 0.05859375,
      "learning_rate": 0.002381810316005649,
      "loss": 0.7887,
      "num_input_tokens_seen": 65205680,
      "step": 112305
    },
    {
      "epoch": 16.72773309502532,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0023807562366412326,
      "loss": 0.7949,
      "num_input_tokens_seen": 65208400,
      "step": 112310
    },
    {
      "epoch": 16.72847780756628,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0023797023704661175,
      "loss": 0.7989,
      "num_input_tokens_seen": 65211312,
      "step": 112315
    },
    {
      "epoch": 16.72922252010724,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.002378648717498104,
      "loss": 0.7952,
      "num_input_tokens_seen": 65213936,
      "step": 112320
    },
    {
      "epoch": 16.7299672326482,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.002377595277754992,
      "loss": 0.7814,
      "num_input_tokens_seen": 65216656,
      "step": 112325
    },
    {
      "epoch": 16.730711945189157,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.0023765420512545763,
      "loss": 0.8032,
      "num_input_tokens_seen": 65219600,
      "step": 112330
    },
    {
      "epoch": 16.731456657730117,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0023754890380146554,
      "loss": 0.8079,
      "num_input_tokens_seen": 65222608,
      "step": 112335
    },
    {
      "epoch": 16.732201370271074,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0023744362380530176,
      "loss": 0.7999,
      "num_input_tokens_seen": 65225520,
      "step": 112340
    },
    {
      "epoch": 16.732946082812035,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0023733836513874424,
      "loss": 0.8242,
      "num_input_tokens_seen": 65228560,
      "step": 112345
    },
    {
      "epoch": 16.733690795352995,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0023723312780357217,
      "loss": 0.8219,
      "num_input_tokens_seen": 65231152,
      "step": 112350
    },
    {
      "epoch": 16.734435507893952,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0023712791180156255,
      "loss": 0.7992,
      "num_input_tokens_seen": 65234128,
      "step": 112355
    },
    {
      "epoch": 16.735180220434913,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.002370227171344939,
      "loss": 0.8043,
      "num_input_tokens_seen": 65236656,
      "step": 112360
    },
    {
      "epoch": 16.73592493297587,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0023691754380414233,
      "loss": 0.8023,
      "num_input_tokens_seen": 65239792,
      "step": 112365
    },
    {
      "epoch": 16.73666964551683,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0023681239181228535,
      "loss": 0.795,
      "num_input_tokens_seen": 65242832,
      "step": 112370
    },
    {
      "epoch": 16.73741435805779,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0023670726116069895,
      "loss": 0.794,
      "num_input_tokens_seen": 65245776,
      "step": 112375
    },
    {
      "epoch": 16.738159070598748,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0023660215185115968,
      "loss": 0.8078,
      "num_input_tokens_seen": 65248624,
      "step": 112380
    },
    {
      "epoch": 16.73890378313971,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0023649706388544313,
      "loss": 0.814,
      "num_input_tokens_seen": 65251440,
      "step": 112385
    },
    {
      "epoch": 16.73964849568067,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0023639199726532435,
      "loss": 0.8,
      "num_input_tokens_seen": 65254320,
      "step": 112390
    },
    {
      "epoch": 16.740393208221626,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.002362869519925786,
      "loss": 0.8049,
      "num_input_tokens_seen": 65257200,
      "step": 112395
    },
    {
      "epoch": 16.741137920762586,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0023618192806897998,
      "loss": 0.8025,
      "num_input_tokens_seen": 65260144,
      "step": 112400
    },
    {
      "epoch": 16.741882633303543,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.002360769254963037,
      "loss": 0.7783,
      "num_input_tokens_seen": 65262864,
      "step": 112405
    },
    {
      "epoch": 16.742627345844504,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0023597194427632273,
      "loss": 0.7993,
      "num_input_tokens_seen": 65265616,
      "step": 112410
    },
    {
      "epoch": 16.743372058385464,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0023586698441081144,
      "loss": 0.7949,
      "num_input_tokens_seen": 65268528,
      "step": 112415
    },
    {
      "epoch": 16.74411677092642,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.002357620459015426,
      "loss": 0.7942,
      "num_input_tokens_seen": 65271472,
      "step": 112420
    },
    {
      "epoch": 16.744861483467382,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0023565712875028874,
      "loss": 0.8271,
      "num_input_tokens_seen": 65274416,
      "step": 112425
    },
    {
      "epoch": 16.745606196008342,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0023555223295882305,
      "loss": 0.8079,
      "num_input_tokens_seen": 65277584,
      "step": 112430
    },
    {
      "epoch": 16.7463509085493,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0023544735852891684,
      "loss": 0.8112,
      "num_input_tokens_seen": 65280528,
      "step": 112435
    },
    {
      "epoch": 16.74709562109026,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0023534250546234274,
      "loss": 0.7922,
      "num_input_tokens_seen": 65283312,
      "step": 112440
    },
    {
      "epoch": 16.747840333631217,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.002352376737608715,
      "loss": 0.7836,
      "num_input_tokens_seen": 65286224,
      "step": 112445
    },
    {
      "epoch": 16.748585046172177,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.00235132863426274,
      "loss": 0.781,
      "num_input_tokens_seen": 65289296,
      "step": 112450
    },
    {
      "epoch": 16.749329758713138,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0023502807446032174,
      "loss": 0.8089,
      "num_input_tokens_seen": 65291856,
      "step": 112455
    },
    {
      "epoch": 16.750074471254095,
      "grad_norm": 0.04296875,
      "learning_rate": 0.002349233068647844,
      "loss": 0.7983,
      "num_input_tokens_seen": 65294704,
      "step": 112460
    },
    {
      "epoch": 16.750819183795056,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0023481856064143207,
      "loss": 0.7926,
      "num_input_tokens_seen": 65297520,
      "step": 112465
    },
    {
      "epoch": 16.751563896336016,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0023471383579203387,
      "loss": 0.792,
      "num_input_tokens_seen": 65300496,
      "step": 112470
    },
    {
      "epoch": 16.752308608876973,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.002346091323183598,
      "loss": 0.8013,
      "num_input_tokens_seen": 65303344,
      "step": 112475
    },
    {
      "epoch": 16.753053321417934,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.002345044502221779,
      "loss": 0.7829,
      "num_input_tokens_seen": 65306320,
      "step": 112480
    },
    {
      "epoch": 16.75379803395889,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0023439978950525768,
      "loss": 0.8126,
      "num_input_tokens_seen": 65309136,
      "step": 112485
    },
    {
      "epoch": 16.75454274649985,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0023429515016936625,
      "loss": 0.7831,
      "num_input_tokens_seen": 65311952,
      "step": 112490
    },
    {
      "epoch": 16.75528745904081,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0023419053221627214,
      "loss": 0.8069,
      "num_input_tokens_seen": 65314896,
      "step": 112495
    },
    {
      "epoch": 16.75603217158177,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0023408593564774255,
      "loss": 0.8037,
      "num_input_tokens_seen": 65317456,
      "step": 112500
    },
    {
      "epoch": 16.75677688412273,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0023398136046554412,
      "loss": 0.7867,
      "num_input_tokens_seen": 65320240,
      "step": 112505
    },
    {
      "epoch": 16.757521596663686,
      "grad_norm": 0.0703125,
      "learning_rate": 0.002338768066714443,
      "loss": 0.799,
      "num_input_tokens_seen": 65323376,
      "step": 112510
    },
    {
      "epoch": 16.758266309204647,
      "grad_norm": 0.03125,
      "learning_rate": 0.002337722742672089,
      "loss": 0.7881,
      "num_input_tokens_seen": 65326416,
      "step": 112515
    },
    {
      "epoch": 16.759011021745607,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0023366776325460395,
      "loss": 0.7896,
      "num_input_tokens_seen": 65329200,
      "step": 112520
    },
    {
      "epoch": 16.759755734286564,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.002335632736353949,
      "loss": 0.7882,
      "num_input_tokens_seen": 65331888,
      "step": 112525
    },
    {
      "epoch": 16.760500446827525,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.002334588054113475,
      "loss": 0.7964,
      "num_input_tokens_seen": 65335056,
      "step": 112530
    },
    {
      "epoch": 16.761245159368485,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.002333543585842258,
      "loss": 0.7902,
      "num_input_tokens_seen": 65338064,
      "step": 112535
    },
    {
      "epoch": 16.761989871909442,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0023324993315579543,
      "loss": 0.7894,
      "num_input_tokens_seen": 65341008,
      "step": 112540
    },
    {
      "epoch": 16.762734584450403,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0023314552912781982,
      "loss": 0.7976,
      "num_input_tokens_seen": 65343568,
      "step": 112545
    },
    {
      "epoch": 16.76347929699136,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0023304114650206267,
      "loss": 0.802,
      "num_input_tokens_seen": 65346320,
      "step": 112550
    },
    {
      "epoch": 16.76422400953232,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.0023293678528028797,
      "loss": 0.7866,
      "num_input_tokens_seen": 65349168,
      "step": 112555
    },
    {
      "epoch": 16.76496872207328,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.002328324454642581,
      "loss": 0.8225,
      "num_input_tokens_seen": 65352208,
      "step": 112560
    },
    {
      "epoch": 16.765713434614238,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.002327281270557365,
      "loss": 0.8077,
      "num_input_tokens_seen": 65355248,
      "step": 112565
    },
    {
      "epoch": 16.7664581471552,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0023262383005648495,
      "loss": 0.7984,
      "num_input_tokens_seen": 65358128,
      "step": 112570
    },
    {
      "epoch": 16.76720285969616,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0023251955446826618,
      "loss": 0.7911,
      "num_input_tokens_seen": 65361232,
      "step": 112575
    },
    {
      "epoch": 16.767947572237116,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.002324153002928411,
      "loss": 0.7934,
      "num_input_tokens_seen": 65364016,
      "step": 112580
    },
    {
      "epoch": 16.768692284778076,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0023231106753197127,
      "loss": 0.8075,
      "num_input_tokens_seen": 65366832,
      "step": 112585
    },
    {
      "epoch": 16.769436997319033,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0023220685618741753,
      "loss": 0.7717,
      "num_input_tokens_seen": 65369488,
      "step": 112590
    },
    {
      "epoch": 16.770181709859994,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0023210266626094013,
      "loss": 0.7752,
      "num_input_tokens_seen": 65372176,
      "step": 112595
    },
    {
      "epoch": 16.770926422400954,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0023199849775429977,
      "loss": 0.8157,
      "num_input_tokens_seen": 65374960,
      "step": 112600
    },
    {
      "epoch": 16.77167113494191,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0023189435066925572,
      "loss": 0.7875,
      "num_input_tokens_seen": 65377936,
      "step": 112605
    },
    {
      "epoch": 16.772415847482872,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0023179022500756812,
      "loss": 0.8622,
      "num_input_tokens_seen": 65380976,
      "step": 112610
    },
    {
      "epoch": 16.773160560023832,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0023168612077099555,
      "loss": 0.7912,
      "num_input_tokens_seen": 65384176,
      "step": 112615
    },
    {
      "epoch": 16.77390527256479,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.0023158203796129705,
      "loss": 0.8042,
      "num_input_tokens_seen": 65387088,
      "step": 112620
    },
    {
      "epoch": 16.77464998510575,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0023147797658023095,
      "loss": 0.7883,
      "num_input_tokens_seen": 65390096,
      "step": 112625
    },
    {
      "epoch": 16.775394697646707,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0023137393662955477,
      "loss": 0.7941,
      "num_input_tokens_seen": 65392976,
      "step": 112630
    },
    {
      "epoch": 16.776139410187668,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.00231269918111027,
      "loss": 0.8009,
      "num_input_tokens_seen": 65395696,
      "step": 112635
    },
    {
      "epoch": 16.776884122728628,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0023116592102640408,
      "loss": 0.8029,
      "num_input_tokens_seen": 65398640,
      "step": 112640
    },
    {
      "epoch": 16.777628835269585,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.002310619453774437,
      "loss": 0.7774,
      "num_input_tokens_seen": 65401520,
      "step": 112645
    },
    {
      "epoch": 16.778373547810546,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0023095799116590204,
      "loss": 0.7931,
      "num_input_tokens_seen": 65404080,
      "step": 112650
    },
    {
      "epoch": 16.779118260351503,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.0023085405839353527,
      "loss": 0.7965,
      "num_input_tokens_seen": 65407024,
      "step": 112655
    },
    {
      "epoch": 16.779862972892463,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.002307501470620994,
      "loss": 0.8024,
      "num_input_tokens_seen": 65410256,
      "step": 112660
    },
    {
      "epoch": 16.780607685433424,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0023064625717334948,
      "loss": 0.8042,
      "num_input_tokens_seen": 65413232,
      "step": 112665
    },
    {
      "epoch": 16.78135239797438,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.002305423887290411,
      "loss": 0.8043,
      "num_input_tokens_seen": 65416048,
      "step": 112670
    },
    {
      "epoch": 16.78209711051534,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.002304385417309285,
      "loss": 0.8,
      "num_input_tokens_seen": 65418704,
      "step": 112675
    },
    {
      "epoch": 16.7828418230563,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0023033471618076687,
      "loss": 0.801,
      "num_input_tokens_seen": 65421392,
      "step": 112680
    },
    {
      "epoch": 16.78358653559726,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0023023091208030943,
      "loss": 0.7886,
      "num_input_tokens_seen": 65424144,
      "step": 112685
    },
    {
      "epoch": 16.78433124813822,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0023012712943131057,
      "loss": 0.7936,
      "num_input_tokens_seen": 65427088,
      "step": 112690
    },
    {
      "epoch": 16.785075960679176,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0023002336823552275,
      "loss": 0.8107,
      "num_input_tokens_seen": 65429872,
      "step": 112695
    },
    {
      "epoch": 16.785820673220137,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0022991962849469977,
      "loss": 0.7896,
      "num_input_tokens_seen": 65432912,
      "step": 112700
    },
    {
      "epoch": 16.786565385761097,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0022981591021059384,
      "loss": 0.7956,
      "num_input_tokens_seen": 65435760,
      "step": 112705
    },
    {
      "epoch": 16.787310098302054,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.002297122133849569,
      "loss": 0.8003,
      "num_input_tokens_seen": 65438448,
      "step": 112710
    },
    {
      "epoch": 16.788054810843015,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.0022960853801954123,
      "loss": 0.7906,
      "num_input_tokens_seen": 65441360,
      "step": 112715
    },
    {
      "epoch": 16.788799523383975,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.0022950488411609825,
      "loss": 0.7983,
      "num_input_tokens_seen": 65444400,
      "step": 112720
    },
    {
      "epoch": 16.789544235924932,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.002294012516763789,
      "loss": 0.7912,
      "num_input_tokens_seen": 65447504,
      "step": 112725
    },
    {
      "epoch": 16.790288948465893,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0022929764070213363,
      "loss": 0.7921,
      "num_input_tokens_seen": 65450544,
      "step": 112730
    },
    {
      "epoch": 16.79103366100685,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.002291940511951137,
      "loss": 0.8058,
      "num_input_tokens_seen": 65453456,
      "step": 112735
    },
    {
      "epoch": 16.79177837354781,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0022909048315706837,
      "loss": 0.8059,
      "num_input_tokens_seen": 65456400,
      "step": 112740
    },
    {
      "epoch": 16.79252308608877,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.002289869365897479,
      "loss": 0.8086,
      "num_input_tokens_seen": 65459280,
      "step": 112745
    },
    {
      "epoch": 16.793267798629728,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.002288834114949014,
      "loss": 0.8233,
      "num_input_tokens_seen": 65462128,
      "step": 112750
    },
    {
      "epoch": 16.79401251117069,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0022877990787427724,
      "loss": 0.7887,
      "num_input_tokens_seen": 65464944,
      "step": 112755
    },
    {
      "epoch": 16.79475722371165,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.00228676425729625,
      "loss": 0.8165,
      "num_input_tokens_seen": 65468112,
      "step": 112760
    },
    {
      "epoch": 16.795501936252606,
      "grad_norm": 0.03125,
      "learning_rate": 0.0022857296506269215,
      "loss": 0.7904,
      "num_input_tokens_seen": 65470992,
      "step": 112765
    },
    {
      "epoch": 16.796246648793566,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.00228469525875227,
      "loss": 0.7886,
      "num_input_tokens_seen": 65473648,
      "step": 112770
    },
    {
      "epoch": 16.796991361334523,
      "grad_norm": 0.044921875,
      "learning_rate": 0.00228366108168977,
      "loss": 0.8207,
      "num_input_tokens_seen": 65476528,
      "step": 112775
    },
    {
      "epoch": 16.797736073875484,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.002282627119456889,
      "loss": 0.8242,
      "num_input_tokens_seen": 65479376,
      "step": 112780
    },
    {
      "epoch": 16.798480786416444,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0022815933720711002,
      "loss": 0.797,
      "num_input_tokens_seen": 65482064,
      "step": 112785
    },
    {
      "epoch": 16.7992254989574,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0022805598395498655,
      "loss": 0.7725,
      "num_input_tokens_seen": 65484848,
      "step": 112790
    },
    {
      "epoch": 16.799970211498362,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.002279526521910643,
      "loss": 0.7832,
      "num_input_tokens_seen": 65487728,
      "step": 112795
    },
    {
      "epoch": 16.800714924039323,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.002278493419170889,
      "loss": 0.8254,
      "num_input_tokens_seen": 65490768,
      "step": 112800
    },
    {
      "epoch": 16.80145963658028,
      "grad_norm": 0.068359375,
      "learning_rate": 0.002277460531348064,
      "loss": 0.7853,
      "num_input_tokens_seen": 65493712,
      "step": 112805
    },
    {
      "epoch": 16.80220434912124,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0022764278584596074,
      "loss": 0.8102,
      "num_input_tokens_seen": 65496368,
      "step": 112810
    },
    {
      "epoch": 16.802949061662197,
      "grad_norm": 0.037109375,
      "learning_rate": 0.002275395400522974,
      "loss": 0.8042,
      "num_input_tokens_seen": 65499280,
      "step": 112815
    },
    {
      "epoch": 16.803693774203158,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.0022743631575556,
      "loss": 0.7862,
      "num_input_tokens_seen": 65502128,
      "step": 112820
    },
    {
      "epoch": 16.804438486744118,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.0022733311295749293,
      "loss": 0.7825,
      "num_input_tokens_seen": 65504848,
      "step": 112825
    },
    {
      "epoch": 16.805183199285075,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.002272299316598396,
      "loss": 0.8084,
      "num_input_tokens_seen": 65507856,
      "step": 112830
    },
    {
      "epoch": 16.805927911826036,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0022712677186434256,
      "loss": 0.8102,
      "num_input_tokens_seen": 65510832,
      "step": 112835
    },
    {
      "epoch": 16.806672624366996,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.002270236335727453,
      "loss": 0.7969,
      "num_input_tokens_seen": 65513712,
      "step": 112840
    },
    {
      "epoch": 16.807417336907953,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0022692051678678995,
      "loss": 0.8009,
      "num_input_tokens_seen": 65516560,
      "step": 112845
    },
    {
      "epoch": 16.808162049448914,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.0022681742150821855,
      "loss": 0.8133,
      "num_input_tokens_seen": 65519376,
      "step": 112850
    },
    {
      "epoch": 16.80890676198987,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.002267143477387726,
      "loss": 0.8026,
      "num_input_tokens_seen": 65522384,
      "step": 112855
    },
    {
      "epoch": 16.80965147453083,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0022661129548019387,
      "loss": 0.7939,
      "num_input_tokens_seen": 65525296,
      "step": 112860
    },
    {
      "epoch": 16.81039618707179,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0022650826473422318,
      "loss": 0.7791,
      "num_input_tokens_seen": 65527888,
      "step": 112865
    },
    {
      "epoch": 16.81114089961275,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0022640525550260064,
      "loss": 0.8021,
      "num_input_tokens_seen": 65530736,
      "step": 112870
    },
    {
      "epoch": 16.81188561215371,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.002263022677870672,
      "loss": 0.7968,
      "num_input_tokens_seen": 65533840,
      "step": 112875
    },
    {
      "epoch": 16.812630324694666,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0022619930158936205,
      "loss": 0.7847,
      "num_input_tokens_seen": 65536784,
      "step": 112880
    },
    {
      "epoch": 16.813375037235627,
      "grad_norm": 0.078125,
      "learning_rate": 0.0022609635691122542,
      "loss": 0.8058,
      "num_input_tokens_seen": 65539728,
      "step": 112885
    },
    {
      "epoch": 16.814119749776587,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0022599343375439584,
      "loss": 0.8128,
      "num_input_tokens_seen": 65542352,
      "step": 112890
    },
    {
      "epoch": 16.814864462317544,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.002258905321206126,
      "loss": 0.8011,
      "num_input_tokens_seen": 65545392,
      "step": 112895
    },
    {
      "epoch": 16.815609174858505,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0022578765201161366,
      "loss": 0.8065,
      "num_input_tokens_seen": 65548208,
      "step": 112900
    },
    {
      "epoch": 16.816353887399465,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.002256847934291376,
      "loss": 0.7966,
      "num_input_tokens_seen": 65550736,
      "step": 112905
    },
    {
      "epoch": 16.817098599940422,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.002255819563749217,
      "loss": 0.8148,
      "num_input_tokens_seen": 65553584,
      "step": 112910
    },
    {
      "epoch": 16.817843312481383,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0022547914085070345,
      "loss": 0.7882,
      "num_input_tokens_seen": 65556496,
      "step": 112915
    },
    {
      "epoch": 16.81858802502234,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.0022537634685821976,
      "loss": 0.7958,
      "num_input_tokens_seen": 65559344,
      "step": 112920
    },
    {
      "epoch": 16.8193327375633,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0022527357439920677,
      "loss": 0.7881,
      "num_input_tokens_seen": 65562288,
      "step": 112925
    },
    {
      "epoch": 16.82007745010426,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.002251708234754017,
      "loss": 0.7839,
      "num_input_tokens_seen": 65565136,
      "step": 112930
    },
    {
      "epoch": 16.820822162645218,
      "grad_norm": 0.33984375,
      "learning_rate": 0.002250680940885393,
      "loss": 0.821,
      "num_input_tokens_seen": 65568624,
      "step": 112935
    },
    {
      "epoch": 16.82156687518618,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0022496538624035617,
      "loss": 0.8211,
      "num_input_tokens_seen": 65571248,
      "step": 112940
    },
    {
      "epoch": 16.82231158772714,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0022486269993258695,
      "loss": 0.8044,
      "num_input_tokens_seen": 65574384,
      "step": 112945
    },
    {
      "epoch": 16.823056300268096,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.002247600351669661,
      "loss": 0.7956,
      "num_input_tokens_seen": 65577392,
      "step": 112950
    },
    {
      "epoch": 16.823801012809056,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0022465739194522864,
      "loss": 0.8014,
      "num_input_tokens_seen": 65580624,
      "step": 112955
    },
    {
      "epoch": 16.824545725350013,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.002245547702691081,
      "loss": 0.7739,
      "num_input_tokens_seen": 65583504,
      "step": 112960
    },
    {
      "epoch": 16.825290437890974,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.002244521701403387,
      "loss": 0.8035,
      "num_input_tokens_seen": 65586256,
      "step": 112965
    },
    {
      "epoch": 16.826035150431935,
      "grad_norm": 0.052734375,
      "learning_rate": 0.002243495915606531,
      "loss": 0.7758,
      "num_input_tokens_seen": 65589168,
      "step": 112970
    },
    {
      "epoch": 16.82677986297289,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0022424703453178497,
      "loss": 0.7875,
      "num_input_tokens_seen": 65592048,
      "step": 112975
    },
    {
      "epoch": 16.827524575513852,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0022414449905546656,
      "loss": 0.8034,
      "num_input_tokens_seen": 65594864,
      "step": 112980
    },
    {
      "epoch": 16.828269288054813,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.002240419851334302,
      "loss": 0.8158,
      "num_input_tokens_seen": 65597776,
      "step": 112985
    },
    {
      "epoch": 16.82901400059577,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0022393949276740765,
      "loss": 0.8093,
      "num_input_tokens_seen": 65600592,
      "step": 112990
    },
    {
      "epoch": 16.82975871313673,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.0022383702195913,
      "loss": 0.7802,
      "num_input_tokens_seen": 65603216,
      "step": 112995
    },
    {
      "epoch": 16.830503425677687,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0022373457271032925,
      "loss": 0.7794,
      "num_input_tokens_seen": 65605968,
      "step": 113000
    },
    {
      "epoch": 16.831248138218648,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.002236321450227355,
      "loss": 0.8157,
      "num_input_tokens_seen": 65608784,
      "step": 113005
    },
    {
      "epoch": 16.831992850759608,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.002235297388980796,
      "loss": 0.7798,
      "num_input_tokens_seen": 65611792,
      "step": 113010
    },
    {
      "epoch": 16.832737563300565,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.0022342735433809116,
      "loss": 0.7853,
      "num_input_tokens_seen": 65614768,
      "step": 113015
    },
    {
      "epoch": 16.833482275841526,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0022332499134450037,
      "loss": 0.8001,
      "num_input_tokens_seen": 65617712,
      "step": 113020
    },
    {
      "epoch": 16.834226988382483,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.002232226499190364,
      "loss": 0.7933,
      "num_input_tokens_seen": 65620720,
      "step": 113025
    },
    {
      "epoch": 16.834971700923443,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.002231203300634278,
      "loss": 0.7986,
      "num_input_tokens_seen": 65623568,
      "step": 113030
    },
    {
      "epoch": 16.835716413464404,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.002230180317794037,
      "loss": 0.8041,
      "num_input_tokens_seen": 65626256,
      "step": 113035
    },
    {
      "epoch": 16.83646112600536,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.0022291575506869178,
      "loss": 0.7954,
      "num_input_tokens_seen": 65629232,
      "step": 113040
    },
    {
      "epoch": 16.83720583854632,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0022281349993302055,
      "loss": 0.7941,
      "num_input_tokens_seen": 65632144,
      "step": 113045
    },
    {
      "epoch": 16.837950551087282,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0022271126637411725,
      "loss": 0.7871,
      "num_input_tokens_seen": 65635056,
      "step": 113050
    },
    {
      "epoch": 16.83869526362824,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0022260905439370877,
      "loss": 0.7772,
      "num_input_tokens_seen": 65637840,
      "step": 113055
    },
    {
      "epoch": 16.8394399761692,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.002225068639935217,
      "loss": 0.8054,
      "num_input_tokens_seen": 65641040,
      "step": 113060
    },
    {
      "epoch": 16.840184688710156,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0022240469517528323,
      "loss": 0.7888,
      "num_input_tokens_seen": 65643760,
      "step": 113065
    },
    {
      "epoch": 16.840929401251117,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.002223025479407189,
      "loss": 0.7904,
      "num_input_tokens_seen": 65646768,
      "step": 113070
    },
    {
      "epoch": 16.841674113792077,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0022220042229155407,
      "loss": 0.7909,
      "num_input_tokens_seen": 65649840,
      "step": 113075
    },
    {
      "epoch": 16.842418826333034,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.002220983182295148,
      "loss": 0.8039,
      "num_input_tokens_seen": 65652432,
      "step": 113080
    },
    {
      "epoch": 16.843163538873995,
      "grad_norm": 0.078125,
      "learning_rate": 0.0022199623575632527,
      "loss": 0.7929,
      "num_input_tokens_seen": 65655536,
      "step": 113085
    },
    {
      "epoch": 16.843908251414955,
      "grad_norm": 0.0245361328125,
      "learning_rate": 0.002218941748737107,
      "loss": 0.8098,
      "num_input_tokens_seen": 65658384,
      "step": 113090
    },
    {
      "epoch": 16.844652963955912,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.0022179213558339477,
      "loss": 0.8105,
      "num_input_tokens_seen": 65661232,
      "step": 113095
    },
    {
      "epoch": 16.845397676496873,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0022169011788710207,
      "loss": 0.7892,
      "num_input_tokens_seen": 65663920,
      "step": 113100
    },
    {
      "epoch": 16.84614238903783,
      "grad_norm": 0.048828125,
      "learning_rate": 0.002215881217865554,
      "loss": 0.7866,
      "num_input_tokens_seen": 65666704,
      "step": 113105
    },
    {
      "epoch": 16.84688710157879,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0022148614728347785,
      "loss": 0.7926,
      "num_input_tokens_seen": 65669808,
      "step": 113110
    },
    {
      "epoch": 16.84763181411975,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0022138419437959275,
      "loss": 0.7899,
      "num_input_tokens_seen": 65672752,
      "step": 113115
    },
    {
      "epoch": 16.848376526660708,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0022128226307662206,
      "loss": 0.7893,
      "num_input_tokens_seen": 65675280,
      "step": 113120
    },
    {
      "epoch": 16.84912123920167,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0022118035337628803,
      "loss": 0.7802,
      "num_input_tokens_seen": 65678256,
      "step": 113125
    },
    {
      "epoch": 16.84986595174263,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0022107846528031183,
      "loss": 0.7917,
      "num_input_tokens_seen": 65681200,
      "step": 113130
    },
    {
      "epoch": 16.850610664283586,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0022097659879041537,
      "loss": 0.7924,
      "num_input_tokens_seen": 65684208,
      "step": 113135
    },
    {
      "epoch": 16.851355376824547,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0022087475390831913,
      "loss": 0.7827,
      "num_input_tokens_seen": 65687344,
      "step": 113140
    },
    {
      "epoch": 16.852100089365504,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.002207729306357442,
      "loss": 0.8043,
      "num_input_tokens_seen": 65690480,
      "step": 113145
    },
    {
      "epoch": 16.852844801906464,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0022067112897441042,
      "loss": 0.7986,
      "num_input_tokens_seen": 65693232,
      "step": 113150
    },
    {
      "epoch": 16.853589514447425,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0022056934892603726,
      "loss": 0.7873,
      "num_input_tokens_seen": 65696240,
      "step": 113155
    },
    {
      "epoch": 16.85433422698838,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0022046759049234492,
      "loss": 0.7739,
      "num_input_tokens_seen": 65699344,
      "step": 113160
    },
    {
      "epoch": 16.855078939529342,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0022036585367505193,
      "loss": 0.8075,
      "num_input_tokens_seen": 65702640,
      "step": 113165
    },
    {
      "epoch": 16.8558236520703,
      "grad_norm": 0.0859375,
      "learning_rate": 0.002202641384758776,
      "loss": 0.7786,
      "num_input_tokens_seen": 65705712,
      "step": 113170
    },
    {
      "epoch": 16.85656836461126,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0022016244489654,
      "loss": 0.7983,
      "num_input_tokens_seen": 65708464,
      "step": 113175
    },
    {
      "epoch": 16.85731307715222,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0022006077293875714,
      "loss": 0.7875,
      "num_input_tokens_seen": 65711504,
      "step": 113180
    },
    {
      "epoch": 16.858057789693177,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.002199591226042463,
      "loss": 0.8006,
      "num_input_tokens_seen": 65714320,
      "step": 113185
    },
    {
      "epoch": 16.858802502234138,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0021985749389472537,
      "loss": 0.8114,
      "num_input_tokens_seen": 65717136,
      "step": 113190
    },
    {
      "epoch": 16.859547214775098,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0021975588681191105,
      "loss": 0.8072,
      "num_input_tokens_seen": 65719920,
      "step": 113195
    },
    {
      "epoch": 16.860291927316055,
      "grad_norm": 0.052734375,
      "learning_rate": 0.002196543013575196,
      "loss": 0.7916,
      "num_input_tokens_seen": 65722640,
      "step": 113200
    },
    {
      "epoch": 16.861036639857016,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.002195527375332677,
      "loss": 0.7832,
      "num_input_tokens_seen": 65725488,
      "step": 113205
    },
    {
      "epoch": 16.861781352397973,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.002194511953408706,
      "loss": 0.7765,
      "num_input_tokens_seen": 65728304,
      "step": 113210
    },
    {
      "epoch": 16.862526064938933,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0021934967478204424,
      "loss": 0.8043,
      "num_input_tokens_seen": 65731088,
      "step": 113215
    },
    {
      "epoch": 16.863270777479894,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.002192481758585033,
      "loss": 0.8107,
      "num_input_tokens_seen": 65734000,
      "step": 113220
    },
    {
      "epoch": 16.86401549002085,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.002191466985719632,
      "loss": 0.7711,
      "num_input_tokens_seen": 65736784,
      "step": 113225
    },
    {
      "epoch": 16.86476020256181,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0021904524292413762,
      "loss": 0.8025,
      "num_input_tokens_seen": 65739568,
      "step": 113230
    },
    {
      "epoch": 16.865504915102772,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0021894380891674062,
      "loss": 0.7924,
      "num_input_tokens_seen": 65742448,
      "step": 113235
    },
    {
      "epoch": 16.86624962764373,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0021884239655148604,
      "loss": 0.7719,
      "num_input_tokens_seen": 65745232,
      "step": 113240
    },
    {
      "epoch": 16.86699434018469,
      "grad_norm": 0.28515625,
      "learning_rate": 0.0021874100583008736,
      "loss": 0.8155,
      "num_input_tokens_seen": 65748080,
      "step": 113245
    },
    {
      "epoch": 16.867739052725646,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.00218639636754257,
      "loss": 0.8169,
      "num_input_tokens_seen": 65751280,
      "step": 113250
    },
    {
      "epoch": 16.868483765266607,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.002185382893257073,
      "loss": 0.8191,
      "num_input_tokens_seen": 65754608,
      "step": 113255
    },
    {
      "epoch": 16.869228477807567,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0021843696354615113,
      "loss": 0.7913,
      "num_input_tokens_seen": 65757616,
      "step": 113260
    },
    {
      "epoch": 16.869973190348524,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0021833565941729977,
      "loss": 0.8036,
      "num_input_tokens_seen": 65760912,
      "step": 113265
    },
    {
      "epoch": 16.870717902889485,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.002182343769408649,
      "loss": 0.79,
      "num_input_tokens_seen": 65764048,
      "step": 113270
    },
    {
      "epoch": 16.871462615430445,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0021813311611855765,
      "loss": 0.7757,
      "num_input_tokens_seen": 65766832,
      "step": 113275
    },
    {
      "epoch": 16.872207327971402,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0021803187695208824,
      "loss": 0.8456,
      "num_input_tokens_seen": 65769744,
      "step": 113280
    },
    {
      "epoch": 16.872952040512363,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.002179306594431676,
      "loss": 0.791,
      "num_input_tokens_seen": 65772496,
      "step": 113285
    },
    {
      "epoch": 16.87369675305332,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0021782946359350515,
      "loss": 0.7961,
      "num_input_tokens_seen": 65775536,
      "step": 113290
    },
    {
      "epoch": 16.87444146559428,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.002177282894048112,
      "loss": 0.8014,
      "num_input_tokens_seen": 65778512,
      "step": 113295
    },
    {
      "epoch": 16.87518617813524,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0021762713687879407,
      "loss": 0.8061,
      "num_input_tokens_seen": 65781360,
      "step": 113300
    },
    {
      "epoch": 16.875930890676198,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.002175260060171634,
      "loss": 0.8015,
      "num_input_tokens_seen": 65784464,
      "step": 113305
    },
    {
      "epoch": 16.87667560321716,
      "grad_norm": 0.056640625,
      "learning_rate": 0.002174248968216275,
      "loss": 0.7779,
      "num_input_tokens_seen": 65787408,
      "step": 113310
    },
    {
      "epoch": 16.87742031575812,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.002173238092938945,
      "loss": 0.7744,
      "num_input_tokens_seen": 65790096,
      "step": 113315
    },
    {
      "epoch": 16.878165028299076,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.00217222743435672,
      "loss": 0.7877,
      "num_input_tokens_seen": 65792912,
      "step": 113320
    },
    {
      "epoch": 16.878909740840037,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0021712169924866715,
      "loss": 0.8059,
      "num_input_tokens_seen": 65795696,
      "step": 113325
    },
    {
      "epoch": 16.879654453380994,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0021702067673458774,
      "loss": 0.7962,
      "num_input_tokens_seen": 65798608,
      "step": 113330
    },
    {
      "epoch": 16.880399165921954,
      "grad_norm": 0.06640625,
      "learning_rate": 0.002169196758951396,
      "loss": 0.8093,
      "num_input_tokens_seen": 65801488,
      "step": 113335
    },
    {
      "epoch": 16.881143878462915,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0021681869673202984,
      "loss": 0.7739,
      "num_input_tokens_seen": 65804176,
      "step": 113340
    },
    {
      "epoch": 16.88188859100387,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.002167177392469638,
      "loss": 0.7997,
      "num_input_tokens_seen": 65806896,
      "step": 113345
    },
    {
      "epoch": 16.882633303544832,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0021661680344164757,
      "loss": 0.8002,
      "num_input_tokens_seen": 65809840,
      "step": 113350
    },
    {
      "epoch": 16.883378016085793,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.0021651588931778597,
      "loss": 0.819,
      "num_input_tokens_seen": 65812720,
      "step": 113355
    },
    {
      "epoch": 16.88412272862675,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0021641499687708363,
      "loss": 0.7847,
      "num_input_tokens_seen": 65815792,
      "step": 113360
    },
    {
      "epoch": 16.88486744116771,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0021631412612124574,
      "loss": 0.8061,
      "num_input_tokens_seen": 65818704,
      "step": 113365
    },
    {
      "epoch": 16.885612153708667,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0021621327705197567,
      "loss": 0.8048,
      "num_input_tokens_seen": 65821680,
      "step": 113370
    },
    {
      "epoch": 16.886356866249628,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.002161124496709778,
      "loss": 0.7696,
      "num_input_tokens_seen": 65824368,
      "step": 113375
    },
    {
      "epoch": 16.88710157879059,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0021601164397995524,
      "loss": 0.7902,
      "num_input_tokens_seen": 65827472,
      "step": 113380
    },
    {
      "epoch": 16.887846291331545,
      "grad_norm": 0.044921875,
      "learning_rate": 0.00215910859980611,
      "loss": 0.8028,
      "num_input_tokens_seen": 65830544,
      "step": 113385
    },
    {
      "epoch": 16.888591003872506,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0021581009767464763,
      "loss": 0.7954,
      "num_input_tokens_seen": 65833680,
      "step": 113390
    },
    {
      "epoch": 16.889335716413463,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0021570935706376718,
      "loss": 0.8112,
      "num_input_tokens_seen": 65836528,
      "step": 113395
    },
    {
      "epoch": 16.890080428954423,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.002156086381496721,
      "loss": 0.8069,
      "num_input_tokens_seen": 65839600,
      "step": 113400
    },
    {
      "epoch": 16.890825141495384,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0021550794093406355,
      "loss": 0.7934,
      "num_input_tokens_seen": 65842672,
      "step": 113405
    },
    {
      "epoch": 16.89156985403634,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.00215407265418643,
      "loss": 0.8037,
      "num_input_tokens_seen": 65845488,
      "step": 113410
    },
    {
      "epoch": 16.8923145665773,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0021530661160511083,
      "loss": 0.794,
      "num_input_tokens_seen": 65848304,
      "step": 113415
    },
    {
      "epoch": 16.893059279118262,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0021520597949516796,
      "loss": 0.8093,
      "num_input_tokens_seen": 65851248,
      "step": 113420
    },
    {
      "epoch": 16.89380399165922,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0021510536909051395,
      "loss": 0.802,
      "num_input_tokens_seen": 65854096,
      "step": 113425
    },
    {
      "epoch": 16.89454870420018,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0021500478039284936,
      "loss": 0.7919,
      "num_input_tokens_seen": 65856816,
      "step": 113430
    },
    {
      "epoch": 16.895293416741136,
      "grad_norm": 0.052734375,
      "learning_rate": 0.002149042134038728,
      "loss": 0.8106,
      "num_input_tokens_seen": 65859536,
      "step": 113435
    },
    {
      "epoch": 16.896038129282097,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.002148036681252832,
      "loss": 0.8094,
      "num_input_tokens_seen": 65862256,
      "step": 113440
    },
    {
      "epoch": 16.896782841823057,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.002147031445587797,
      "loss": 0.7891,
      "num_input_tokens_seen": 65865072,
      "step": 113445
    },
    {
      "epoch": 16.897527554364014,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0021460264270606014,
      "loss": 0.7866,
      "num_input_tokens_seen": 65867792,
      "step": 113450
    },
    {
      "epoch": 16.898272266904975,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.002145021625688227,
      "loss": 0.7966,
      "num_input_tokens_seen": 65870608,
      "step": 113455
    },
    {
      "epoch": 16.899016979445936,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.002144017041487643,
      "loss": 0.8081,
      "num_input_tokens_seen": 65873584,
      "step": 113460
    },
    {
      "epoch": 16.899761691986892,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0021430126744758277,
      "loss": 0.7913,
      "num_input_tokens_seen": 65876784,
      "step": 113465
    },
    {
      "epoch": 16.900506404527853,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0021420085246697473,
      "loss": 0.8181,
      "num_input_tokens_seen": 65879664,
      "step": 113470
    },
    {
      "epoch": 16.90125111706881,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00214100459208636,
      "loss": 0.7996,
      "num_input_tokens_seen": 65882512,
      "step": 113475
    },
    {
      "epoch": 16.90199582960977,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.002140000876742634,
      "loss": 0.8272,
      "num_input_tokens_seen": 65885520,
      "step": 113480
    },
    {
      "epoch": 16.90274054215073,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.002138997378655521,
      "loss": 0.7677,
      "num_input_tokens_seen": 65888144,
      "step": 113485
    },
    {
      "epoch": 16.903485254691688,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0021379940978419787,
      "loss": 0.8489,
      "num_input_tokens_seen": 65890896,
      "step": 113490
    },
    {
      "epoch": 16.90422996723265,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.00213699103431895,
      "loss": 0.8068,
      "num_input_tokens_seen": 65894000,
      "step": 113495
    },
    {
      "epoch": 16.90497467977361,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.002135988188103387,
      "loss": 0.7786,
      "num_input_tokens_seen": 65896816,
      "step": 113500
    },
    {
      "epoch": 16.905719392314566,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0021349855592122306,
      "loss": 0.7848,
      "num_input_tokens_seen": 65899568,
      "step": 113505
    },
    {
      "epoch": 16.906464104855527,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0021339831476624178,
      "loss": 0.7939,
      "num_input_tokens_seen": 65902384,
      "step": 113510
    },
    {
      "epoch": 16.907208817396484,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.002132980953470879,
      "loss": 0.7896,
      "num_input_tokens_seen": 65905264,
      "step": 113515
    },
    {
      "epoch": 16.907953529937444,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0021319789766545538,
      "loss": 0.7909,
      "num_input_tokens_seen": 65908144,
      "step": 113520
    },
    {
      "epoch": 16.908698242478405,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0021309772172303644,
      "loss": 0.791,
      "num_input_tokens_seen": 65911056,
      "step": 113525
    },
    {
      "epoch": 16.90944295501936,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.002129975675215232,
      "loss": 0.7789,
      "num_input_tokens_seen": 65913936,
      "step": 113530
    },
    {
      "epoch": 16.910187667560322,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.0021289743506260838,
      "loss": 0.7972,
      "num_input_tokens_seen": 65917424,
      "step": 113535
    },
    {
      "epoch": 16.91093238010128,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.002127973243479829,
      "loss": 0.8,
      "num_input_tokens_seen": 65920272,
      "step": 113540
    },
    {
      "epoch": 16.91167709264224,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0021269723537933882,
      "loss": 0.789,
      "num_input_tokens_seen": 65923248,
      "step": 113545
    },
    {
      "epoch": 16.9124218051832,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0021259716815836636,
      "loss": 0.7984,
      "num_input_tokens_seen": 65926000,
      "step": 113550
    },
    {
      "epoch": 16.913166517724157,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0021249712268675606,
      "loss": 0.8134,
      "num_input_tokens_seen": 65929744,
      "step": 113555
    },
    {
      "epoch": 16.913911230265118,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.002123970989661987,
      "loss": 0.818,
      "num_input_tokens_seen": 65932432,
      "step": 113560
    },
    {
      "epoch": 16.91465594280608,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0021229709699838333,
      "loss": 0.8181,
      "num_input_tokens_seen": 65935152,
      "step": 113565
    },
    {
      "epoch": 16.915400655347035,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.002121971167850002,
      "loss": 0.7865,
      "num_input_tokens_seen": 65937776,
      "step": 113570
    },
    {
      "epoch": 16.916145367887996,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.002120971583277378,
      "loss": 0.8143,
      "num_input_tokens_seen": 65940336,
      "step": 113575
    },
    {
      "epoch": 16.916890080428953,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0021199722162828474,
      "loss": 0.7909,
      "num_input_tokens_seen": 65943184,
      "step": 113580
    },
    {
      "epoch": 16.917634792969913,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0021189730668832946,
      "loss": 0.8248,
      "num_input_tokens_seen": 65946032,
      "step": 113585
    },
    {
      "epoch": 16.918379505510874,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.002117974135095603,
      "loss": 0.8017,
      "num_input_tokens_seen": 65948816,
      "step": 113590
    },
    {
      "epoch": 16.91912421805183,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0021169754209366468,
      "loss": 0.8014,
      "num_input_tokens_seen": 65951792,
      "step": 113595
    },
    {
      "epoch": 16.91986893059279,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.002115976924423292,
      "loss": 0.8281,
      "num_input_tokens_seen": 65954864,
      "step": 113600
    },
    {
      "epoch": 16.920613643133752,
      "grad_norm": 0.07421875,
      "learning_rate": 0.002114978645572417,
      "loss": 0.7963,
      "num_input_tokens_seen": 65957552,
      "step": 113605
    },
    {
      "epoch": 16.92135835567471,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0021139805844008768,
      "loss": 0.7927,
      "num_input_tokens_seen": 65960240,
      "step": 113610
    },
    {
      "epoch": 16.92210306821567,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0021129827409255425,
      "loss": 0.8,
      "num_input_tokens_seen": 65963088,
      "step": 113615
    },
    {
      "epoch": 16.922847780756626,
      "grad_norm": 0.078125,
      "learning_rate": 0.002111985115163264,
      "loss": 0.7927,
      "num_input_tokens_seen": 65966000,
      "step": 113620
    },
    {
      "epoch": 16.923592493297587,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.002110987707130901,
      "loss": 0.8237,
      "num_input_tokens_seen": 65969072,
      "step": 113625
    },
    {
      "epoch": 16.924337205838548,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.002109990516845301,
      "loss": 0.8097,
      "num_input_tokens_seen": 65972080,
      "step": 113630
    },
    {
      "epoch": 16.925081918379504,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0021089935443233078,
      "loss": 0.8004,
      "num_input_tokens_seen": 65974928,
      "step": 113635
    },
    {
      "epoch": 16.925826630920465,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0021079967895817698,
      "loss": 0.8206,
      "num_input_tokens_seen": 65977808,
      "step": 113640
    },
    {
      "epoch": 16.926571343461426,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.002107000252637522,
      "loss": 0.7852,
      "num_input_tokens_seen": 65980528,
      "step": 113645
    },
    {
      "epoch": 16.927316056002383,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0021060039335074024,
      "loss": 0.8126,
      "num_input_tokens_seen": 65983056,
      "step": 113650
    },
    {
      "epoch": 16.928060768543343,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0021050078322082384,
      "loss": 0.8082,
      "num_input_tokens_seen": 65986160,
      "step": 113655
    },
    {
      "epoch": 16.9288054810843,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0021040119487568637,
      "loss": 0.7872,
      "num_input_tokens_seen": 65989488,
      "step": 113660
    },
    {
      "epoch": 16.92955019362526,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0021030162831700973,
      "loss": 0.7941,
      "num_input_tokens_seen": 65992496,
      "step": 113665
    },
    {
      "epoch": 16.93029490616622,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.002102020835464765,
      "loss": 0.7929,
      "num_input_tokens_seen": 65995664,
      "step": 113670
    },
    {
      "epoch": 16.931039618707178,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0021010256056576827,
      "loss": 0.8203,
      "num_input_tokens_seen": 65998352,
      "step": 113675
    },
    {
      "epoch": 16.93178433124814,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.002100030593765659,
      "loss": 0.7853,
      "num_input_tokens_seen": 66001168,
      "step": 113680
    },
    {
      "epoch": 16.932529043789096,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.002099035799805512,
      "loss": 0.7717,
      "num_input_tokens_seen": 66004016,
      "step": 113685
    },
    {
      "epoch": 16.933273756330056,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0020980412237940386,
      "loss": 0.7916,
      "num_input_tokens_seen": 66006704,
      "step": 113690
    },
    {
      "epoch": 16.934018468871017,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0020970468657480496,
      "loss": 0.7794,
      "num_input_tokens_seen": 66009968,
      "step": 113695
    },
    {
      "epoch": 16.934763181411974,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0020960527256843374,
      "loss": 0.7823,
      "num_input_tokens_seen": 66013008,
      "step": 113700
    },
    {
      "epoch": 16.935507893952934,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.002095058803619701,
      "loss": 0.796,
      "num_input_tokens_seen": 66015888,
      "step": 113705
    },
    {
      "epoch": 16.936252606493895,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0020940650995709314,
      "loss": 0.7925,
      "num_input_tokens_seen": 66018704,
      "step": 113710
    },
    {
      "epoch": 16.93699731903485,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0020930716135548156,
      "loss": 0.8119,
      "num_input_tokens_seen": 66021712,
      "step": 113715
    },
    {
      "epoch": 16.937742031575812,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0020920783455881348,
      "loss": 0.7966,
      "num_input_tokens_seen": 66024272,
      "step": 113720
    },
    {
      "epoch": 16.93848674411677,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.002091085295687669,
      "loss": 0.7838,
      "num_input_tokens_seen": 66027056,
      "step": 113725
    },
    {
      "epoch": 16.93923145665773,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0020900924638702,
      "loss": 0.7977,
      "num_input_tokens_seen": 66029680,
      "step": 113730
    },
    {
      "epoch": 16.93997616919869,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.002089099850152496,
      "loss": 0.7848,
      "num_input_tokens_seen": 66032624,
      "step": 113735
    },
    {
      "epoch": 16.940720881739647,
      "grad_norm": 0.056640625,
      "learning_rate": 0.00208810745455133,
      "loss": 0.7748,
      "num_input_tokens_seen": 66035600,
      "step": 113740
    },
    {
      "epoch": 16.941465594280608,
      "grad_norm": 0.060546875,
      "learning_rate": 0.002087115277083462,
      "loss": 0.8035,
      "num_input_tokens_seen": 66038544,
      "step": 113745
    },
    {
      "epoch": 16.94221030682157,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0020861233177656606,
      "loss": 0.8141,
      "num_input_tokens_seen": 66041456,
      "step": 113750
    },
    {
      "epoch": 16.942955019362525,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0020851315766146823,
      "loss": 0.8175,
      "num_input_tokens_seen": 66044432,
      "step": 113755
    },
    {
      "epoch": 16.943699731903486,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.002084140053647275,
      "loss": 0.8037,
      "num_input_tokens_seen": 66047440,
      "step": 113760
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0020831487488801987,
      "loss": 0.7916,
      "num_input_tokens_seen": 66050000,
      "step": 113765
    },
    {
      "epoch": 16.945189156985403,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.002082157662330192,
      "loss": 0.7867,
      "num_input_tokens_seen": 66052944,
      "step": 113770
    },
    {
      "epoch": 16.945933869526364,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.002081166794014006,
      "loss": 0.7842,
      "num_input_tokens_seen": 66056048,
      "step": 113775
    },
    {
      "epoch": 16.94667858206732,
      "grad_norm": 0.2255859375,
      "learning_rate": 0.002080176143948376,
      "loss": 0.8011,
      "num_input_tokens_seen": 66058992,
      "step": 113780
    },
    {
      "epoch": 16.94742329460828,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0020791857121500406,
      "loss": 0.8037,
      "num_input_tokens_seen": 66061872,
      "step": 113785
    },
    {
      "epoch": 16.948168007149242,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.002078195498635729,
      "loss": 0.7972,
      "num_input_tokens_seen": 66065040,
      "step": 113790
    },
    {
      "epoch": 16.9489127196902,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.00207720550342217,
      "loss": 0.7943,
      "num_input_tokens_seen": 66067856,
      "step": 113795
    },
    {
      "epoch": 16.94965743223116,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.002076215726526092,
      "loss": 0.7889,
      "num_input_tokens_seen": 66070512,
      "step": 113800
    },
    {
      "epoch": 16.950402144772116,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0020752261679642097,
      "loss": 0.8017,
      "num_input_tokens_seen": 66073424,
      "step": 113805
    },
    {
      "epoch": 16.951146857313077,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0020742368277532498,
      "loss": 0.7801,
      "num_input_tokens_seen": 66076528,
      "step": 113810
    },
    {
      "epoch": 16.951891569854038,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0020732477059099185,
      "loss": 0.8047,
      "num_input_tokens_seen": 66079760,
      "step": 113815
    },
    {
      "epoch": 16.952636282394995,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0020722588024509312,
      "loss": 0.8244,
      "num_input_tokens_seen": 66082896,
      "step": 113820
    },
    {
      "epoch": 16.953380994935955,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00207127011739299,
      "loss": 0.7763,
      "num_input_tokens_seen": 66086000,
      "step": 113825
    },
    {
      "epoch": 16.954125707476916,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.002070281650752804,
      "loss": 0.8083,
      "num_input_tokens_seen": 66088976,
      "step": 113830
    },
    {
      "epoch": 16.954870420017873,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0020692934025470667,
      "loss": 0.8016,
      "num_input_tokens_seen": 66091856,
      "step": 113835
    },
    {
      "epoch": 16.955615132558833,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0020683053727924747,
      "loss": 0.7958,
      "num_input_tokens_seen": 66094704,
      "step": 113840
    },
    {
      "epoch": 16.95635984509979,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0020673175615057186,
      "loss": 0.8084,
      "num_input_tokens_seen": 66097840,
      "step": 113845
    },
    {
      "epoch": 16.95710455764075,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.002066329968703491,
      "loss": 0.801,
      "num_input_tokens_seen": 66100560,
      "step": 113850
    },
    {
      "epoch": 16.95784927018171,
      "grad_norm": 0.0546875,
      "learning_rate": 0.002065342594402473,
      "loss": 0.7907,
      "num_input_tokens_seen": 66103568,
      "step": 113855
    },
    {
      "epoch": 16.958593982722668,
      "grad_norm": 0.09375,
      "learning_rate": 0.002064355438619343,
      "loss": 0.8038,
      "num_input_tokens_seen": 66106672,
      "step": 113860
    },
    {
      "epoch": 16.95933869526363,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0020633685013707838,
      "loss": 0.804,
      "num_input_tokens_seen": 66109904,
      "step": 113865
    },
    {
      "epoch": 16.96008340780459,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0020623817826734615,
      "loss": 0.7981,
      "num_input_tokens_seen": 66112656,
      "step": 113870
    },
    {
      "epoch": 16.960828120345546,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.002061395282544053,
      "loss": 0.7951,
      "num_input_tokens_seen": 66115504,
      "step": 113875
    },
    {
      "epoch": 16.961572832886507,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0020604090009992216,
      "loss": 0.7878,
      "num_input_tokens_seen": 66118608,
      "step": 113880
    },
    {
      "epoch": 16.962317545427464,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0020594229380556254,
      "loss": 0.8088,
      "num_input_tokens_seen": 66121648,
      "step": 113885
    },
    {
      "epoch": 16.963062257968424,
      "grad_norm": 0.06640625,
      "learning_rate": 0.00205843709372993,
      "loss": 0.8426,
      "num_input_tokens_seen": 66124432,
      "step": 113890
    },
    {
      "epoch": 16.963806970509385,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.002057451468038783,
      "loss": 0.7945,
      "num_input_tokens_seen": 66127184,
      "step": 113895
    },
    {
      "epoch": 16.964551683050342,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.002056466060998844,
      "loss": 0.7999,
      "num_input_tokens_seen": 66130064,
      "step": 113900
    },
    {
      "epoch": 16.965296395591302,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.002055480872626753,
      "loss": 0.7797,
      "num_input_tokens_seen": 66132848,
      "step": 113905
    },
    {
      "epoch": 16.96604110813226,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.002054495902939158,
      "loss": 0.7975,
      "num_input_tokens_seen": 66135856,
      "step": 113910
    },
    {
      "epoch": 16.96678582067322,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0020535111519526983,
      "loss": 0.7803,
      "num_input_tokens_seen": 66138608,
      "step": 113915
    },
    {
      "epoch": 16.96753053321418,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.002052526619684004,
      "loss": 0.8005,
      "num_input_tokens_seen": 66141232,
      "step": 113920
    },
    {
      "epoch": 16.968275245755137,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.002051542306149719,
      "loss": 0.8002,
      "num_input_tokens_seen": 66144208,
      "step": 113925
    },
    {
      "epoch": 16.969019958296098,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.002050558211366461,
      "loss": 0.767,
      "num_input_tokens_seen": 66146896,
      "step": 113930
    },
    {
      "epoch": 16.96976467083706,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0020495743353508653,
      "loss": 0.8021,
      "num_input_tokens_seen": 66150000,
      "step": 113935
    },
    {
      "epoch": 16.970509383378015,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0020485906781195456,
      "loss": 0.7945,
      "num_input_tokens_seen": 66152880,
      "step": 113940
    },
    {
      "epoch": 16.971254095918976,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0020476072396891265,
      "loss": 0.7694,
      "num_input_tokens_seen": 66155504,
      "step": 113945
    },
    {
      "epoch": 16.971998808459933,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0020466240200762162,
      "loss": 0.8008,
      "num_input_tokens_seen": 66158896,
      "step": 113950
    },
    {
      "epoch": 16.972743521000893,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0020456410192974304,
      "loss": 0.8045,
      "num_input_tokens_seen": 66161968,
      "step": 113955
    },
    {
      "epoch": 16.973488233541854,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.002044658237369372,
      "loss": 0.7982,
      "num_input_tokens_seen": 66164592,
      "step": 113960
    },
    {
      "epoch": 16.97423294608281,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.002043675674308645,
      "loss": 0.7759,
      "num_input_tokens_seen": 66167152,
      "step": 113965
    },
    {
      "epoch": 16.97497765862377,
      "grad_norm": 0.060546875,
      "learning_rate": 0.00204269333013185,
      "loss": 0.8036,
      "num_input_tokens_seen": 66169968,
      "step": 113970
    },
    {
      "epoch": 16.975722371164732,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.002041711204855583,
      "loss": 0.7814,
      "num_input_tokens_seen": 66173136,
      "step": 113975
    },
    {
      "epoch": 16.97646708370569,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0020407292984964353,
      "loss": 0.8332,
      "num_input_tokens_seen": 66176368,
      "step": 113980
    },
    {
      "epoch": 16.97721179624665,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0020397476110709904,
      "loss": 0.797,
      "num_input_tokens_seen": 66179344,
      "step": 113985
    },
    {
      "epoch": 16.977956508787607,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.002038766142595842,
      "loss": 0.806,
      "num_input_tokens_seen": 66182512,
      "step": 113990
    },
    {
      "epoch": 16.978701221328567,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0020377848930875665,
      "loss": 0.8022,
      "num_input_tokens_seen": 66185296,
      "step": 113995
    },
    {
      "epoch": 16.979445933869528,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.002036803862562737,
      "loss": 0.779,
      "num_input_tokens_seen": 66188240,
      "step": 114000
    },
    {
      "epoch": 16.980190646410485,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0020358230510379343,
      "loss": 0.8235,
      "num_input_tokens_seen": 66190960,
      "step": 114005
    },
    {
      "epoch": 16.980935358951445,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.002034842458529723,
      "loss": 0.7966,
      "num_input_tokens_seen": 66193648,
      "step": 114010
    },
    {
      "epoch": 16.981680071492406,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0020338620850546734,
      "loss": 0.7788,
      "num_input_tokens_seen": 66196656,
      "step": 114015
    },
    {
      "epoch": 16.982424784033363,
      "grad_norm": 0.083984375,
      "learning_rate": 0.002032881930629342,
      "loss": 0.772,
      "num_input_tokens_seen": 66199568,
      "step": 114020
    },
    {
      "epoch": 16.983169496574323,
      "grad_norm": 0.06640625,
      "learning_rate": 0.002031901995270294,
      "loss": 0.7821,
      "num_input_tokens_seen": 66202384,
      "step": 114025
    },
    {
      "epoch": 16.98391420911528,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.00203092227899408,
      "loss": 0.8181,
      "num_input_tokens_seen": 66205104,
      "step": 114030
    },
    {
      "epoch": 16.98465892165624,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0020299427818172545,
      "loss": 0.7855,
      "num_input_tokens_seen": 66207824,
      "step": 114035
    },
    {
      "epoch": 16.9854036341972,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0020289635037563656,
      "loss": 0.8073,
      "num_input_tokens_seen": 66210512,
      "step": 114040
    },
    {
      "epoch": 16.986148346738158,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0020279844448279527,
      "loss": 0.7727,
      "num_input_tokens_seen": 66213552,
      "step": 114045
    },
    {
      "epoch": 16.98689305927912,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00202700560504856,
      "loss": 0.8012,
      "num_input_tokens_seen": 66216304,
      "step": 114050
    },
    {
      "epoch": 16.987637771820076,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0020260269844347184,
      "loss": 0.7921,
      "num_input_tokens_seen": 66219312,
      "step": 114055
    },
    {
      "epoch": 16.988382484361036,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.002025048583002969,
      "loss": 0.7944,
      "num_input_tokens_seen": 66222320,
      "step": 114060
    },
    {
      "epoch": 16.989127196901997,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0020240704007698325,
      "loss": 0.7719,
      "num_input_tokens_seen": 66225488,
      "step": 114065
    },
    {
      "epoch": 16.989871909442954,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0020230924377518415,
      "loss": 0.7866,
      "num_input_tokens_seen": 66228560,
      "step": 114070
    },
    {
      "epoch": 16.990616621983914,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0020221146939655155,
      "loss": 0.7955,
      "num_input_tokens_seen": 66232112,
      "step": 114075
    },
    {
      "epoch": 16.991361334524875,
      "grad_norm": 0.080078125,
      "learning_rate": 0.002021137169427369,
      "loss": 0.7871,
      "num_input_tokens_seen": 66234896,
      "step": 114080
    },
    {
      "epoch": 16.992106047065832,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0020201598641539204,
      "loss": 0.7902,
      "num_input_tokens_seen": 66237488,
      "step": 114085
    },
    {
      "epoch": 16.992850759606792,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0020191827781616766,
      "loss": 0.7846,
      "num_input_tokens_seen": 66240496,
      "step": 114090
    },
    {
      "epoch": 16.99359547214775,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0020182059114671495,
      "loss": 0.8211,
      "num_input_tokens_seen": 66243728,
      "step": 114095
    },
    {
      "epoch": 16.99434018468871,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.002017229264086836,
      "loss": 0.8125,
      "num_input_tokens_seen": 66246896,
      "step": 114100
    },
    {
      "epoch": 16.99508489722967,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0020162528360372425,
      "loss": 0.781,
      "num_input_tokens_seen": 66250000,
      "step": 114105
    },
    {
      "epoch": 16.995829609770627,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0020152766273348614,
      "loss": 0.8166,
      "num_input_tokens_seen": 66252816,
      "step": 114110
    },
    {
      "epoch": 16.996574322311588,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.002014300637996184,
      "loss": 0.8052,
      "num_input_tokens_seen": 66255696,
      "step": 114115
    },
    {
      "epoch": 16.99731903485255,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0020133248680377007,
      "loss": 0.7966,
      "num_input_tokens_seen": 66258736,
      "step": 114120
    },
    {
      "epoch": 16.998063747393505,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.00201234931747589,
      "loss": 0.7915,
      "num_input_tokens_seen": 66261424,
      "step": 114125
    },
    {
      "epoch": 16.998808459934466,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.00201137398632724,
      "loss": 0.7969,
      "num_input_tokens_seen": 66264176,
      "step": 114130
    },
    {
      "epoch": 16.999553172475423,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0020103988746082244,
      "loss": 0.7752,
      "num_input_tokens_seen": 66267120,
      "step": 114135
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.8002009987831116,
      "eval_runtime": 70.4795,
      "eval_samples_per_second": 42.339,
      "eval_steps_per_second": 10.585,
      "num_input_tokens_seen": 66268336,
      "step": 114138
    },
    {
      "epoch": 17.000297885016384,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0020094239823353205,
      "loss": 0.7933,
      "num_input_tokens_seen": 66269776,
      "step": 114140
    },
    {
      "epoch": 17.001042597557344,
      "grad_norm": 0.04296875,
      "learning_rate": 0.002008449309524991,
      "loss": 0.7942,
      "num_input_tokens_seen": 66272496,
      "step": 114145
    },
    {
      "epoch": 17.0017873100983,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00200747485619371,
      "loss": 0.7792,
      "num_input_tokens_seen": 66275248,
      "step": 114150
    },
    {
      "epoch": 17.00253202263926,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0020065006223579374,
      "loss": 0.8194,
      "num_input_tokens_seen": 66277776,
      "step": 114155
    },
    {
      "epoch": 17.003276735180222,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0020055266080341266,
      "loss": 0.8288,
      "num_input_tokens_seen": 66280656,
      "step": 114160
    },
    {
      "epoch": 17.00402144772118,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0020045528132387395,
      "loss": 0.8041,
      "num_input_tokens_seen": 66283888,
      "step": 114165
    },
    {
      "epoch": 17.00476616026214,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0020035792379882247,
      "loss": 0.7977,
      "num_input_tokens_seen": 66286864,
      "step": 114170
    },
    {
      "epoch": 17.005510872803097,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.002002605882299028,
      "loss": 0.7872,
      "num_input_tokens_seen": 66289808,
      "step": 114175
    },
    {
      "epoch": 17.006255585344057,
      "grad_norm": 0.060546875,
      "learning_rate": 0.0020016327461875967,
      "loss": 0.7784,
      "num_input_tokens_seen": 66292848,
      "step": 114180
    },
    {
      "epoch": 17.007000297885018,
      "grad_norm": 0.056640625,
      "learning_rate": 0.00200065982967037,
      "loss": 0.7896,
      "num_input_tokens_seen": 66295760,
      "step": 114185
    },
    {
      "epoch": 17.007745010425975,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.001999687132763779,
      "loss": 0.8064,
      "num_input_tokens_seen": 66298768,
      "step": 114190
    },
    {
      "epoch": 17.008489722966935,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0019987146554842653,
      "loss": 0.8092,
      "num_input_tokens_seen": 66301680,
      "step": 114195
    },
    {
      "epoch": 17.009234435507896,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0019977423978482533,
      "loss": 0.7695,
      "num_input_tokens_seen": 66304688,
      "step": 114200
    },
    {
      "epoch": 17.009979148048853,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0019967703598721665,
      "loss": 0.7921,
      "num_input_tokens_seen": 66307536,
      "step": 114205
    },
    {
      "epoch": 17.010723860589813,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0019957985415724297,
      "loss": 0.7871,
      "num_input_tokens_seen": 66310352,
      "step": 114210
    },
    {
      "epoch": 17.01146857313077,
      "grad_norm": 0.05859375,
      "learning_rate": 0.001994826942965459,
      "loss": 0.7809,
      "num_input_tokens_seen": 66313360,
      "step": 114215
    },
    {
      "epoch": 17.01221328567173,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0019938555640676707,
      "loss": 0.7971,
      "num_input_tokens_seen": 66316176,
      "step": 114220
    },
    {
      "epoch": 17.01295799821269,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0019928844048954703,
      "loss": 0.7983,
      "num_input_tokens_seen": 66318928,
      "step": 114225
    },
    {
      "epoch": 17.01370271075365,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0019919134654652723,
      "loss": 0.7883,
      "num_input_tokens_seen": 66321680,
      "step": 114230
    },
    {
      "epoch": 17.01444742329461,
      "grad_norm": 0.0235595703125,
      "learning_rate": 0.0019909427457934746,
      "loss": 0.7925,
      "num_input_tokens_seen": 66324336,
      "step": 114235
    },
    {
      "epoch": 17.015192135835566,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0019899722458964785,
      "loss": 0.797,
      "num_input_tokens_seen": 66327056,
      "step": 114240
    },
    {
      "epoch": 17.015936848376526,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0019890019657906777,
      "loss": 0.7949,
      "num_input_tokens_seen": 66330128,
      "step": 114245
    },
    {
      "epoch": 17.016681560917487,
      "grad_norm": 0.04296875,
      "learning_rate": 0.001988031905492462,
      "loss": 0.7846,
      "num_input_tokens_seen": 66333136,
      "step": 114250
    },
    {
      "epoch": 17.017426273458444,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0019870620650182256,
      "loss": 0.7726,
      "num_input_tokens_seen": 66336016,
      "step": 114255
    },
    {
      "epoch": 17.018170985999404,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0019860924443843454,
      "loss": 0.7971,
      "num_input_tokens_seen": 66339120,
      "step": 114260
    },
    {
      "epoch": 17.018915698540365,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.001985123043607211,
      "loss": 0.7785,
      "num_input_tokens_seen": 66341936,
      "step": 114265
    },
    {
      "epoch": 17.019660411081322,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.001984153862703193,
      "loss": 0.7965,
      "num_input_tokens_seen": 66344944,
      "step": 114270
    },
    {
      "epoch": 17.020405123622282,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.001983184901688668,
      "loss": 0.7958,
      "num_input_tokens_seen": 66347696,
      "step": 114275
    },
    {
      "epoch": 17.02114983616324,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.001982216160580005,
      "loss": 0.8007,
      "num_input_tokens_seen": 66350736,
      "step": 114280
    },
    {
      "epoch": 17.0218945487042,
      "grad_norm": 0.087890625,
      "learning_rate": 0.001981247639393566,
      "loss": 0.7865,
      "num_input_tokens_seen": 66353776,
      "step": 114285
    },
    {
      "epoch": 17.02263926124516,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0019802793381457195,
      "loss": 0.8071,
      "num_input_tokens_seen": 66356592,
      "step": 114290
    },
    {
      "epoch": 17.023383973786117,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0019793112568528186,
      "loss": 0.7852,
      "num_input_tokens_seen": 66359344,
      "step": 114295
    },
    {
      "epoch": 17.024128686327078,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.001978343395531224,
      "loss": 0.7765,
      "num_input_tokens_seen": 66362096,
      "step": 114300
    },
    {
      "epoch": 17.02487339886804,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0019773757541972823,
      "loss": 0.7926,
      "num_input_tokens_seen": 66364944,
      "step": 114305
    },
    {
      "epoch": 17.025618111408996,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0019764083328673407,
      "loss": 0.839,
      "num_input_tokens_seen": 66367696,
      "step": 114310
    },
    {
      "epoch": 17.026362823949956,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.001975441131557745,
      "loss": 0.7788,
      "num_input_tokens_seen": 66370832,
      "step": 114315
    },
    {
      "epoch": 17.027107536490913,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0019744741502848305,
      "loss": 0.8149,
      "num_input_tokens_seen": 66373808,
      "step": 114320
    },
    {
      "epoch": 17.027852249031874,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.001973507389064939,
      "loss": 0.7817,
      "num_input_tokens_seen": 66376560,
      "step": 114325
    },
    {
      "epoch": 17.028596961572834,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0019725408479143985,
      "loss": 0.7626,
      "num_input_tokens_seen": 66379312,
      "step": 114330
    },
    {
      "epoch": 17.02934167411379,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0019715745268495425,
      "loss": 0.8002,
      "num_input_tokens_seen": 66382224,
      "step": 114335
    },
    {
      "epoch": 17.03008638665475,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.00197060842588669,
      "loss": 0.8153,
      "num_input_tokens_seen": 66385200,
      "step": 114340
    },
    {
      "epoch": 17.030831099195712,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0019696425450421696,
      "loss": 0.8174,
      "num_input_tokens_seen": 66388272,
      "step": 114345
    },
    {
      "epoch": 17.03157581173667,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.001968676884332291,
      "loss": 0.7913,
      "num_input_tokens_seen": 66391216,
      "step": 114350
    },
    {
      "epoch": 17.03232052427763,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.001967711443773375,
      "loss": 0.8022,
      "num_input_tokens_seen": 66394320,
      "step": 114355
    },
    {
      "epoch": 17.033065236818587,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.001966746223381728,
      "loss": 0.8034,
      "num_input_tokens_seen": 66397264,
      "step": 114360
    },
    {
      "epoch": 17.033809949359547,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0019657812231736564,
      "loss": 0.7838,
      "num_input_tokens_seen": 66400208,
      "step": 114365
    },
    {
      "epoch": 17.034554661900508,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0019648164431654647,
      "loss": 0.8097,
      "num_input_tokens_seen": 66403056,
      "step": 114370
    },
    {
      "epoch": 17.035299374441465,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.001963851883373453,
      "loss": 0.7951,
      "num_input_tokens_seen": 66405936,
      "step": 114375
    },
    {
      "epoch": 17.036044086982425,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.001962887543813912,
      "loss": 0.7833,
      "num_input_tokens_seen": 66408912,
      "step": 114380
    },
    {
      "epoch": 17.036788799523382,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0019619234245031336,
      "loss": 0.7833,
      "num_input_tokens_seen": 66411856,
      "step": 114385
    },
    {
      "epoch": 17.037533512064343,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0019609595254574115,
      "loss": 0.7999,
      "num_input_tokens_seen": 66414576,
      "step": 114390
    },
    {
      "epoch": 17.038278224605303,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.001959995846693024,
      "loss": 0.7706,
      "num_input_tokens_seen": 66417328,
      "step": 114395
    },
    {
      "epoch": 17.03902293714626,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.001959032388226252,
      "loss": 0.7963,
      "num_input_tokens_seen": 66420048,
      "step": 114400
    },
    {
      "epoch": 17.03976764968722,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0019580691500733767,
      "loss": 0.7802,
      "num_input_tokens_seen": 66422672,
      "step": 114405
    },
    {
      "epoch": 17.04051236222818,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0019571061322506637,
      "loss": 0.7678,
      "num_input_tokens_seen": 66425520,
      "step": 114410
    },
    {
      "epoch": 17.04125707476914,
      "grad_norm": 0.0625,
      "learning_rate": 0.001956143334774389,
      "loss": 0.8168,
      "num_input_tokens_seen": 66428400,
      "step": 114415
    },
    {
      "epoch": 17.0420017873101,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.001955180757660814,
      "loss": 0.8018,
      "num_input_tokens_seen": 66431760,
      "step": 114420
    },
    {
      "epoch": 17.042746499851056,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0019542184009262034,
      "loss": 0.8045,
      "num_input_tokens_seen": 66434992,
      "step": 114425
    },
    {
      "epoch": 17.043491212392016,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.001953256264586812,
      "loss": 0.761,
      "num_input_tokens_seen": 66437904,
      "step": 114430
    },
    {
      "epoch": 17.044235924932977,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0019522943486588982,
      "loss": 0.8035,
      "num_input_tokens_seen": 66440688,
      "step": 114435
    },
    {
      "epoch": 17.044980637473934,
      "grad_norm": 0.068359375,
      "learning_rate": 0.001951332653158712,
      "loss": 0.7734,
      "num_input_tokens_seen": 66443696,
      "step": 114440
    },
    {
      "epoch": 17.045725350014894,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0019503711781024974,
      "loss": 0.8035,
      "num_input_tokens_seen": 66446256,
      "step": 114445
    },
    {
      "epoch": 17.046470062555855,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0019494099235064977,
      "loss": 0.7975,
      "num_input_tokens_seen": 66449552,
      "step": 114450
    },
    {
      "epoch": 17.047214775096812,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0019484488893869516,
      "loss": 0.7951,
      "num_input_tokens_seen": 66452688,
      "step": 114455
    },
    {
      "epoch": 17.047959487637772,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.001947488075760098,
      "loss": 0.8009,
      "num_input_tokens_seen": 66455696,
      "step": 114460
    },
    {
      "epoch": 17.04870420017873,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.0019465274826421658,
      "loss": 0.8014,
      "num_input_tokens_seen": 66458416,
      "step": 114465
    },
    {
      "epoch": 17.04944891271969,
      "grad_norm": 0.0546875,
      "learning_rate": 0.001945567110049388,
      "loss": 0.8101,
      "num_input_tokens_seen": 66461328,
      "step": 114470
    },
    {
      "epoch": 17.05019362526065,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0019446069579979824,
      "loss": 0.7941,
      "num_input_tokens_seen": 66464112,
      "step": 114475
    },
    {
      "epoch": 17.050938337801608,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0019436470265041755,
      "loss": 0.8175,
      "num_input_tokens_seen": 66467088,
      "step": 114480
    },
    {
      "epoch": 17.051683050342568,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0019426873155841845,
      "loss": 0.803,
      "num_input_tokens_seen": 66470160,
      "step": 114485
    },
    {
      "epoch": 17.05242776288353,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0019417278252542147,
      "loss": 0.7946,
      "num_input_tokens_seen": 66473104,
      "step": 114490
    },
    {
      "epoch": 17.053172475424486,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0019407685555304865,
      "loss": 0.7695,
      "num_input_tokens_seen": 66476208,
      "step": 114495
    },
    {
      "epoch": 17.053917187965446,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0019398095064292002,
      "loss": 0.7985,
      "num_input_tokens_seen": 66479344,
      "step": 114500
    },
    {
      "epoch": 17.054661900506403,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.0019388506779665548,
      "loss": 0.7926,
      "num_input_tokens_seen": 66482256,
      "step": 114505
    },
    {
      "epoch": 17.055406613047364,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0019378920701587554,
      "loss": 0.8035,
      "num_input_tokens_seen": 66485616,
      "step": 114510
    },
    {
      "epoch": 17.056151325588324,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.001936933683021994,
      "loss": 0.7888,
      "num_input_tokens_seen": 66488720,
      "step": 114515
    },
    {
      "epoch": 17.05689603812928,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.001935975516572463,
      "loss": 0.7773,
      "num_input_tokens_seen": 66491568,
      "step": 114520
    },
    {
      "epoch": 17.05764075067024,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0019350175708263422,
      "loss": 0.7882,
      "num_input_tokens_seen": 66494416,
      "step": 114525
    },
    {
      "epoch": 17.058385463211202,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.001934059845799826,
      "loss": 0.7955,
      "num_input_tokens_seen": 66497520,
      "step": 114530
    },
    {
      "epoch": 17.05913017575216,
      "grad_norm": 0.4140625,
      "learning_rate": 0.0019331023415090842,
      "loss": 0.8048,
      "num_input_tokens_seen": 66500592,
      "step": 114535
    },
    {
      "epoch": 17.05987488829312,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0019321450579703025,
      "loss": 0.7825,
      "num_input_tokens_seen": 66503760,
      "step": 114540
    },
    {
      "epoch": 17.060619600834077,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0019311879951996447,
      "loss": 0.7948,
      "num_input_tokens_seen": 66506704,
      "step": 114545
    },
    {
      "epoch": 17.061364313375037,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.001930231153213286,
      "loss": 0.7824,
      "num_input_tokens_seen": 66509744,
      "step": 114550
    },
    {
      "epoch": 17.062109025915998,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0019292745320273852,
      "loss": 0.7996,
      "num_input_tokens_seen": 66512944,
      "step": 114555
    },
    {
      "epoch": 17.062853738456955,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.001928318131658111,
      "loss": 0.8044,
      "num_input_tokens_seen": 66516080,
      "step": 114560
    },
    {
      "epoch": 17.063598450997915,
      "grad_norm": 0.060546875,
      "learning_rate": 0.0019273619521216155,
      "loss": 0.8073,
      "num_input_tokens_seen": 66518896,
      "step": 114565
    },
    {
      "epoch": 17.064343163538872,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0019264059934340544,
      "loss": 0.8039,
      "num_input_tokens_seen": 66521648,
      "step": 114570
    },
    {
      "epoch": 17.065087876079833,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.001925450255611576,
      "loss": 0.7846,
      "num_input_tokens_seen": 66524528,
      "step": 114575
    },
    {
      "epoch": 17.065832588620793,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0019244947386703225,
      "loss": 0.7881,
      "num_input_tokens_seen": 66527856,
      "step": 114580
    },
    {
      "epoch": 17.06657730116175,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0019235394426264463,
      "loss": 0.7938,
      "num_input_tokens_seen": 66530512,
      "step": 114585
    },
    {
      "epoch": 17.06732201370271,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0019225843674960774,
      "loss": 0.8004,
      "num_input_tokens_seen": 66533360,
      "step": 114590
    },
    {
      "epoch": 17.06806672624367,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0019216295132953563,
      "loss": 0.7923,
      "num_input_tokens_seen": 66536208,
      "step": 114595
    },
    {
      "epoch": 17.06881143878463,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0019206748800404138,
      "loss": 0.7846,
      "num_input_tokens_seen": 66538864,
      "step": 114600
    },
    {
      "epoch": 17.06955615132559,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0019197204677473716,
      "loss": 0.7868,
      "num_input_tokens_seen": 66541584,
      "step": 114605
    },
    {
      "epoch": 17.070300863866546,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.001918766276432362,
      "loss": 0.7918,
      "num_input_tokens_seen": 66544400,
      "step": 114610
    },
    {
      "epoch": 17.071045576407506,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0019178123061114971,
      "loss": 0.8141,
      "num_input_tokens_seen": 66547056,
      "step": 114615
    },
    {
      "epoch": 17.071790288948467,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0019168585568009005,
      "loss": 0.7842,
      "num_input_tokens_seen": 66550192,
      "step": 114620
    },
    {
      "epoch": 17.072535001489424,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0019159050285166778,
      "loss": 0.8052,
      "num_input_tokens_seen": 66553264,
      "step": 114625
    },
    {
      "epoch": 17.073279714030384,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.001914951721274946,
      "loss": 0.8119,
      "num_input_tokens_seen": 66556336,
      "step": 114630
    },
    {
      "epoch": 17.074024426571345,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0019139986350918058,
      "loss": 0.7708,
      "num_input_tokens_seen": 66559440,
      "step": 114635
    },
    {
      "epoch": 17.074769139112302,
      "grad_norm": 0.03125,
      "learning_rate": 0.0019130457699833574,
      "loss": 0.8296,
      "num_input_tokens_seen": 66562416,
      "step": 114640
    },
    {
      "epoch": 17.075513851653263,
      "grad_norm": 0.69140625,
      "learning_rate": 0.0019120931259656998,
      "loss": 0.8609,
      "num_input_tokens_seen": 66565136,
      "step": 114645
    },
    {
      "epoch": 17.07625856419422,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0019111407030549248,
      "loss": 0.8032,
      "num_input_tokens_seen": 66567984,
      "step": 114650
    },
    {
      "epoch": 17.07700327673518,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0019101885012671282,
      "loss": 0.8023,
      "num_input_tokens_seen": 66570768,
      "step": 114655
    },
    {
      "epoch": 17.07774798927614,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0019092365206183886,
      "loss": 0.8009,
      "num_input_tokens_seen": 66573648,
      "step": 114660
    },
    {
      "epoch": 17.078492701817098,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0019082847611247965,
      "loss": 0.7964,
      "num_input_tokens_seen": 66576752,
      "step": 114665
    },
    {
      "epoch": 17.079237414358058,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.001907333222802424,
      "loss": 0.7994,
      "num_input_tokens_seen": 66579856,
      "step": 114670
    },
    {
      "epoch": 17.07998212689902,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0019063819056673536,
      "loss": 0.7915,
      "num_input_tokens_seen": 66582608,
      "step": 114675
    },
    {
      "epoch": 17.080726839439976,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0019054308097356536,
      "loss": 0.8296,
      "num_input_tokens_seen": 66585840,
      "step": 114680
    },
    {
      "epoch": 17.081471551980936,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0019044799350233866,
      "loss": 0.7936,
      "num_input_tokens_seen": 66588944,
      "step": 114685
    },
    {
      "epoch": 17.082216264521893,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0019035292815466258,
      "loss": 0.8119,
      "num_input_tokens_seen": 66591696,
      "step": 114690
    },
    {
      "epoch": 17.082960977062854,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.001902578849321424,
      "loss": 0.795,
      "num_input_tokens_seen": 66595088,
      "step": 114695
    },
    {
      "epoch": 17.083705689603814,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0019016286383638447,
      "loss": 0.7881,
      "num_input_tokens_seen": 66597872,
      "step": 114700
    },
    {
      "epoch": 17.08445040214477,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0019006786486899351,
      "loss": 0.8068,
      "num_input_tokens_seen": 66600752,
      "step": 114705
    },
    {
      "epoch": 17.08519511468573,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0018997288803157474,
      "loss": 0.7855,
      "num_input_tokens_seen": 66603632,
      "step": 114710
    },
    {
      "epoch": 17.085939827226692,
      "grad_norm": 0.0546875,
      "learning_rate": 0.001898779333257322,
      "loss": 0.7918,
      "num_input_tokens_seen": 66606608,
      "step": 114715
    },
    {
      "epoch": 17.08668453976765,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.001897830007530708,
      "loss": 0.7956,
      "num_input_tokens_seen": 66609488,
      "step": 114720
    },
    {
      "epoch": 17.08742925230861,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0018968809031519407,
      "loss": 0.7848,
      "num_input_tokens_seen": 66612560,
      "step": 114725
    },
    {
      "epoch": 17.088173964849567,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.0018959320201370488,
      "loss": 0.8069,
      "num_input_tokens_seen": 66615440,
      "step": 114730
    },
    {
      "epoch": 17.088918677390527,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0018949833585020697,
      "loss": 0.7831,
      "num_input_tokens_seen": 66618224,
      "step": 114735
    },
    {
      "epoch": 17.089663389931488,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.001894034918263024,
      "loss": 0.8186,
      "num_input_tokens_seen": 66621008,
      "step": 114740
    },
    {
      "epoch": 17.090408102472445,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0018930866994359419,
      "loss": 0.805,
      "num_input_tokens_seen": 66623856,
      "step": 114745
    },
    {
      "epoch": 17.091152815013405,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0018921387020368356,
      "loss": 0.802,
      "num_input_tokens_seen": 66626512,
      "step": 114750
    },
    {
      "epoch": 17.091897527554362,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0018911909260817276,
      "loss": 0.7952,
      "num_input_tokens_seen": 66629456,
      "step": 114755
    },
    {
      "epoch": 17.092642240095323,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.001890243371586625,
      "loss": 0.7743,
      "num_input_tokens_seen": 66632368,
      "step": 114760
    },
    {
      "epoch": 17.093386952636283,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0018892960385675332,
      "loss": 0.8067,
      "num_input_tokens_seen": 66635440,
      "step": 114765
    },
    {
      "epoch": 17.09413166517724,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0018883489270404628,
      "loss": 0.8075,
      "num_input_tokens_seen": 66638480,
      "step": 114770
    },
    {
      "epoch": 17.0948763777182,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0018874020370214112,
      "loss": 0.808,
      "num_input_tokens_seen": 66641488,
      "step": 114775
    },
    {
      "epoch": 17.09562109025916,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0018864553685263752,
      "loss": 0.7927,
      "num_input_tokens_seen": 66644432,
      "step": 114780
    },
    {
      "epoch": 17.09636580280012,
      "grad_norm": 0.388671875,
      "learning_rate": 0.0018855089215713437,
      "loss": 0.8289,
      "num_input_tokens_seen": 66647248,
      "step": 114785
    },
    {
      "epoch": 17.09711051534108,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0018845626961723145,
      "loss": 0.8051,
      "num_input_tokens_seen": 66650064,
      "step": 114790
    },
    {
      "epoch": 17.097855227882036,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0018836166923452656,
      "loss": 0.7848,
      "num_input_tokens_seen": 66653040,
      "step": 114795
    },
    {
      "epoch": 17.098599940422996,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0018826709101061833,
      "loss": 0.7807,
      "num_input_tokens_seen": 66655760,
      "step": 114800
    },
    {
      "epoch": 17.099344652963957,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.0018817253494710462,
      "loss": 0.7966,
      "num_input_tokens_seen": 66658448,
      "step": 114805
    },
    {
      "epoch": 17.100089365504914,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0018807800104558214,
      "loss": 0.8024,
      "num_input_tokens_seen": 66661456,
      "step": 114810
    },
    {
      "epoch": 17.100834078045875,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0018798348930764896,
      "loss": 0.8022,
      "num_input_tokens_seen": 66664272,
      "step": 114815
    },
    {
      "epoch": 17.101578790586835,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.001878889997349008,
      "loss": 0.8042,
      "num_input_tokens_seen": 66666992,
      "step": 114820
    },
    {
      "epoch": 17.102323503127792,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0018779453232893471,
      "loss": 0.8127,
      "num_input_tokens_seen": 66669968,
      "step": 114825
    },
    {
      "epoch": 17.103068215668753,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0018770008709134644,
      "loss": 0.7774,
      "num_input_tokens_seen": 66672912,
      "step": 114830
    },
    {
      "epoch": 17.10381292820971,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.00187605664023731,
      "loss": 0.8142,
      "num_input_tokens_seen": 66675728,
      "step": 114835
    },
    {
      "epoch": 17.10455764075067,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.001875112631276845,
      "loss": 0.7983,
      "num_input_tokens_seen": 66678832,
      "step": 114840
    },
    {
      "epoch": 17.10530235329163,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0018741688440480114,
      "loss": 0.7921,
      "num_input_tokens_seen": 66681552,
      "step": 114845
    },
    {
      "epoch": 17.106047065832588,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0018732252785667546,
      "loss": 0.7878,
      "num_input_tokens_seen": 66684240,
      "step": 114850
    },
    {
      "epoch": 17.106791778373548,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0018722819348490122,
      "loss": 0.7944,
      "num_input_tokens_seen": 66686992,
      "step": 114855
    },
    {
      "epoch": 17.10753649091451,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0018713388129107276,
      "loss": 0.8079,
      "num_input_tokens_seen": 66690000,
      "step": 114860
    },
    {
      "epoch": 17.108281203455466,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.001870395912767827,
      "loss": 0.7991,
      "num_input_tokens_seen": 66692688,
      "step": 114865
    },
    {
      "epoch": 17.109025915996426,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0018694532344362457,
      "loss": 0.7828,
      "num_input_tokens_seen": 66695632,
      "step": 114870
    },
    {
      "epoch": 17.109770628537383,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0018685107779319043,
      "loss": 0.7921,
      "num_input_tokens_seen": 66698416,
      "step": 114875
    },
    {
      "epoch": 17.110515341078344,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00186756854327073,
      "loss": 0.8098,
      "num_input_tokens_seen": 66701520,
      "step": 114880
    },
    {
      "epoch": 17.111260053619304,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0018666265304686386,
      "loss": 0.7736,
      "num_input_tokens_seen": 66704560,
      "step": 114885
    },
    {
      "epoch": 17.11200476616026,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0018656847395415405,
      "loss": 0.8099,
      "num_input_tokens_seen": 66707408,
      "step": 114890
    },
    {
      "epoch": 17.11274947870122,
      "grad_norm": 0.0306396484375,
      "learning_rate": 0.0018647431705053512,
      "loss": 0.7917,
      "num_input_tokens_seen": 66710320,
      "step": 114895
    },
    {
      "epoch": 17.113494191242182,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0018638018233759783,
      "loss": 0.7907,
      "num_input_tokens_seen": 66713072,
      "step": 114900
    },
    {
      "epoch": 17.11423890378314,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0018628606981693208,
      "loss": 0.7887,
      "num_input_tokens_seen": 66716080,
      "step": 114905
    },
    {
      "epoch": 17.1149836163241,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0018619197949012773,
      "loss": 0.8045,
      "num_input_tokens_seen": 66718992,
      "step": 114910
    },
    {
      "epoch": 17.115728328865057,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0018609791135877484,
      "loss": 0.811,
      "num_input_tokens_seen": 66721936,
      "step": 114915
    },
    {
      "epoch": 17.116473041406017,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0018600386542446251,
      "loss": 0.8197,
      "num_input_tokens_seen": 66724496,
      "step": 114920
    },
    {
      "epoch": 17.117217753946978,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0018590984168877893,
      "loss": 0.7872,
      "num_input_tokens_seen": 66727504,
      "step": 114925
    },
    {
      "epoch": 17.117962466487935,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0018581584015331337,
      "loss": 0.7661,
      "num_input_tokens_seen": 66730352,
      "step": 114930
    },
    {
      "epoch": 17.118707179028895,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0018572186081965301,
      "loss": 0.813,
      "num_input_tokens_seen": 66733648,
      "step": 114935
    },
    {
      "epoch": 17.119451891569852,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0018562790368938664,
      "loss": 0.8012,
      "num_input_tokens_seen": 66736784,
      "step": 114940
    },
    {
      "epoch": 17.120196604110813,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0018553396876410044,
      "loss": 0.7883,
      "num_input_tokens_seen": 66739792,
      "step": 114945
    },
    {
      "epoch": 17.120941316651773,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0018544005604538233,
      "loss": 0.7914,
      "num_input_tokens_seen": 66743088,
      "step": 114950
    },
    {
      "epoch": 17.12168602919273,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0018534616553481803,
      "loss": 0.809,
      "num_input_tokens_seen": 66745840,
      "step": 114955
    },
    {
      "epoch": 17.12243074173369,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0018525229723399443,
      "loss": 0.7842,
      "num_input_tokens_seen": 66748624,
      "step": 114960
    },
    {
      "epoch": 17.12317545427465,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0018515845114449713,
      "loss": 0.7933,
      "num_input_tokens_seen": 66751536,
      "step": 114965
    },
    {
      "epoch": 17.12392016681561,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0018506462726791133,
      "loss": 0.8021,
      "num_input_tokens_seen": 66754576,
      "step": 114970
    },
    {
      "epoch": 17.12466487935657,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.001849708256058221,
      "loss": 0.778,
      "num_input_tokens_seen": 66757840,
      "step": 114975
    },
    {
      "epoch": 17.125409591897526,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.001848770461598142,
      "loss": 0.7815,
      "num_input_tokens_seen": 66760720,
      "step": 114980
    },
    {
      "epoch": 17.126154304438487,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0018478328893147217,
      "loss": 0.7779,
      "num_input_tokens_seen": 66763504,
      "step": 114985
    },
    {
      "epoch": 17.126899016979447,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.001846895539223794,
      "loss": 0.7895,
      "num_input_tokens_seen": 66766256,
      "step": 114990
    },
    {
      "epoch": 17.127643729520404,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0018459584113411998,
      "loss": 0.7915,
      "num_input_tokens_seen": 66769136,
      "step": 114995
    },
    {
      "epoch": 17.128388442061365,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0018450215056827711,
      "loss": 0.8092,
      "num_input_tokens_seen": 66772112,
      "step": 115000
    },
    {
      "epoch": 17.129133154602325,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.0018440848222643308,
      "loss": 0.7874,
      "num_input_tokens_seen": 66775024,
      "step": 115005
    },
    {
      "epoch": 17.129877867143282,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0018431483611017073,
      "loss": 0.8041,
      "num_input_tokens_seen": 66777840,
      "step": 115010
    },
    {
      "epoch": 17.130622579684243,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.00184221212221072,
      "loss": 0.803,
      "num_input_tokens_seen": 66781104,
      "step": 115015
    },
    {
      "epoch": 17.1313672922252,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0018412761056071864,
      "loss": 0.8131,
      "num_input_tokens_seen": 66783760,
      "step": 115020
    },
    {
      "epoch": 17.13211200476616,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.0018403403113069182,
      "loss": 0.8062,
      "num_input_tokens_seen": 66786544,
      "step": 115025
    },
    {
      "epoch": 17.13285671730712,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0018394047393257267,
      "loss": 0.8031,
      "num_input_tokens_seen": 66789360,
      "step": 115030
    },
    {
      "epoch": 17.133601429848078,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0018384693896794173,
      "loss": 0.776,
      "num_input_tokens_seen": 66792368,
      "step": 115035
    },
    {
      "epoch": 17.134346142389038,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.001837534262383791,
      "loss": 0.7841,
      "num_input_tokens_seen": 66795216,
      "step": 115040
    },
    {
      "epoch": 17.13509085493,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0018365993574546461,
      "loss": 0.7743,
      "num_input_tokens_seen": 66798160,
      "step": 115045
    },
    {
      "epoch": 17.135835567470956,
      "grad_norm": 0.046875,
      "learning_rate": 0.0018356646749077726,
      "loss": 0.8073,
      "num_input_tokens_seen": 66800880,
      "step": 115050
    },
    {
      "epoch": 17.136580280011916,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0018347302147589672,
      "loss": 0.7954,
      "num_input_tokens_seen": 66803952,
      "step": 115055
    },
    {
      "epoch": 17.137324992552873,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.0018337959770240108,
      "loss": 0.7978,
      "num_input_tokens_seen": 66806672,
      "step": 115060
    },
    {
      "epoch": 17.138069705093834,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0018328619617186941,
      "loss": 0.7826,
      "num_input_tokens_seen": 66809648,
      "step": 115065
    },
    {
      "epoch": 17.138814417634794,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0018319281688587896,
      "loss": 0.782,
      "num_input_tokens_seen": 66812656,
      "step": 115070
    },
    {
      "epoch": 17.13955913017575,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0018309945984600778,
      "loss": 0.8106,
      "num_input_tokens_seen": 66815280,
      "step": 115075
    },
    {
      "epoch": 17.140303842716712,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0018300612505383245,
      "loss": 0.8012,
      "num_input_tokens_seen": 66818384,
      "step": 115080
    },
    {
      "epoch": 17.14104855525767,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0018291281251093037,
      "loss": 0.7977,
      "num_input_tokens_seen": 66821584,
      "step": 115085
    },
    {
      "epoch": 17.14179326779863,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0018281952221887764,
      "loss": 0.7869,
      "num_input_tokens_seen": 66824528,
      "step": 115090
    },
    {
      "epoch": 17.14253798033959,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0018272625417925013,
      "loss": 0.785,
      "num_input_tokens_seen": 66827248,
      "step": 115095
    },
    {
      "epoch": 17.143282692880547,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0018263300839362411,
      "loss": 0.8161,
      "num_input_tokens_seen": 66830224,
      "step": 115100
    },
    {
      "epoch": 17.144027405421507,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0018253978486357447,
      "loss": 0.7941,
      "num_input_tokens_seen": 66833232,
      "step": 115105
    },
    {
      "epoch": 17.144772117962468,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.001824465835906761,
      "loss": 0.7973,
      "num_input_tokens_seen": 66836432,
      "step": 115110
    },
    {
      "epoch": 17.145516830503425,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0018235340457650345,
      "loss": 0.8037,
      "num_input_tokens_seen": 66839184,
      "step": 115115
    },
    {
      "epoch": 17.146261543044385,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0018226024782263105,
      "loss": 0.7986,
      "num_input_tokens_seen": 66841968,
      "step": 115120
    },
    {
      "epoch": 17.147006255585342,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.0018216711333063234,
      "loss": 0.7814,
      "num_input_tokens_seen": 66845072,
      "step": 115125
    },
    {
      "epoch": 17.147750968126303,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.001820740011020807,
      "loss": 0.8075,
      "num_input_tokens_seen": 66847888,
      "step": 115130
    },
    {
      "epoch": 17.148495680667263,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0018198091113854958,
      "loss": 0.8193,
      "num_input_tokens_seen": 66850768,
      "step": 115135
    },
    {
      "epoch": 17.14924039320822,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0018188784344161117,
      "loss": 0.8023,
      "num_input_tokens_seen": 66853968,
      "step": 115140
    },
    {
      "epoch": 17.14998510574918,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0018179479801283826,
      "loss": 0.8138,
      "num_input_tokens_seen": 66856816,
      "step": 115145
    },
    {
      "epoch": 17.15072981829014,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0018170177485380206,
      "loss": 0.7762,
      "num_input_tokens_seen": 66859312,
      "step": 115150
    },
    {
      "epoch": 17.1514745308311,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0018160877396607483,
      "loss": 0.8124,
      "num_input_tokens_seen": 66862352,
      "step": 115155
    },
    {
      "epoch": 17.15221924337206,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0018151579535122747,
      "loss": 0.8029,
      "num_input_tokens_seen": 66865104,
      "step": 115160
    },
    {
      "epoch": 17.152963955913016,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.001814228390108304,
      "loss": 0.7764,
      "num_input_tokens_seen": 66868208,
      "step": 115165
    },
    {
      "epoch": 17.153708668453977,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0018132990494645468,
      "loss": 0.8087,
      "num_input_tokens_seen": 66871248,
      "step": 115170
    },
    {
      "epoch": 17.154453380994937,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.001812369931596699,
      "loss": 0.8019,
      "num_input_tokens_seen": 66874192,
      "step": 115175
    },
    {
      "epoch": 17.155198093535894,
      "grad_norm": 0.04296875,
      "learning_rate": 0.001811441036520458,
      "loss": 0.7825,
      "num_input_tokens_seen": 66877200,
      "step": 115180
    },
    {
      "epoch": 17.155942806076855,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0018105123642515146,
      "loss": 0.7976,
      "num_input_tokens_seen": 66880176,
      "step": 115185
    },
    {
      "epoch": 17.156687518617815,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0018095839148055614,
      "loss": 0.7989,
      "num_input_tokens_seen": 66882960,
      "step": 115190
    },
    {
      "epoch": 17.157432231158772,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.001808655688198279,
      "loss": 0.7912,
      "num_input_tokens_seen": 66885936,
      "step": 115195
    },
    {
      "epoch": 17.158176943699733,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0018077276844453549,
      "loss": 0.7827,
      "num_input_tokens_seen": 66888688,
      "step": 115200
    },
    {
      "epoch": 17.15892165624069,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0018067999035624631,
      "loss": 0.8006,
      "num_input_tokens_seen": 66891792,
      "step": 115205
    },
    {
      "epoch": 17.15966636878165,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0018058723455652748,
      "loss": 0.7948,
      "num_input_tokens_seen": 66894832,
      "step": 115210
    },
    {
      "epoch": 17.16041108132261,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.001804945010469467,
      "loss": 0.8306,
      "num_input_tokens_seen": 66897520,
      "step": 115215
    },
    {
      "epoch": 17.161155793863568,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.001804017898290699,
      "loss": 0.7991,
      "num_input_tokens_seen": 66900240,
      "step": 115220
    },
    {
      "epoch": 17.16190050640453,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.00180309100904464,
      "loss": 0.8022,
      "num_input_tokens_seen": 66903152,
      "step": 115225
    },
    {
      "epoch": 17.16264521894549,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.001802164342746944,
      "loss": 0.7991,
      "num_input_tokens_seen": 66906224,
      "step": 115230
    },
    {
      "epoch": 17.163389931486446,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0018012378994132704,
      "loss": 0.8014,
      "num_input_tokens_seen": 66908816,
      "step": 115235
    },
    {
      "epoch": 17.164134644027406,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0018003116790592632,
      "loss": 0.8073,
      "num_input_tokens_seen": 66912048,
      "step": 115240
    },
    {
      "epoch": 17.164879356568363,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.001799385681700578,
      "loss": 0.8111,
      "num_input_tokens_seen": 66914864,
      "step": 115245
    },
    {
      "epoch": 17.165624069109324,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0017984599073528561,
      "loss": 0.7845,
      "num_input_tokens_seen": 66917520,
      "step": 115250
    },
    {
      "epoch": 17.166368781650284,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.001797534356031733,
      "loss": 0.785,
      "num_input_tokens_seen": 66920272,
      "step": 115255
    },
    {
      "epoch": 17.16711349419124,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0017966090277528512,
      "loss": 0.7706,
      "num_input_tokens_seen": 66923056,
      "step": 115260
    },
    {
      "epoch": 17.167858206732202,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0017956839225318382,
      "loss": 0.7881,
      "num_input_tokens_seen": 66926064,
      "step": 115265
    },
    {
      "epoch": 17.16860291927316,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0017947590403843267,
      "loss": 0.7859,
      "num_input_tokens_seen": 66930096,
      "step": 115270
    },
    {
      "epoch": 17.16934763181412,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0017938343813259387,
      "loss": 0.7764,
      "num_input_tokens_seen": 66932752,
      "step": 115275
    },
    {
      "epoch": 17.17009234435508,
      "grad_norm": 0.078125,
      "learning_rate": 0.001792909945372299,
      "loss": 0.792,
      "num_input_tokens_seen": 66936144,
      "step": 115280
    },
    {
      "epoch": 17.170837056896037,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0017919857325390215,
      "loss": 0.7947,
      "num_input_tokens_seen": 66939120,
      "step": 115285
    },
    {
      "epoch": 17.171581769436997,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0017910617428417203,
      "loss": 0.7766,
      "num_input_tokens_seen": 66942032,
      "step": 115290
    },
    {
      "epoch": 17.172326481977958,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0017901379762960078,
      "loss": 0.8001,
      "num_input_tokens_seen": 66944880,
      "step": 115295
    },
    {
      "epoch": 17.173071194518915,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0017892144329174884,
      "loss": 0.8438,
      "num_input_tokens_seen": 66947760,
      "step": 115300
    },
    {
      "epoch": 17.173815907059875,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0017882911127217631,
      "loss": 0.7905,
      "num_input_tokens_seen": 66950608,
      "step": 115305
    },
    {
      "epoch": 17.174560619600832,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0017873680157244308,
      "loss": 0.7984,
      "num_input_tokens_seen": 66953424,
      "step": 115310
    },
    {
      "epoch": 17.175305332141793,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0017864451419410893,
      "loss": 0.8267,
      "num_input_tokens_seen": 66956208,
      "step": 115315
    },
    {
      "epoch": 17.176050044682754,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.001785522491387324,
      "loss": 0.8021,
      "num_input_tokens_seen": 66958928,
      "step": 115320
    },
    {
      "epoch": 17.17679475722371,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.001784600064078728,
      "loss": 0.8131,
      "num_input_tokens_seen": 66961840,
      "step": 115325
    },
    {
      "epoch": 17.17753946976467,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0017836778600308833,
      "loss": 0.7843,
      "num_input_tokens_seen": 66964816,
      "step": 115330
    },
    {
      "epoch": 17.17828418230563,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.001782755879259366,
      "loss": 0.802,
      "num_input_tokens_seen": 66967760,
      "step": 115335
    },
    {
      "epoch": 17.17902889484659,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0017818341217797572,
      "loss": 0.7883,
      "num_input_tokens_seen": 66970544,
      "step": 115340
    },
    {
      "epoch": 17.17977360738755,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0017809125876076242,
      "loss": 0.8007,
      "num_input_tokens_seen": 66973488,
      "step": 115345
    },
    {
      "epoch": 17.180518319928506,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0017799912767585413,
      "loss": 0.7846,
      "num_input_tokens_seen": 66976240,
      "step": 115350
    },
    {
      "epoch": 17.181263032469467,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.0017790701892480643,
      "loss": 0.7631,
      "num_input_tokens_seen": 66979088,
      "step": 115355
    },
    {
      "epoch": 17.182007745010427,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.001778149325091764,
      "loss": 0.7913,
      "num_input_tokens_seen": 66981488,
      "step": 115360
    },
    {
      "epoch": 17.182752457551384,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.001777228684305193,
      "loss": 0.8025,
      "num_input_tokens_seen": 66984144,
      "step": 115365
    },
    {
      "epoch": 17.183497170092345,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0017763082669039025,
      "loss": 0.7852,
      "num_input_tokens_seen": 66986960,
      "step": 115370
    },
    {
      "epoch": 17.184241882633305,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0017753880729034464,
      "loss": 0.7817,
      "num_input_tokens_seen": 66989808,
      "step": 115375
    },
    {
      "epoch": 17.184986595174262,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0017744681023193626,
      "loss": 0.7977,
      "num_input_tokens_seen": 66992560,
      "step": 115380
    },
    {
      "epoch": 17.185731307715223,
      "grad_norm": 0.078125,
      "learning_rate": 0.0017735483551672032,
      "loss": 0.7891,
      "num_input_tokens_seen": 66995728,
      "step": 115385
    },
    {
      "epoch": 17.18647602025618,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0017726288314624976,
      "loss": 0.8092,
      "num_input_tokens_seen": 66998640,
      "step": 115390
    },
    {
      "epoch": 17.18722073279714,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.001771709531220787,
      "loss": 0.7925,
      "num_input_tokens_seen": 67001264,
      "step": 115395
    },
    {
      "epoch": 17.1879654453381,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.001770790454457597,
      "loss": 0.8155,
      "num_input_tokens_seen": 67003920,
      "step": 115400
    },
    {
      "epoch": 17.188710157879058,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0017698716011884585,
      "loss": 0.8058,
      "num_input_tokens_seen": 67006704,
      "step": 115405
    },
    {
      "epoch": 17.18945487042002,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0017689529714288942,
      "loss": 0.8118,
      "num_input_tokens_seen": 67009424,
      "step": 115410
    },
    {
      "epoch": 17.19019958296098,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0017680345651944183,
      "loss": 0.7716,
      "num_input_tokens_seen": 67012208,
      "step": 115415
    },
    {
      "epoch": 17.190944295501936,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0017671163825005536,
      "loss": 0.7721,
      "num_input_tokens_seen": 67015088,
      "step": 115420
    },
    {
      "epoch": 17.191689008042896,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.001766198423362804,
      "loss": 0.7796,
      "num_input_tokens_seen": 67018064,
      "step": 115425
    },
    {
      "epoch": 17.192433720583853,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0017652806877966859,
      "loss": 0.7896,
      "num_input_tokens_seen": 67021008,
      "step": 115430
    },
    {
      "epoch": 17.193178433124814,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0017643631758176997,
      "loss": 0.8105,
      "num_input_tokens_seen": 67024080,
      "step": 115435
    },
    {
      "epoch": 17.193923145665774,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.001763445887441345,
      "loss": 0.8233,
      "num_input_tokens_seen": 67027056,
      "step": 115440
    },
    {
      "epoch": 17.19466785820673,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0017625288226831193,
      "loss": 0.7853,
      "num_input_tokens_seen": 67029936,
      "step": 115445
    },
    {
      "epoch": 17.195412570747692,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0017616119815585117,
      "loss": 0.7908,
      "num_input_tokens_seen": 67032784,
      "step": 115450
    },
    {
      "epoch": 17.19615728328865,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0017606953640830186,
      "loss": 0.8217,
      "num_input_tokens_seen": 67035408,
      "step": 115455
    },
    {
      "epoch": 17.19690199582961,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0017597789702721172,
      "loss": 0.7936,
      "num_input_tokens_seen": 67038352,
      "step": 115460
    },
    {
      "epoch": 17.19764670837057,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.001758862800141297,
      "loss": 0.7831,
      "num_input_tokens_seen": 67041264,
      "step": 115465
    },
    {
      "epoch": 17.198391420911527,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0017579468537060289,
      "loss": 0.8139,
      "num_input_tokens_seen": 67044176,
      "step": 115470
    },
    {
      "epoch": 17.199136133452487,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0017570311309817936,
      "loss": 0.8063,
      "num_input_tokens_seen": 67046768,
      "step": 115475
    },
    {
      "epoch": 17.199880845993448,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0017561156319840542,
      "loss": 0.7739,
      "num_input_tokens_seen": 67049648,
      "step": 115480
    },
    {
      "epoch": 17.200625558534405,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0017552003567282847,
      "loss": 0.7989,
      "num_input_tokens_seen": 67052720,
      "step": 115485
    },
    {
      "epoch": 17.201370271075366,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0017542853052299428,
      "loss": 0.8179,
      "num_input_tokens_seen": 67055344,
      "step": 115490
    },
    {
      "epoch": 17.202114983616323,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0017533704775044845,
      "loss": 0.7922,
      "num_input_tokens_seen": 67058160,
      "step": 115495
    },
    {
      "epoch": 17.202859696157283,
      "grad_norm": 0.046875,
      "learning_rate": 0.001752455873567374,
      "loss": 0.796,
      "num_input_tokens_seen": 67061072,
      "step": 115500
    },
    {
      "epoch": 17.203604408698244,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0017515414934340557,
      "loss": 0.7757,
      "num_input_tokens_seen": 67063952,
      "step": 115505
    },
    {
      "epoch": 17.2043491212392,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0017506273371199787,
      "loss": 0.7685,
      "num_input_tokens_seen": 67066960,
      "step": 115510
    },
    {
      "epoch": 17.20509383378016,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0017497134046405842,
      "loss": 0.7963,
      "num_input_tokens_seen": 67069840,
      "step": 115515
    },
    {
      "epoch": 17.20583854632112,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0017487996960113183,
      "loss": 0.8012,
      "num_input_tokens_seen": 67072528,
      "step": 115520
    },
    {
      "epoch": 17.20658325886208,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0017478862112476133,
      "loss": 0.7979,
      "num_input_tokens_seen": 67075696,
      "step": 115525
    },
    {
      "epoch": 17.20732797140304,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0017469729503648988,
      "loss": 0.821,
      "num_input_tokens_seen": 67078352,
      "step": 115530
    },
    {
      "epoch": 17.208072683943996,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0017460599133786092,
      "loss": 0.7836,
      "num_input_tokens_seen": 67081424,
      "step": 115535
    },
    {
      "epoch": 17.208817396484957,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0017451471003041635,
      "loss": 0.8169,
      "num_input_tokens_seen": 67084304,
      "step": 115540
    },
    {
      "epoch": 17.209562109025917,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0017442345111569895,
      "loss": 0.783,
      "num_input_tokens_seen": 67087344,
      "step": 115545
    },
    {
      "epoch": 17.210306821566874,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0017433221459524982,
      "loss": 0.7882,
      "num_input_tokens_seen": 67090448,
      "step": 115550
    },
    {
      "epoch": 17.211051534107835,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0017424100047061075,
      "loss": 0.7958,
      "num_input_tokens_seen": 67093392,
      "step": 115555
    },
    {
      "epoch": 17.211796246648795,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0017414980874332264,
      "loss": 0.8092,
      "num_input_tokens_seen": 67096560,
      "step": 115560
    },
    {
      "epoch": 17.212540959189752,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0017405863941492594,
      "loss": 0.7949,
      "num_input_tokens_seen": 67099376,
      "step": 115565
    },
    {
      "epoch": 17.213285671730713,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0017396749248696091,
      "loss": 0.7889,
      "num_input_tokens_seen": 67102096,
      "step": 115570
    },
    {
      "epoch": 17.21403038427167,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.00173876367960967,
      "loss": 0.8002,
      "num_input_tokens_seen": 67104848,
      "step": 115575
    },
    {
      "epoch": 17.21477509681263,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0017378526583848447,
      "loss": 0.785,
      "num_input_tokens_seen": 67107824,
      "step": 115580
    },
    {
      "epoch": 17.21551980935359,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0017369418612105158,
      "loss": 0.7953,
      "num_input_tokens_seen": 67110640,
      "step": 115585
    },
    {
      "epoch": 17.216264521894548,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0017360312881020779,
      "loss": 0.7772,
      "num_input_tokens_seen": 67113520,
      "step": 115590
    },
    {
      "epoch": 17.21700923443551,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0017351209390749067,
      "loss": 0.7681,
      "num_input_tokens_seen": 67116400,
      "step": 115595
    },
    {
      "epoch": 17.217753946976465,
      "grad_norm": 0.330078125,
      "learning_rate": 0.0017342108141443883,
      "loss": 0.8243,
      "num_input_tokens_seen": 67119440,
      "step": 115600
    },
    {
      "epoch": 17.218498659517426,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.0017333009133258975,
      "loss": 0.8136,
      "num_input_tokens_seen": 67122512,
      "step": 115605
    },
    {
      "epoch": 17.219243372058386,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0017323912366347981,
      "loss": 0.8073,
      "num_input_tokens_seen": 67125264,
      "step": 115610
    },
    {
      "epoch": 17.219988084599343,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0017314817840864698,
      "loss": 0.8061,
      "num_input_tokens_seen": 67128144,
      "step": 115615
    },
    {
      "epoch": 17.220732797140304,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0017305725556962669,
      "loss": 0.7782,
      "num_input_tokens_seen": 67131248,
      "step": 115620
    },
    {
      "epoch": 17.221477509681264,
      "grad_norm": 0.0703125,
      "learning_rate": 0.001729663551479557,
      "loss": 0.7809,
      "num_input_tokens_seen": 67134256,
      "step": 115625
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 0.2158203125,
      "learning_rate": 0.0017287547714516947,
      "loss": 0.8011,
      "num_input_tokens_seen": 67137264,
      "step": 115630
    },
    {
      "epoch": 17.222966934763182,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0017278462156280327,
      "loss": 0.807,
      "num_input_tokens_seen": 67139984,
      "step": 115635
    },
    {
      "epoch": 17.22371164730414,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0017269378840239153,
      "loss": 0.8041,
      "num_input_tokens_seen": 67142736,
      "step": 115640
    },
    {
      "epoch": 17.2244563598451,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.001726029776654695,
      "loss": 0.7818,
      "num_input_tokens_seen": 67145648,
      "step": 115645
    },
    {
      "epoch": 17.22520107238606,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0017251218935357114,
      "loss": 0.7975,
      "num_input_tokens_seen": 67148496,
      "step": 115650
    },
    {
      "epoch": 17.225945784927017,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0017242142346822992,
      "loss": 0.7934,
      "num_input_tokens_seen": 67151568,
      "step": 115655
    },
    {
      "epoch": 17.226690497467978,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0017233068001097972,
      "loss": 0.803,
      "num_input_tokens_seen": 67154352,
      "step": 115660
    },
    {
      "epoch": 17.227435210008938,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0017223995898335287,
      "loss": 0.7852,
      "num_input_tokens_seen": 67157264,
      "step": 115665
    },
    {
      "epoch": 17.228179922549895,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0017214926038688293,
      "loss": 0.7733,
      "num_input_tokens_seen": 67159824,
      "step": 115670
    },
    {
      "epoch": 17.228924635090856,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.001720585842231012,
      "loss": 0.8082,
      "num_input_tokens_seen": 67162640,
      "step": 115675
    },
    {
      "epoch": 17.229669347631813,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0017196793049354046,
      "loss": 0.7972,
      "num_input_tokens_seen": 67165456,
      "step": 115680
    },
    {
      "epoch": 17.230414060172773,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0017187729919973132,
      "loss": 0.7699,
      "num_input_tokens_seen": 67168176,
      "step": 115685
    },
    {
      "epoch": 17.231158772713734,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0017178669034320587,
      "loss": 0.7767,
      "num_input_tokens_seen": 67171184,
      "step": 115690
    },
    {
      "epoch": 17.23190348525469,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.001716961039254942,
      "loss": 0.797,
      "num_input_tokens_seen": 67173808,
      "step": 115695
    },
    {
      "epoch": 17.23264819779565,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0017160553994812681,
      "loss": 0.79,
      "num_input_tokens_seen": 67176656,
      "step": 115700
    },
    {
      "epoch": 17.23339291033661,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.001715149984126338,
      "loss": 0.7936,
      "num_input_tokens_seen": 67179664,
      "step": 115705
    },
    {
      "epoch": 17.23413762287757,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0017142447932054438,
      "loss": 0.7823,
      "num_input_tokens_seen": 67182384,
      "step": 115710
    },
    {
      "epoch": 17.23488233541853,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.001713339826733884,
      "loss": 0.8138,
      "num_input_tokens_seen": 67185488,
      "step": 115715
    },
    {
      "epoch": 17.235627047959486,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0017124350847269392,
      "loss": 0.819,
      "num_input_tokens_seen": 67188336,
      "step": 115720
    },
    {
      "epoch": 17.236371760500447,
      "grad_norm": 0.0625,
      "learning_rate": 0.0017115305671999042,
      "loss": 0.7839,
      "num_input_tokens_seen": 67191376,
      "step": 115725
    },
    {
      "epoch": 17.237116473041407,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0017106262741680533,
      "loss": 0.8016,
      "num_input_tokens_seen": 67194288,
      "step": 115730
    },
    {
      "epoch": 17.237861185582364,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0017097222056466608,
      "loss": 0.8199,
      "num_input_tokens_seen": 67197296,
      "step": 115735
    },
    {
      "epoch": 17.238605898123325,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0017088183616510082,
      "loss": 0.7915,
      "num_input_tokens_seen": 67200656,
      "step": 115740
    },
    {
      "epoch": 17.239350610664285,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0017079147421963563,
      "loss": 0.8178,
      "num_input_tokens_seen": 67203568,
      "step": 115745
    },
    {
      "epoch": 17.240095323205242,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0017070113472979797,
      "loss": 0.7987,
      "num_input_tokens_seen": 67206608,
      "step": 115750
    },
    {
      "epoch": 17.240840035746203,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0017061081769711329,
      "loss": 0.8127,
      "num_input_tokens_seen": 67209520,
      "step": 115755
    },
    {
      "epoch": 17.24158474828716,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00170520523123108,
      "loss": 0.8098,
      "num_input_tokens_seen": 67212240,
      "step": 115760
    },
    {
      "epoch": 17.24232946082812,
      "grad_norm": 0.4765625,
      "learning_rate": 0.001704302510093074,
      "loss": 0.8243,
      "num_input_tokens_seen": 67215440,
      "step": 115765
    },
    {
      "epoch": 17.24307417336908,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0017034000135723614,
      "loss": 0.7854,
      "num_input_tokens_seen": 67218512,
      "step": 115770
    },
    {
      "epoch": 17.243818885910038,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0017024977416841929,
      "loss": 0.7833,
      "num_input_tokens_seen": 67221392,
      "step": 115775
    },
    {
      "epoch": 17.244563598451,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0017015956944438065,
      "loss": 0.7868,
      "num_input_tokens_seen": 67224336,
      "step": 115780
    },
    {
      "epoch": 17.245308310991955,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0017006938718664483,
      "loss": 0.7851,
      "num_input_tokens_seen": 67227280,
      "step": 115785
    },
    {
      "epoch": 17.246053023532916,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.0016997922739673476,
      "loss": 0.8246,
      "num_input_tokens_seen": 67230128,
      "step": 115790
    },
    {
      "epoch": 17.246797736073876,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0016988909007617408,
      "loss": 0.7977,
      "num_input_tokens_seen": 67233072,
      "step": 115795
    },
    {
      "epoch": 17.247542448614833,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.0016979897522648524,
      "loss": 0.7879,
      "num_input_tokens_seen": 67235952,
      "step": 115800
    },
    {
      "epoch": 17.248287161155794,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0016970888284919083,
      "loss": 0.7993,
      "num_input_tokens_seen": 67238896,
      "step": 115805
    },
    {
      "epoch": 17.249031873696755,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.00169618812945813,
      "loss": 0.7712,
      "num_input_tokens_seen": 67241392,
      "step": 115810
    },
    {
      "epoch": 17.24977658623771,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0016952876551787266,
      "loss": 0.8151,
      "num_input_tokens_seen": 67244144,
      "step": 115815
    },
    {
      "epoch": 17.250521298778672,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.001694387405668921,
      "loss": 0.7918,
      "num_input_tokens_seen": 67247216,
      "step": 115820
    },
    {
      "epoch": 17.25126601131963,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0016934873809439132,
      "loss": 0.8221,
      "num_input_tokens_seen": 67250256,
      "step": 115825
    },
    {
      "epoch": 17.25201072386059,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0016925875810189154,
      "loss": 0.7656,
      "num_input_tokens_seen": 67253040,
      "step": 115830
    },
    {
      "epoch": 17.25275543640155,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0016916880059091243,
      "loss": 0.7908,
      "num_input_tokens_seen": 67255952,
      "step": 115835
    },
    {
      "epoch": 17.253500148942507,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.001690788655629739,
      "loss": 0.7971,
      "num_input_tokens_seen": 67258640,
      "step": 115840
    },
    {
      "epoch": 17.254244861483468,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0016898895301959492,
      "loss": 0.7897,
      "num_input_tokens_seen": 67261584,
      "step": 115845
    },
    {
      "epoch": 17.254989574024428,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0016889906296229512,
      "loss": 0.8107,
      "num_input_tokens_seen": 67264528,
      "step": 115850
    },
    {
      "epoch": 17.255734286565385,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0016880919539259292,
      "loss": 0.7869,
      "num_input_tokens_seen": 67267536,
      "step": 115855
    },
    {
      "epoch": 17.256478999106346,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0016871935031200595,
      "loss": 0.7879,
      "num_input_tokens_seen": 67270320,
      "step": 115860
    },
    {
      "epoch": 17.257223711647303,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.001686295277220527,
      "loss": 0.7855,
      "num_input_tokens_seen": 67273104,
      "step": 115865
    },
    {
      "epoch": 17.257968424188263,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0016853972762425039,
      "loss": 0.7862,
      "num_input_tokens_seen": 67276080,
      "step": 115870
    },
    {
      "epoch": 17.258713136729224,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0016844995002011618,
      "loss": 0.7941,
      "num_input_tokens_seen": 67279152,
      "step": 115875
    },
    {
      "epoch": 17.25945784927018,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0016836019491116671,
      "loss": 0.8021,
      "num_input_tokens_seen": 67282192,
      "step": 115880
    },
    {
      "epoch": 17.26020256181114,
      "grad_norm": 0.064453125,
      "learning_rate": 0.001682704622989184,
      "loss": 0.7678,
      "num_input_tokens_seen": 67284880,
      "step": 115885
    },
    {
      "epoch": 17.2609472743521,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.001681807521848872,
      "loss": 0.8194,
      "num_input_tokens_seen": 67287728,
      "step": 115890
    },
    {
      "epoch": 17.26169198689306,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.0016809106457058858,
      "loss": 0.8089,
      "num_input_tokens_seen": 67290576,
      "step": 115895
    },
    {
      "epoch": 17.26243669943402,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0016800139945753784,
      "loss": 0.8225,
      "num_input_tokens_seen": 67293808,
      "step": 115900
    },
    {
      "epoch": 17.263181411974976,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0016791175684724923,
      "loss": 0.7962,
      "num_input_tokens_seen": 67296432,
      "step": 115905
    },
    {
      "epoch": 17.263926124515937,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0016782213674123808,
      "loss": 0.8026,
      "num_input_tokens_seen": 67299344,
      "step": 115910
    },
    {
      "epoch": 17.264670837056897,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0016773253914101748,
      "loss": 0.7739,
      "num_input_tokens_seen": 67302384,
      "step": 115915
    },
    {
      "epoch": 17.265415549597854,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.001676429640481019,
      "loss": 0.809,
      "num_input_tokens_seen": 67305552,
      "step": 115920
    },
    {
      "epoch": 17.266160262138815,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0016755341146400414,
      "loss": 0.8269,
      "num_input_tokens_seen": 67308336,
      "step": 115925
    },
    {
      "epoch": 17.266904974679775,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0016746388139023747,
      "loss": 0.806,
      "num_input_tokens_seen": 67311440,
      "step": 115930
    },
    {
      "epoch": 17.267649687220732,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.0016737437382831422,
      "loss": 0.7833,
      "num_input_tokens_seen": 67314352,
      "step": 115935
    },
    {
      "epoch": 17.268394399761693,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0016728488877974612,
      "loss": 0.7739,
      "num_input_tokens_seen": 67317168,
      "step": 115940
    },
    {
      "epoch": 17.26913911230265,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0016719542624604566,
      "loss": 0.7871,
      "num_input_tokens_seen": 67319920,
      "step": 115945
    },
    {
      "epoch": 17.26988382484361,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0016710598622872346,
      "loss": 0.775,
      "num_input_tokens_seen": 67322960,
      "step": 115950
    },
    {
      "epoch": 17.27062853738457,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0016701656872929148,
      "loss": 0.8023,
      "num_input_tokens_seen": 67325776,
      "step": 115955
    },
    {
      "epoch": 17.271373249925528,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0016692717374925969,
      "loss": 0.7967,
      "num_input_tokens_seen": 67329008,
      "step": 115960
    },
    {
      "epoch": 17.27211796246649,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0016683780129013836,
      "loss": 0.7858,
      "num_input_tokens_seen": 67331760,
      "step": 115965
    },
    {
      "epoch": 17.272862675007445,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0016674845135343746,
      "loss": 0.7782,
      "num_input_tokens_seen": 67334864,
      "step": 115970
    },
    {
      "epoch": 17.273607387548406,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0016665912394066595,
      "loss": 0.7911,
      "num_input_tokens_seen": 67337840,
      "step": 115975
    },
    {
      "epoch": 17.274352100089367,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0016656981905333395,
      "loss": 0.7791,
      "num_input_tokens_seen": 67340752,
      "step": 115980
    },
    {
      "epoch": 17.275096812630323,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.001664805366929491,
      "loss": 0.7901,
      "num_input_tokens_seen": 67343856,
      "step": 115985
    },
    {
      "epoch": 17.275841525171284,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0016639127686102068,
      "loss": 0.8028,
      "num_input_tokens_seen": 67346704,
      "step": 115990
    },
    {
      "epoch": 17.276586237712245,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0016630203955905597,
      "loss": 0.7945,
      "num_input_tokens_seen": 67349360,
      "step": 115995
    },
    {
      "epoch": 17.2773309502532,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0016621282478856297,
      "loss": 0.7999,
      "num_input_tokens_seen": 67352752,
      "step": 116000
    },
    {
      "epoch": 17.278075662794162,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0016612363255104844,
      "loss": 0.8026,
      "num_input_tokens_seen": 67355440,
      "step": 116005
    },
    {
      "epoch": 17.27882037533512,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0016603446284801987,
      "loss": 0.7852,
      "num_input_tokens_seen": 67358256,
      "step": 116010
    },
    {
      "epoch": 17.27956508787608,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0016594531568098303,
      "loss": 0.7849,
      "num_input_tokens_seen": 67361072,
      "step": 116015
    },
    {
      "epoch": 17.28030980041704,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0016585619105144406,
      "loss": 0.7897,
      "num_input_tokens_seen": 67363888,
      "step": 116020
    },
    {
      "epoch": 17.281054512957997,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.0016576708896090908,
      "loss": 0.7726,
      "num_input_tokens_seen": 67366768,
      "step": 116025
    },
    {
      "epoch": 17.281799225498958,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0016567800941088306,
      "loss": 0.7956,
      "num_input_tokens_seen": 67369872,
      "step": 116030
    },
    {
      "epoch": 17.282543938039918,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0016558895240287095,
      "loss": 0.807,
      "num_input_tokens_seen": 67372944,
      "step": 116035
    },
    {
      "epoch": 17.283288650580875,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0016549991793837692,
      "loss": 0.782,
      "num_input_tokens_seen": 67376208,
      "step": 116040
    },
    {
      "epoch": 17.284033363121836,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0016541090601890573,
      "loss": 0.7833,
      "num_input_tokens_seen": 67378832,
      "step": 116045
    },
    {
      "epoch": 17.284778075662793,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0016532191664596085,
      "loss": 0.7974,
      "num_input_tokens_seen": 67381776,
      "step": 116050
    },
    {
      "epoch": 17.285522788203753,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0016523294982104541,
      "loss": 0.7982,
      "num_input_tokens_seen": 67384784,
      "step": 116055
    },
    {
      "epoch": 17.286267500744714,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0016514400554566305,
      "loss": 0.7954,
      "num_input_tokens_seen": 67387824,
      "step": 116060
    },
    {
      "epoch": 17.28701221328567,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0016505508382131555,
      "loss": 0.8011,
      "num_input_tokens_seen": 67390832,
      "step": 116065
    },
    {
      "epoch": 17.28775692582663,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0016496618464950606,
      "loss": 0.7942,
      "num_input_tokens_seen": 67393808,
      "step": 116070
    },
    {
      "epoch": 17.288501638367592,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0016487730803173572,
      "loss": 0.7947,
      "num_input_tokens_seen": 67396944,
      "step": 116075
    },
    {
      "epoch": 17.28924635090855,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0016478845396950663,
      "loss": 0.7932,
      "num_input_tokens_seen": 67399696,
      "step": 116080
    },
    {
      "epoch": 17.28999106344951,
      "grad_norm": 0.11328125,
      "learning_rate": 0.001646996224643193,
      "loss": 0.7903,
      "num_input_tokens_seen": 67402288,
      "step": 116085
    },
    {
      "epoch": 17.290735775990466,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0016461081351767496,
      "loss": 0.7874,
      "num_input_tokens_seen": 67405136,
      "step": 116090
    },
    {
      "epoch": 17.291480488531427,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0016452202713107367,
      "loss": 0.7888,
      "num_input_tokens_seen": 67407920,
      "step": 116095
    },
    {
      "epoch": 17.292225201072387,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0016443326330601532,
      "loss": 0.7843,
      "num_input_tokens_seen": 67410832,
      "step": 116100
    },
    {
      "epoch": 17.292969913613344,
      "grad_norm": 0.2001953125,
      "learning_rate": 0.0016434452204399973,
      "loss": 0.8064,
      "num_input_tokens_seen": 67413616,
      "step": 116105
    },
    {
      "epoch": 17.293714626154305,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0016425580334652556,
      "loss": 0.7958,
      "num_input_tokens_seen": 67416528,
      "step": 116110
    },
    {
      "epoch": 17.294459338695262,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0016416710721509225,
      "loss": 0.7898,
      "num_input_tokens_seen": 67419504,
      "step": 116115
    },
    {
      "epoch": 17.295204051236222,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0016407843365119777,
      "loss": 0.7828,
      "num_input_tokens_seen": 67422384,
      "step": 116120
    },
    {
      "epoch": 17.295948763777183,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0016398978265634077,
      "loss": 0.8025,
      "num_input_tokens_seen": 67425488,
      "step": 116125
    },
    {
      "epoch": 17.29669347631814,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0016390115423201839,
      "loss": 0.7926,
      "num_input_tokens_seen": 67428592,
      "step": 116130
    },
    {
      "epoch": 17.2974381888591,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0016381254837972775,
      "loss": 0.8057,
      "num_input_tokens_seen": 67431312,
      "step": 116135
    },
    {
      "epoch": 17.29818290140006,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0016372396510096648,
      "loss": 0.8132,
      "num_input_tokens_seen": 67434128,
      "step": 116140
    },
    {
      "epoch": 17.298927613941018,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0016363540439723022,
      "loss": 0.8063,
      "num_input_tokens_seen": 67436912,
      "step": 116145
    },
    {
      "epoch": 17.29967232648198,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.0016354686627001612,
      "loss": 0.7752,
      "num_input_tokens_seen": 67439760,
      "step": 116150
    },
    {
      "epoch": 17.300417039022935,
      "grad_norm": 0.091796875,
      "learning_rate": 0.00163458350720819,
      "loss": 0.7826,
      "num_input_tokens_seen": 67442672,
      "step": 116155
    },
    {
      "epoch": 17.301161751563896,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0016336985775113476,
      "loss": 0.7788,
      "num_input_tokens_seen": 67445328,
      "step": 116160
    },
    {
      "epoch": 17.301906464104857,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0016328138736245845,
      "loss": 0.7993,
      "num_input_tokens_seen": 67448176,
      "step": 116165
    },
    {
      "epoch": 17.302651176645814,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0016319293955628466,
      "loss": 0.7925,
      "num_input_tokens_seen": 67450960,
      "step": 116170
    },
    {
      "epoch": 17.303395889186774,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.001631045143341072,
      "loss": 0.7917,
      "num_input_tokens_seen": 67454000,
      "step": 116175
    },
    {
      "epoch": 17.304140601727735,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.0016301611169742007,
      "loss": 0.7926,
      "num_input_tokens_seen": 67456976,
      "step": 116180
    },
    {
      "epoch": 17.30488531426869,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0016292773164771723,
      "loss": 0.7951,
      "num_input_tokens_seen": 67459856,
      "step": 116185
    },
    {
      "epoch": 17.305630026809652,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0016283937418649097,
      "loss": 0.7949,
      "num_input_tokens_seen": 67462640,
      "step": 116190
    },
    {
      "epoch": 17.30637473935061,
      "grad_norm": 0.09375,
      "learning_rate": 0.0016275103931523492,
      "loss": 0.7921,
      "num_input_tokens_seen": 67465840,
      "step": 116195
    },
    {
      "epoch": 17.30711945189157,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.001626627270354406,
      "loss": 0.8808,
      "num_input_tokens_seen": 67468656,
      "step": 116200
    },
    {
      "epoch": 17.30786416443253,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0016257443734860044,
      "loss": 0.7998,
      "num_input_tokens_seen": 67471600,
      "step": 116205
    },
    {
      "epoch": 17.308608876973487,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0016248617025620608,
      "loss": 0.8312,
      "num_input_tokens_seen": 67474256,
      "step": 116210
    },
    {
      "epoch": 17.309353589514448,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0016239792575974802,
      "loss": 0.7969,
      "num_input_tokens_seen": 67477072,
      "step": 116215
    },
    {
      "epoch": 17.31009830205541,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.001623097038607179,
      "loss": 0.7768,
      "num_input_tokens_seen": 67479920,
      "step": 116220
    },
    {
      "epoch": 17.310843014596365,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.001622215045606057,
      "loss": 0.7845,
      "num_input_tokens_seen": 67482704,
      "step": 116225
    },
    {
      "epoch": 17.311587727137326,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0016213332786090167,
      "loss": 0.795,
      "num_input_tokens_seen": 67485584,
      "step": 116230
    },
    {
      "epoch": 17.312332439678283,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0016204517376309485,
      "loss": 0.7948,
      "num_input_tokens_seen": 67488432,
      "step": 116235
    },
    {
      "epoch": 17.313077152219243,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0016195704226867535,
      "loss": 0.798,
      "num_input_tokens_seen": 67491248,
      "step": 116240
    },
    {
      "epoch": 17.313821864760204,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0016186893337913149,
      "loss": 0.7837,
      "num_input_tokens_seen": 67494256,
      "step": 116245
    },
    {
      "epoch": 17.31456657730116,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0016178084709595226,
      "loss": 0.8004,
      "num_input_tokens_seen": 67496880,
      "step": 116250
    },
    {
      "epoch": 17.31531128984212,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0016169278342062542,
      "loss": 0.7717,
      "num_input_tokens_seen": 67499760,
      "step": 116255
    },
    {
      "epoch": 17.316056002383082,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.001616047423546385,
      "loss": 0.8096,
      "num_input_tokens_seen": 67502544,
      "step": 116260
    },
    {
      "epoch": 17.31680071492404,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0016151672389947945,
      "loss": 0.7744,
      "num_input_tokens_seen": 67505360,
      "step": 116265
    },
    {
      "epoch": 17.317545427465,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0016142872805663478,
      "loss": 0.7795,
      "num_input_tokens_seen": 67508336,
      "step": 116270
    },
    {
      "epoch": 17.318290140005956,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0016134075482759158,
      "loss": 0.7771,
      "num_input_tokens_seen": 67511376,
      "step": 116275
    },
    {
      "epoch": 17.319034852546917,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0016125280421383536,
      "loss": 0.8228,
      "num_input_tokens_seen": 67514448,
      "step": 116280
    },
    {
      "epoch": 17.319779565087877,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0016116487621685275,
      "loss": 0.8,
      "num_input_tokens_seen": 67517104,
      "step": 116285
    },
    {
      "epoch": 17.320524277628834,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0016107697083812894,
      "loss": 0.8171,
      "num_input_tokens_seen": 67519856,
      "step": 116290
    },
    {
      "epoch": 17.321268990169795,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.001609890880791487,
      "loss": 0.8041,
      "num_input_tokens_seen": 67522832,
      "step": 116295
    },
    {
      "epoch": 17.322013702710752,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.00160901227941397,
      "loss": 0.8192,
      "num_input_tokens_seen": 67525712,
      "step": 116300
    },
    {
      "epoch": 17.322758415251712,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0016081339042635784,
      "loss": 0.8008,
      "num_input_tokens_seen": 67528528,
      "step": 116305
    },
    {
      "epoch": 17.323503127792673,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0016072557553551553,
      "loss": 0.7797,
      "num_input_tokens_seen": 67531312,
      "step": 116310
    },
    {
      "epoch": 17.32424784033363,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0016063778327035305,
      "loss": 0.7981,
      "num_input_tokens_seen": 67534032,
      "step": 116315
    },
    {
      "epoch": 17.32499255287459,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0016055001363235438,
      "loss": 0.8366,
      "num_input_tokens_seen": 67536752,
      "step": 116320
    },
    {
      "epoch": 17.32573726541555,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0016046226662300167,
      "loss": 0.7818,
      "num_input_tokens_seen": 67539792,
      "step": 116325
    },
    {
      "epoch": 17.326481977956508,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0016037454224377772,
      "loss": 0.7813,
      "num_input_tokens_seen": 67542704,
      "step": 116330
    },
    {
      "epoch": 17.32722669049747,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0016028684049616438,
      "loss": 0.797,
      "num_input_tokens_seen": 67545712,
      "step": 116335
    },
    {
      "epoch": 17.327971403038426,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0016019916138164292,
      "loss": 0.7926,
      "num_input_tokens_seen": 67548592,
      "step": 116340
    },
    {
      "epoch": 17.328716115579386,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0016011150490169517,
      "loss": 0.802,
      "num_input_tokens_seen": 67551632,
      "step": 116345
    },
    {
      "epoch": 17.329460828120347,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0016002387105780145,
      "loss": 0.7954,
      "num_input_tokens_seen": 67554800,
      "step": 116350
    },
    {
      "epoch": 17.330205540661304,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0015993625985144294,
      "loss": 0.8172,
      "num_input_tokens_seen": 67557552,
      "step": 116355
    },
    {
      "epoch": 17.330950253202264,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0015984867128409923,
      "loss": 0.7903,
      "num_input_tokens_seen": 67560944,
      "step": 116360
    },
    {
      "epoch": 17.331694965743225,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0015976110535725018,
      "loss": 0.8029,
      "num_input_tokens_seen": 67563664,
      "step": 116365
    },
    {
      "epoch": 17.33243967828418,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0015967356207237475,
      "loss": 0.8139,
      "num_input_tokens_seen": 67566320,
      "step": 116370
    },
    {
      "epoch": 17.333184390825142,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.001595860414309526,
      "loss": 0.8179,
      "num_input_tokens_seen": 67569136,
      "step": 116375
    },
    {
      "epoch": 17.3339291033661,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.001594985434344619,
      "loss": 0.8097,
      "num_input_tokens_seen": 67571792,
      "step": 116380
    },
    {
      "epoch": 17.33467381590706,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0015941106808438077,
      "loss": 0.8105,
      "num_input_tokens_seen": 67574832,
      "step": 116385
    },
    {
      "epoch": 17.33541852844802,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.001593236153821872,
      "loss": 0.8065,
      "num_input_tokens_seen": 67577904,
      "step": 116390
    },
    {
      "epoch": 17.336163240988977,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0015923618532935834,
      "loss": 0.7985,
      "num_input_tokens_seen": 67580816,
      "step": 116395
    },
    {
      "epoch": 17.336907953529938,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0015914877792737169,
      "loss": 0.8031,
      "num_input_tokens_seen": 67583760,
      "step": 116400
    },
    {
      "epoch": 17.3376526660709,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.001590613931777034,
      "loss": 0.8138,
      "num_input_tokens_seen": 67586960,
      "step": 116405
    },
    {
      "epoch": 17.338397378611855,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0015897403108183028,
      "loss": 0.7923,
      "num_input_tokens_seen": 67589776,
      "step": 116410
    },
    {
      "epoch": 17.339142091152816,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0015888669164122798,
      "loss": 0.8061,
      "num_input_tokens_seen": 67592592,
      "step": 116415
    },
    {
      "epoch": 17.339886803693773,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0015879937485737167,
      "loss": 0.7937,
      "num_input_tokens_seen": 67595760,
      "step": 116420
    },
    {
      "epoch": 17.340631516234733,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0015871208073173714,
      "loss": 0.7906,
      "num_input_tokens_seen": 67598864,
      "step": 116425
    },
    {
      "epoch": 17.341376228775694,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0015862480926579874,
      "loss": 0.8041,
      "num_input_tokens_seen": 67601744,
      "step": 116430
    },
    {
      "epoch": 17.34212094131665,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0015853756046103078,
      "loss": 0.7888,
      "num_input_tokens_seen": 67604592,
      "step": 116435
    },
    {
      "epoch": 17.34286565385761,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0015845033431890708,
      "loss": 0.783,
      "num_input_tokens_seen": 67607376,
      "step": 116440
    },
    {
      "epoch": 17.343610366398572,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.001583631308409018,
      "loss": 0.7892,
      "num_input_tokens_seen": 67609872,
      "step": 116445
    },
    {
      "epoch": 17.34435507893953,
      "grad_norm": 0.07421875,
      "learning_rate": 0.001582759500284876,
      "loss": 0.7867,
      "num_input_tokens_seen": 67613456,
      "step": 116450
    },
    {
      "epoch": 17.34509979148049,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0015818879188313794,
      "loss": 0.7977,
      "num_input_tokens_seen": 67616336,
      "step": 116455
    },
    {
      "epoch": 17.345844504021446,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0015810165640632466,
      "loss": 0.8121,
      "num_input_tokens_seen": 67618992,
      "step": 116460
    },
    {
      "epoch": 17.346589216562407,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0015801454359951994,
      "loss": 0.8266,
      "num_input_tokens_seen": 67622032,
      "step": 116465
    },
    {
      "epoch": 17.347333929103367,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0015792745346419573,
      "loss": 0.7855,
      "num_input_tokens_seen": 67624784,
      "step": 116470
    },
    {
      "epoch": 17.348078641644324,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0015784038600182304,
      "loss": 0.7708,
      "num_input_tokens_seen": 67627856,
      "step": 116475
    },
    {
      "epoch": 17.348823354185285,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0015775334121387301,
      "loss": 0.793,
      "num_input_tokens_seen": 67630544,
      "step": 116480
    },
    {
      "epoch": 17.349568066726242,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0015766631910181582,
      "loss": 0.7865,
      "num_input_tokens_seen": 67633360,
      "step": 116485
    },
    {
      "epoch": 17.350312779267203,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.001575793196671223,
      "loss": 0.7835,
      "num_input_tokens_seen": 67636240,
      "step": 116490
    },
    {
      "epoch": 17.351057491808163,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0015749234291126174,
      "loss": 0.8409,
      "num_input_tokens_seen": 67639376,
      "step": 116495
    },
    {
      "epoch": 17.35180220434912,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0015740538883570365,
      "loss": 0.8168,
      "num_input_tokens_seen": 67642352,
      "step": 116500
    },
    {
      "epoch": 17.35254691689008,
      "grad_norm": 0.046875,
      "learning_rate": 0.0015731845744191685,
      "loss": 0.804,
      "num_input_tokens_seen": 67645136,
      "step": 116505
    },
    {
      "epoch": 17.35329162943104,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0015723154873136984,
      "loss": 0.8058,
      "num_input_tokens_seen": 67647888,
      "step": 116510
    },
    {
      "epoch": 17.354036341971998,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0015714466270553146,
      "loss": 0.8062,
      "num_input_tokens_seen": 67651184,
      "step": 116515
    },
    {
      "epoch": 17.35478105451296,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0015705779936586867,
      "loss": 0.8089,
      "num_input_tokens_seen": 67654448,
      "step": 116520
    },
    {
      "epoch": 17.355525767053916,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.0015697095871384996,
      "loss": 0.791,
      "num_input_tokens_seen": 67657232,
      "step": 116525
    },
    {
      "epoch": 17.356270479594876,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0015688414075094154,
      "loss": 0.7833,
      "num_input_tokens_seen": 67659888,
      "step": 116530
    },
    {
      "epoch": 17.357015192135837,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0015679734547861068,
      "loss": 0.7877,
      "num_input_tokens_seen": 67662544,
      "step": 116535
    },
    {
      "epoch": 17.357759904676794,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0015671057289832356,
      "loss": 0.8088,
      "num_input_tokens_seen": 67665552,
      "step": 116540
    },
    {
      "epoch": 17.358504617217754,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0015662382301154586,
      "loss": 0.8003,
      "num_input_tokens_seen": 67668688,
      "step": 116545
    },
    {
      "epoch": 17.359249329758715,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0015653709581974356,
      "loss": 0.8032,
      "num_input_tokens_seen": 67671536,
      "step": 116550
    },
    {
      "epoch": 17.35999404229967,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0015645039132438148,
      "loss": 0.7938,
      "num_input_tokens_seen": 67674288,
      "step": 116555
    },
    {
      "epoch": 17.360738754840632,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0015636370952692464,
      "loss": 0.7915,
      "num_input_tokens_seen": 67677008,
      "step": 116560
    },
    {
      "epoch": 17.36148346738159,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.00156277050428837,
      "loss": 0.7914,
      "num_input_tokens_seen": 67679664,
      "step": 116565
    },
    {
      "epoch": 17.36222817992255,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0015619041403158323,
      "loss": 0.7887,
      "num_input_tokens_seen": 67682544,
      "step": 116570
    },
    {
      "epoch": 17.36297289246351,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0015610380033662652,
      "loss": 0.7788,
      "num_input_tokens_seen": 67685616,
      "step": 116575
    },
    {
      "epoch": 17.363717605004467,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0015601720934543,
      "loss": 0.7986,
      "num_input_tokens_seen": 67688688,
      "step": 116580
    },
    {
      "epoch": 17.364462317545428,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0015593064105945702,
      "loss": 0.789,
      "num_input_tokens_seen": 67691472,
      "step": 116585
    },
    {
      "epoch": 17.36520703008639,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0015584409548016957,
      "loss": 0.794,
      "num_input_tokens_seen": 67694352,
      "step": 116590
    },
    {
      "epoch": 17.365951742627345,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0015575757260903017,
      "loss": 0.8001,
      "num_input_tokens_seen": 67697200,
      "step": 116595
    },
    {
      "epoch": 17.366696455168306,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0015567107244750028,
      "loss": 0.8117,
      "num_input_tokens_seen": 67699952,
      "step": 116600
    },
    {
      "epoch": 17.367441167709263,
      "grad_norm": 0.0625,
      "learning_rate": 0.0015558459499704141,
      "loss": 0.8052,
      "num_input_tokens_seen": 67702928,
      "step": 116605
    },
    {
      "epoch": 17.368185880250223,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0015549814025911423,
      "loss": 0.798,
      "num_input_tokens_seen": 67705872,
      "step": 116610
    },
    {
      "epoch": 17.368930592791184,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0015541170823517958,
      "loss": 0.8028,
      "num_input_tokens_seen": 67708528,
      "step": 116615
    },
    {
      "epoch": 17.36967530533214,
      "grad_norm": 0.060546875,
      "learning_rate": 0.0015532529892669778,
      "loss": 0.8019,
      "num_input_tokens_seen": 67711568,
      "step": 116620
    },
    {
      "epoch": 17.3704200178731,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0015523891233512814,
      "loss": 0.823,
      "num_input_tokens_seen": 67714480,
      "step": 116625
    },
    {
      "epoch": 17.37116473041406,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0015515254846193054,
      "loss": 0.7963,
      "num_input_tokens_seen": 67717296,
      "step": 116630
    },
    {
      "epoch": 17.37190944295502,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0015506620730856345,
      "loss": 0.7946,
      "num_input_tokens_seen": 67720144,
      "step": 116635
    },
    {
      "epoch": 17.37265415549598,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0015497988887648606,
      "loss": 0.7944,
      "num_input_tokens_seen": 67723056,
      "step": 116640
    },
    {
      "epoch": 17.373398868036936,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0015489359316715617,
      "loss": 0.7898,
      "num_input_tokens_seen": 67725936,
      "step": 116645
    },
    {
      "epoch": 17.374143580577897,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.001548073201820323,
      "loss": 0.7933,
      "num_input_tokens_seen": 67728880,
      "step": 116650
    },
    {
      "epoch": 17.374888293118858,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0015472106992257145,
      "loss": 0.7857,
      "num_input_tokens_seen": 67731856,
      "step": 116655
    },
    {
      "epoch": 17.375633005659815,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0015463484239023062,
      "loss": 0.7827,
      "num_input_tokens_seen": 67734448,
      "step": 116660
    },
    {
      "epoch": 17.376377718200775,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0015454863758646714,
      "loss": 0.7967,
      "num_input_tokens_seen": 67737360,
      "step": 116665
    },
    {
      "epoch": 17.377122430741732,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0015446245551273652,
      "loss": 0.8116,
      "num_input_tokens_seen": 67740432,
      "step": 116670
    },
    {
      "epoch": 17.377867143282693,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0015437629617049541,
      "loss": 0.8042,
      "num_input_tokens_seen": 67743344,
      "step": 116675
    },
    {
      "epoch": 17.378611855823653,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00154290159561199,
      "loss": 0.793,
      "num_input_tokens_seen": 67746128,
      "step": 116680
    },
    {
      "epoch": 17.37935656836461,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0015420404568630296,
      "loss": 0.7872,
      "num_input_tokens_seen": 67749200,
      "step": 116685
    },
    {
      "epoch": 17.38010128090557,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.0015411795454726162,
      "loss": 0.7796,
      "num_input_tokens_seen": 67751920,
      "step": 116690
    },
    {
      "epoch": 17.38084599344653,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0015403188614552964,
      "loss": 0.8077,
      "num_input_tokens_seen": 67754864,
      "step": 116695
    },
    {
      "epoch": 17.381590705987488,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0015394584048256086,
      "loss": 0.7921,
      "num_input_tokens_seen": 67757744,
      "step": 116700
    },
    {
      "epoch": 17.38233541852845,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0015385981755980881,
      "loss": 0.83,
      "num_input_tokens_seen": 67760400,
      "step": 116705
    },
    {
      "epoch": 17.383080131069406,
      "grad_norm": 0.17578125,
      "learning_rate": 0.001537738173787273,
      "loss": 0.8011,
      "num_input_tokens_seen": 67763280,
      "step": 116710
    },
    {
      "epoch": 17.383824843610366,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0015368783994076868,
      "loss": 0.7813,
      "num_input_tokens_seen": 67766064,
      "step": 116715
    },
    {
      "epoch": 17.384569556151327,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.0015360188524738577,
      "loss": 0.7927,
      "num_input_tokens_seen": 67768752,
      "step": 116720
    },
    {
      "epoch": 17.385314268692284,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0015351595330003042,
      "loss": 0.8047,
      "num_input_tokens_seen": 67771728,
      "step": 116725
    },
    {
      "epoch": 17.386058981233244,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0015343004410015482,
      "loss": 0.7813,
      "num_input_tokens_seen": 67774864,
      "step": 116730
    },
    {
      "epoch": 17.386803693774205,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.001533441576492101,
      "loss": 0.7866,
      "num_input_tokens_seen": 67777968,
      "step": 116735
    },
    {
      "epoch": 17.38754840631516,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0015325829394864665,
      "loss": 0.7907,
      "num_input_tokens_seen": 67781232,
      "step": 116740
    },
    {
      "epoch": 17.388293118856122,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.001531724529999161,
      "loss": 0.7809,
      "num_input_tokens_seen": 67784112,
      "step": 116745
    },
    {
      "epoch": 17.38903783139708,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0015308663480446765,
      "loss": 0.7956,
      "num_input_tokens_seen": 67787120,
      "step": 116750
    },
    {
      "epoch": 17.38978254393804,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0015300083936375197,
      "loss": 0.7815,
      "num_input_tokens_seen": 67790032,
      "step": 116755
    },
    {
      "epoch": 17.390527256479,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.001529150666792179,
      "loss": 0.8092,
      "num_input_tokens_seen": 67793264,
      "step": 116760
    },
    {
      "epoch": 17.391271969019957,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0015282931675231475,
      "loss": 0.785,
      "num_input_tokens_seen": 67796016,
      "step": 116765
    },
    {
      "epoch": 17.392016681560918,
      "grad_norm": 0.06640625,
      "learning_rate": 0.001527435895844909,
      "loss": 0.8036,
      "num_input_tokens_seen": 67798896,
      "step": 116770
    },
    {
      "epoch": 17.39276139410188,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.0015265788517719503,
      "loss": 0.803,
      "num_input_tokens_seen": 67801680,
      "step": 116775
    },
    {
      "epoch": 17.393506106642835,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0015257220353187477,
      "loss": 0.7857,
      "num_input_tokens_seen": 67804592,
      "step": 116780
    },
    {
      "epoch": 17.394250819183796,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0015248654464997751,
      "loss": 0.7842,
      "num_input_tokens_seen": 67807632,
      "step": 116785
    },
    {
      "epoch": 17.394995531724753,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.0015240090853295073,
      "loss": 0.789,
      "num_input_tokens_seen": 67810512,
      "step": 116790
    },
    {
      "epoch": 17.395740244265713,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0015231529518224063,
      "loss": 0.8029,
      "num_input_tokens_seen": 67813168,
      "step": 116795
    },
    {
      "epoch": 17.396484956806674,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0015222970459929418,
      "loss": 0.7896,
      "num_input_tokens_seen": 67815856,
      "step": 116800
    },
    {
      "epoch": 17.39722966934763,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0015214413678555677,
      "loss": 0.781,
      "num_input_tokens_seen": 67818576,
      "step": 116805
    },
    {
      "epoch": 17.39797438188859,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0015205859174247471,
      "loss": 0.7852,
      "num_input_tokens_seen": 67821200,
      "step": 116810
    },
    {
      "epoch": 17.39871909442955,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0015197306947149253,
      "loss": 0.7955,
      "num_input_tokens_seen": 67823984,
      "step": 116815
    },
    {
      "epoch": 17.39946380697051,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0015188756997405505,
      "loss": 0.7939,
      "num_input_tokens_seen": 67826704,
      "step": 116820
    },
    {
      "epoch": 17.40020851951147,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0015180209325160732,
      "loss": 0.8184,
      "num_input_tokens_seen": 67829712,
      "step": 116825
    },
    {
      "epoch": 17.400953232052427,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0015171663930559297,
      "loss": 0.7813,
      "num_input_tokens_seen": 67832528,
      "step": 116830
    },
    {
      "epoch": 17.401697944593387,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0015163120813745557,
      "loss": 0.786,
      "num_input_tokens_seen": 67835440,
      "step": 116835
    },
    {
      "epoch": 17.402442657134348,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0015154579974863823,
      "loss": 0.7801,
      "num_input_tokens_seen": 67838768,
      "step": 116840
    },
    {
      "epoch": 17.403187369675305,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0015146041414058435,
      "loss": 0.8036,
      "num_input_tokens_seen": 67841776,
      "step": 116845
    },
    {
      "epoch": 17.403932082216265,
      "grad_norm": 0.02783203125,
      "learning_rate": 0.0015137505131473594,
      "loss": 0.8003,
      "num_input_tokens_seen": 67844368,
      "step": 116850
    },
    {
      "epoch": 17.404676794757222,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0015128971127253564,
      "loss": 0.8077,
      "num_input_tokens_seen": 67847376,
      "step": 116855
    },
    {
      "epoch": 17.405421507298183,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0015120439401542485,
      "loss": 0.8252,
      "num_input_tokens_seen": 67850096,
      "step": 116860
    },
    {
      "epoch": 17.406166219839143,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.001511190995448447,
      "loss": 0.7728,
      "num_input_tokens_seen": 67852720,
      "step": 116865
    },
    {
      "epoch": 17.4069109323801,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0015103382786223673,
      "loss": 0.8058,
      "num_input_tokens_seen": 67855376,
      "step": 116870
    },
    {
      "epoch": 17.40765564492106,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0015094857896904095,
      "loss": 0.8077,
      "num_input_tokens_seen": 67858320,
      "step": 116875
    },
    {
      "epoch": 17.40840035746202,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0015086335286669805,
      "loss": 0.7765,
      "num_input_tokens_seen": 67861168,
      "step": 116880
    },
    {
      "epoch": 17.409145070002978,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0015077814955664754,
      "loss": 0.8007,
      "num_input_tokens_seen": 67863920,
      "step": 116885
    },
    {
      "epoch": 17.40988978254394,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0015069296904032907,
      "loss": 0.7951,
      "num_input_tokens_seen": 67866608,
      "step": 116890
    },
    {
      "epoch": 17.410634495084896,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0015060781131918104,
      "loss": 0.7945,
      "num_input_tokens_seen": 67869616,
      "step": 116895
    },
    {
      "epoch": 17.411379207625856,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0015052267639464312,
      "loss": 0.7984,
      "num_input_tokens_seen": 67872464,
      "step": 116900
    },
    {
      "epoch": 17.412123920166817,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0015043756426815279,
      "loss": 0.8183,
      "num_input_tokens_seen": 67875088,
      "step": 116905
    },
    {
      "epoch": 17.412868632707774,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0015035247494114795,
      "loss": 0.8032,
      "num_input_tokens_seen": 67878320,
      "step": 116910
    },
    {
      "epoch": 17.413613345248734,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0015026740841506674,
      "loss": 0.8009,
      "num_input_tokens_seen": 67881488,
      "step": 116915
    },
    {
      "epoch": 17.414358057789695,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0015018236469134538,
      "loss": 0.7832,
      "num_input_tokens_seen": 67884400,
      "step": 116920
    },
    {
      "epoch": 17.415102770330652,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0015009734377142152,
      "loss": 0.8053,
      "num_input_tokens_seen": 67887504,
      "step": 116925
    },
    {
      "epoch": 17.415847482871612,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.001500123456567307,
      "loss": 0.7951,
      "num_input_tokens_seen": 67890512,
      "step": 116930
    },
    {
      "epoch": 17.41659219541257,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0014992737034870945,
      "loss": 0.8095,
      "num_input_tokens_seen": 67893712,
      "step": 116935
    },
    {
      "epoch": 17.41733690795353,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0014984241784879325,
      "loss": 0.7883,
      "num_input_tokens_seen": 67896176,
      "step": 116940
    },
    {
      "epoch": 17.41808162049449,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0014975748815841683,
      "loss": 0.7929,
      "num_input_tokens_seen": 67899248,
      "step": 116945
    },
    {
      "epoch": 17.418826333035447,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0014967258127901568,
      "loss": 0.8068,
      "num_input_tokens_seen": 67902032,
      "step": 116950
    },
    {
      "epoch": 17.419571045576408,
      "grad_norm": 0.1484375,
      "learning_rate": 0.001495876972120238,
      "loss": 0.8195,
      "num_input_tokens_seen": 67905168,
      "step": 116955
    },
    {
      "epoch": 17.42031575811737,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0014950283595887525,
      "loss": 0.7786,
      "num_input_tokens_seen": 67907856,
      "step": 116960
    },
    {
      "epoch": 17.421060470658325,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0014941799752100338,
      "loss": 0.8063,
      "num_input_tokens_seen": 67910704,
      "step": 116965
    },
    {
      "epoch": 17.421805183199286,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0014933318189984217,
      "loss": 0.8299,
      "num_input_tokens_seen": 67913328,
      "step": 116970
    },
    {
      "epoch": 17.422549895740243,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0014924838909682369,
      "loss": 0.7963,
      "num_input_tokens_seen": 67916144,
      "step": 116975
    },
    {
      "epoch": 17.423294608281203,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.001491636191133811,
      "loss": 0.8049,
      "num_input_tokens_seen": 67918832,
      "step": 116980
    },
    {
      "epoch": 17.424039320822164,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0014907887195094642,
      "loss": 0.8118,
      "num_input_tokens_seen": 67921872,
      "step": 116985
    },
    {
      "epoch": 17.42478403336312,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.001489941476109507,
      "loss": 0.7944,
      "num_input_tokens_seen": 67924688,
      "step": 116990
    },
    {
      "epoch": 17.42552874590408,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0014890944609482625,
      "loss": 0.8044,
      "num_input_tokens_seen": 67927472,
      "step": 116995
    },
    {
      "epoch": 17.42627345844504,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0014882476740400313,
      "loss": 0.7863,
      "num_input_tokens_seen": 67930512,
      "step": 117000
    },
    {
      "epoch": 17.427018170986,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0014874011153991251,
      "loss": 0.807,
      "num_input_tokens_seen": 67933392,
      "step": 117005
    },
    {
      "epoch": 17.42776288352696,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0014865547850398408,
      "loss": 0.8018,
      "num_input_tokens_seen": 67936400,
      "step": 117010
    },
    {
      "epoch": 17.428507596067917,
      "grad_norm": 0.08984375,
      "learning_rate": 0.001485708682976482,
      "loss": 0.8003,
      "num_input_tokens_seen": 67939600,
      "step": 117015
    },
    {
      "epoch": 17.429252308608877,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0014848628092233407,
      "loss": 0.7926,
      "num_input_tokens_seen": 67942480,
      "step": 117020
    },
    {
      "epoch": 17.429997021149838,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0014840171637947052,
      "loss": 0.8092,
      "num_input_tokens_seen": 67945328,
      "step": 117025
    },
    {
      "epoch": 17.430741733690795,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0014831717467048626,
      "loss": 0.7958,
      "num_input_tokens_seen": 67948080,
      "step": 117030
    },
    {
      "epoch": 17.431486446231755,
      "grad_norm": 0.3984375,
      "learning_rate": 0.0014823265579680916,
      "loss": 0.8404,
      "num_input_tokens_seen": 67950672,
      "step": 117035
    },
    {
      "epoch": 17.432231158772712,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0014814815975986789,
      "loss": 0.8183,
      "num_input_tokens_seen": 67953648,
      "step": 117040
    },
    {
      "epoch": 17.432975871313673,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0014806368656108914,
      "loss": 0.7787,
      "num_input_tokens_seen": 67956592,
      "step": 117045
    },
    {
      "epoch": 17.433720583854633,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0014797923620190078,
      "loss": 0.7962,
      "num_input_tokens_seen": 67959504,
      "step": 117050
    },
    {
      "epoch": 17.43446529639559,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0014789480868372866,
      "loss": 0.8199,
      "num_input_tokens_seen": 67962320,
      "step": 117055
    },
    {
      "epoch": 17.43521000893655,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00147810404008,
      "loss": 0.7941,
      "num_input_tokens_seen": 67965040,
      "step": 117060
    },
    {
      "epoch": 17.43595472147751,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0014772602217614011,
      "loss": 0.8074,
      "num_input_tokens_seen": 67967760,
      "step": 117065
    },
    {
      "epoch": 17.43669943401847,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0014764166318957438,
      "loss": 0.7952,
      "num_input_tokens_seen": 67970736,
      "step": 117070
    },
    {
      "epoch": 17.43744414655943,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0014755732704972851,
      "loss": 0.815,
      "num_input_tokens_seen": 67973712,
      "step": 117075
    },
    {
      "epoch": 17.438188859100386,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0014747301375802684,
      "loss": 0.7843,
      "num_input_tokens_seen": 67976368,
      "step": 117080
    },
    {
      "epoch": 17.438933571641346,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.0014738872331589426,
      "loss": 0.7965,
      "num_input_tokens_seen": 67978928,
      "step": 117085
    },
    {
      "epoch": 17.439678284182307,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0014730445572475443,
      "loss": 0.8082,
      "num_input_tokens_seen": 67981584,
      "step": 117090
    },
    {
      "epoch": 17.440422996723264,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0014722021098603104,
      "loss": 0.8094,
      "num_input_tokens_seen": 67984464,
      "step": 117095
    },
    {
      "epoch": 17.441167709264224,
      "grad_norm": 0.060546875,
      "learning_rate": 0.0014713598910114734,
      "loss": 0.8156,
      "num_input_tokens_seen": 67987344,
      "step": 117100
    },
    {
      "epoch": 17.441912421805185,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.001470517900715258,
      "loss": 0.8831,
      "num_input_tokens_seen": 67990448,
      "step": 117105
    },
    {
      "epoch": 17.442657134346142,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0014696761389858932,
      "loss": 0.8194,
      "num_input_tokens_seen": 67993360,
      "step": 117110
    },
    {
      "epoch": 17.443401846887102,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0014688346058375971,
      "loss": 0.8131,
      "num_input_tokens_seen": 67996048,
      "step": 117115
    },
    {
      "epoch": 17.44414655942806,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0014679933012845907,
      "loss": 0.7964,
      "num_input_tokens_seen": 67998992,
      "step": 117120
    },
    {
      "epoch": 17.44489127196902,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0014671522253410806,
      "loss": 0.7877,
      "num_input_tokens_seen": 68002192,
      "step": 117125
    },
    {
      "epoch": 17.44563598450998,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0014663113780212821,
      "loss": 0.7796,
      "num_input_tokens_seen": 68005072,
      "step": 117130
    },
    {
      "epoch": 17.446380697050937,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0014654707593393956,
      "loss": 0.7864,
      "num_input_tokens_seen": 68007984,
      "step": 117135
    },
    {
      "epoch": 17.447125409591898,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.001464630369309628,
      "loss": 0.7945,
      "num_input_tokens_seen": 68010832,
      "step": 117140
    },
    {
      "epoch": 17.447870122132855,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.0014637902079461728,
      "loss": 0.7996,
      "num_input_tokens_seen": 68013520,
      "step": 117145
    },
    {
      "epoch": 17.448614834673815,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0014629502752632223,
      "loss": 0.7753,
      "num_input_tokens_seen": 68016304,
      "step": 117150
    },
    {
      "epoch": 17.449359547214776,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0014621105712749715,
      "loss": 0.7848,
      "num_input_tokens_seen": 68019376,
      "step": 117155
    },
    {
      "epoch": 17.450104259755733,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0014612710959956026,
      "loss": 0.8104,
      "num_input_tokens_seen": 68022192,
      "step": 117160
    },
    {
      "epoch": 17.450848972296694,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.0014604318494392976,
      "loss": 0.7805,
      "num_input_tokens_seen": 68025264,
      "step": 117165
    },
    {
      "epoch": 17.451593684837654,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0014595928316202332,
      "loss": 0.79,
      "num_input_tokens_seen": 68027984,
      "step": 117170
    },
    {
      "epoch": 17.45233839737861,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0014587540425525902,
      "loss": 0.8191,
      "num_input_tokens_seen": 68030672,
      "step": 117175
    },
    {
      "epoch": 17.45308310991957,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0014579154822505335,
      "loss": 0.8073,
      "num_input_tokens_seen": 68033392,
      "step": 117180
    },
    {
      "epoch": 17.45382782246053,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0014570771507282287,
      "loss": 0.7964,
      "num_input_tokens_seen": 68036272,
      "step": 117185
    },
    {
      "epoch": 17.45457253500149,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0014562390479998443,
      "loss": 0.8053,
      "num_input_tokens_seen": 68038960,
      "step": 117190
    },
    {
      "epoch": 17.45531724754245,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0014554011740795325,
      "loss": 0.7909,
      "num_input_tokens_seen": 68041904,
      "step": 117195
    },
    {
      "epoch": 17.456061960083407,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.001454563528981455,
      "loss": 0.7895,
      "num_input_tokens_seen": 68044752,
      "step": 117200
    },
    {
      "epoch": 17.456806672624367,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0014537261127197559,
      "loss": 0.8113,
      "num_input_tokens_seen": 68047440,
      "step": 117205
    },
    {
      "epoch": 17.457551385165328,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0014528889253085902,
      "loss": 0.8056,
      "num_input_tokens_seen": 68050448,
      "step": 117210
    },
    {
      "epoch": 17.458296097706285,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0014520519667620983,
      "loss": 0.7933,
      "num_input_tokens_seen": 68053328,
      "step": 117215
    },
    {
      "epoch": 17.459040810247245,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0014512152370944175,
      "loss": 0.8067,
      "num_input_tokens_seen": 68056048,
      "step": 117220
    },
    {
      "epoch": 17.459785522788202,
      "grad_norm": 0.03515625,
      "learning_rate": 0.001450378736319683,
      "loss": 0.7967,
      "num_input_tokens_seen": 68058640,
      "step": 117225
    },
    {
      "epoch": 17.460530235329163,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0014495424644520298,
      "loss": 0.791,
      "num_input_tokens_seen": 68061584,
      "step": 117230
    },
    {
      "epoch": 17.461274947870123,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0014487064215055856,
      "loss": 0.7925,
      "num_input_tokens_seen": 68064432,
      "step": 117235
    },
    {
      "epoch": 17.46201966041108,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.0014478706074944703,
      "loss": 0.7843,
      "num_input_tokens_seen": 68067312,
      "step": 117240
    },
    {
      "epoch": 17.46276437295204,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.001447035022432811,
      "loss": 0.7962,
      "num_input_tokens_seen": 68070064,
      "step": 117245
    },
    {
      "epoch": 17.463509085493,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0014461996663347166,
      "loss": 0.7915,
      "num_input_tokens_seen": 68073136,
      "step": 117250
    },
    {
      "epoch": 17.46425379803396,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.001445364539214307,
      "loss": 0.8006,
      "num_input_tokens_seen": 68076304,
      "step": 117255
    },
    {
      "epoch": 17.46499851057492,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0014445296410856867,
      "loss": 0.795,
      "num_input_tokens_seen": 68079184,
      "step": 117260
    },
    {
      "epoch": 17.465743223115876,
      "grad_norm": 0.13671875,
      "learning_rate": 0.001443694971962957,
      "loss": 0.7985,
      "num_input_tokens_seen": 68082064,
      "step": 117265
    },
    {
      "epoch": 17.466487935656836,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.001442860531860227,
      "loss": 0.7952,
      "num_input_tokens_seen": 68085232,
      "step": 117270
    },
    {
      "epoch": 17.467232648197797,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0014420263207915872,
      "loss": 0.7813,
      "num_input_tokens_seen": 68088176,
      "step": 117275
    },
    {
      "epoch": 17.467977360738754,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0014411923387711344,
      "loss": 0.7943,
      "num_input_tokens_seen": 68091504,
      "step": 117280
    },
    {
      "epoch": 17.468722073279714,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0014403585858129575,
      "loss": 0.8162,
      "num_input_tokens_seen": 68094224,
      "step": 117285
    },
    {
      "epoch": 17.469466785820675,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.00143952506193114,
      "loss": 0.7805,
      "num_input_tokens_seen": 68097456,
      "step": 117290
    },
    {
      "epoch": 17.470211498361632,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0014386917671397625,
      "loss": 0.8035,
      "num_input_tokens_seen": 68100336,
      "step": 117295
    },
    {
      "epoch": 17.470956210902592,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.001437858701452907,
      "loss": 0.8047,
      "num_input_tokens_seen": 68103184,
      "step": 117300
    },
    {
      "epoch": 17.47170092344355,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.001437025864884644,
      "loss": 0.7768,
      "num_input_tokens_seen": 68105872,
      "step": 117305
    },
    {
      "epoch": 17.47244563598451,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.001436193257449042,
      "loss": 0.7995,
      "num_input_tokens_seen": 68108656,
      "step": 117310
    },
    {
      "epoch": 17.47319034852547,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0014353608791601719,
      "loss": 0.8169,
      "num_input_tokens_seen": 68111376,
      "step": 117315
    },
    {
      "epoch": 17.473935061066427,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0014345287300320919,
      "loss": 0.7955,
      "num_input_tokens_seen": 68114384,
      "step": 117320
    },
    {
      "epoch": 17.474679773607388,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0014336968100788626,
      "loss": 0.8084,
      "num_input_tokens_seen": 68117072,
      "step": 117325
    },
    {
      "epoch": 17.47542448614835,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0014328651193145348,
      "loss": 0.7797,
      "num_input_tokens_seen": 68120080,
      "step": 117330
    },
    {
      "epoch": 17.476169198689306,
      "grad_norm": 0.05859375,
      "learning_rate": 0.001432033657753165,
      "loss": 0.8118,
      "num_input_tokens_seen": 68122992,
      "step": 117335
    },
    {
      "epoch": 17.476913911230266,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0014312024254087973,
      "loss": 0.7774,
      "num_input_tokens_seen": 68125680,
      "step": 117340
    },
    {
      "epoch": 17.477658623771223,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0014303714222954705,
      "loss": 0.791,
      "num_input_tokens_seen": 68128944,
      "step": 117345
    },
    {
      "epoch": 17.478403336312184,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0014295406484272299,
      "loss": 0.7987,
      "num_input_tokens_seen": 68131824,
      "step": 117350
    },
    {
      "epoch": 17.479148048853144,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0014287101038181077,
      "loss": 0.8006,
      "num_input_tokens_seen": 68134832,
      "step": 117355
    },
    {
      "epoch": 17.4798927613941,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0014278797884821359,
      "loss": 0.7867,
      "num_input_tokens_seen": 68137648,
      "step": 117360
    },
    {
      "epoch": 17.48063747393506,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0014270497024333367,
      "loss": 0.8,
      "num_input_tokens_seen": 68140368,
      "step": 117365
    },
    {
      "epoch": 17.48138218647602,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0014262198456857405,
      "loss": 0.78,
      "num_input_tokens_seen": 68143440,
      "step": 117370
    },
    {
      "epoch": 17.48212689901698,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0014253902182533628,
      "loss": 0.8063,
      "num_input_tokens_seen": 68146064,
      "step": 117375
    },
    {
      "epoch": 17.48287161155794,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0014245608201502223,
      "loss": 0.7945,
      "num_input_tokens_seen": 68148944,
      "step": 117380
    },
    {
      "epoch": 17.483616324098897,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0014237316513903313,
      "loss": 0.7927,
      "num_input_tokens_seen": 68151984,
      "step": 117385
    },
    {
      "epoch": 17.484361036639857,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0014229027119876901,
      "loss": 0.7984,
      "num_input_tokens_seen": 68155120,
      "step": 117390
    },
    {
      "epoch": 17.485105749180818,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0014220740019563127,
      "loss": 0.7833,
      "num_input_tokens_seen": 68157904,
      "step": 117395
    },
    {
      "epoch": 17.485850461721775,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0014212455213101925,
      "loss": 0.7854,
      "num_input_tokens_seen": 68160528,
      "step": 117400
    },
    {
      "epoch": 17.486595174262735,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0014204172700633321,
      "loss": 0.8182,
      "num_input_tokens_seen": 68163152,
      "step": 117405
    },
    {
      "epoch": 17.487339886803692,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.001419589248229715,
      "loss": 0.7969,
      "num_input_tokens_seen": 68166064,
      "step": 117410
    },
    {
      "epoch": 17.488084599344653,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0014187614558233402,
      "loss": 0.7883,
      "num_input_tokens_seen": 68168848,
      "step": 117415
    },
    {
      "epoch": 17.488829311885613,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0014179338928581864,
      "loss": 0.8138,
      "num_input_tokens_seen": 68172112,
      "step": 117420
    },
    {
      "epoch": 17.48957402442657,
      "grad_norm": 0.064453125,
      "learning_rate": 0.001417106559348234,
      "loss": 0.7769,
      "num_input_tokens_seen": 68174768,
      "step": 117425
    },
    {
      "epoch": 17.49031873696753,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.001416279455307462,
      "loss": 0.7976,
      "num_input_tokens_seen": 68177680,
      "step": 117430
    },
    {
      "epoch": 17.49106344950849,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0014154525807498408,
      "loss": 0.7964,
      "num_input_tokens_seen": 68180656,
      "step": 117435
    },
    {
      "epoch": 17.49180816204945,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0014146259356893442,
      "loss": 0.8026,
      "num_input_tokens_seen": 68183504,
      "step": 117440
    },
    {
      "epoch": 17.49255287459041,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.001413799520139931,
      "loss": 0.8094,
      "num_input_tokens_seen": 68186256,
      "step": 117445
    },
    {
      "epoch": 17.493297587131366,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0014129733341155704,
      "loss": 0.793,
      "num_input_tokens_seen": 68189200,
      "step": 117450
    },
    {
      "epoch": 17.494042299672326,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0014121473776302124,
      "loss": 0.79,
      "num_input_tokens_seen": 68192304,
      "step": 117455
    },
    {
      "epoch": 17.494787012213287,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0014113216506978176,
      "loss": 0.793,
      "num_input_tokens_seen": 68195440,
      "step": 117460
    },
    {
      "epoch": 17.495531724754244,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0014104961533323317,
      "loss": 0.8082,
      "num_input_tokens_seen": 68198192,
      "step": 117465
    },
    {
      "epoch": 17.496276437295204,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0014096708855477003,
      "loss": 0.7927,
      "num_input_tokens_seen": 68200944,
      "step": 117470
    },
    {
      "epoch": 17.497021149836165,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0014088458473578685,
      "loss": 0.7915,
      "num_input_tokens_seen": 68203856,
      "step": 117475
    },
    {
      "epoch": 17.497765862377122,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0014080210387767704,
      "loss": 0.8076,
      "num_input_tokens_seen": 68206608,
      "step": 117480
    },
    {
      "epoch": 17.498510574918082,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.001407196459818345,
      "loss": 0.8173,
      "num_input_tokens_seen": 68209456,
      "step": 117485
    },
    {
      "epoch": 17.49925528745904,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.001406372110496521,
      "loss": 0.8246,
      "num_input_tokens_seen": 68212400,
      "step": 117490
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0014055479908252238,
      "loss": 0.8032,
      "num_input_tokens_seen": 68215440,
      "step": 117495
    },
    {
      "epoch": 17.50074471254096,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0014047241008183725,
      "loss": 0.8185,
      "num_input_tokens_seen": 68218288,
      "step": 117500
    },
    {
      "epoch": 17.501489425081918,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0014039004404898942,
      "loss": 0.7743,
      "num_input_tokens_seen": 68221520,
      "step": 117505
    },
    {
      "epoch": 17.502234137622878,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0014030770098536992,
      "loss": 0.7885,
      "num_input_tokens_seen": 68224432,
      "step": 117510
    },
    {
      "epoch": 17.502978850163835,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.001402253808923697,
      "loss": 0.7877,
      "num_input_tokens_seen": 68227408,
      "step": 117515
    },
    {
      "epoch": 17.503723562704796,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0014014308377137974,
      "loss": 0.808,
      "num_input_tokens_seen": 68230288,
      "step": 117520
    },
    {
      "epoch": 17.504468275245756,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0014006080962379013,
      "loss": 0.8259,
      "num_input_tokens_seen": 68233168,
      "step": 117525
    },
    {
      "epoch": 17.505212987786713,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0013997855845099111,
      "loss": 0.7951,
      "num_input_tokens_seen": 68236176,
      "step": 117530
    },
    {
      "epoch": 17.505957700327674,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0013989633025437204,
      "loss": 0.7841,
      "num_input_tokens_seen": 68238992,
      "step": 117535
    },
    {
      "epoch": 17.506702412868634,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0013981412503532213,
      "loss": 0.7747,
      "num_input_tokens_seen": 68242032,
      "step": 117540
    },
    {
      "epoch": 17.50744712540959,
      "grad_norm": 0.0546875,
      "learning_rate": 0.001397319427952303,
      "loss": 0.8112,
      "num_input_tokens_seen": 68245168,
      "step": 117545
    },
    {
      "epoch": 17.50819183795055,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0013964978353548478,
      "loss": 0.7922,
      "num_input_tokens_seen": 68248048,
      "step": 117550
    },
    {
      "epoch": 17.50893655049151,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0013956764725747328,
      "loss": 0.8303,
      "num_input_tokens_seen": 68251088,
      "step": 117555
    },
    {
      "epoch": 17.50968126303247,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.00139485533962584,
      "loss": 0.8085,
      "num_input_tokens_seen": 68253872,
      "step": 117560
    },
    {
      "epoch": 17.51042597557343,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.001394034436522037,
      "loss": 0.7986,
      "num_input_tokens_seen": 68256944,
      "step": 117565
    },
    {
      "epoch": 17.511170688114387,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0013932137632771922,
      "loss": 0.7975,
      "num_input_tokens_seen": 68260112,
      "step": 117570
    },
    {
      "epoch": 17.511915400655347,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0013923933199051753,
      "loss": 0.8021,
      "num_input_tokens_seen": 68262704,
      "step": 117575
    },
    {
      "epoch": 17.512660113196308,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0013915731064198394,
      "loss": 0.799,
      "num_input_tokens_seen": 68265552,
      "step": 117580
    },
    {
      "epoch": 17.513404825737265,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0013907531228350473,
      "loss": 0.7908,
      "num_input_tokens_seen": 68268432,
      "step": 117585
    },
    {
      "epoch": 17.514149538278225,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0013899333691646493,
      "loss": 0.7957,
      "num_input_tokens_seen": 68271376,
      "step": 117590
    },
    {
      "epoch": 17.514894250819182,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0013891138454224926,
      "loss": 0.8028,
      "num_input_tokens_seen": 68274192,
      "step": 117595
    },
    {
      "epoch": 17.515638963360143,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0013882945516224248,
      "loss": 0.7701,
      "num_input_tokens_seen": 68277328,
      "step": 117600
    },
    {
      "epoch": 17.516383675901103,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0013874754877782846,
      "loss": 0.7852,
      "num_input_tokens_seen": 68280048,
      "step": 117605
    },
    {
      "epoch": 17.51712838844206,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0013866566539039143,
      "loss": 0.8079,
      "num_input_tokens_seen": 68282896,
      "step": 117610
    },
    {
      "epoch": 17.51787310098302,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0013858380500131428,
      "loss": 0.8035,
      "num_input_tokens_seen": 68285488,
      "step": 117615
    },
    {
      "epoch": 17.51861781352398,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0013850196761198024,
      "loss": 0.811,
      "num_input_tokens_seen": 68288240,
      "step": 117620
    },
    {
      "epoch": 17.51936252606494,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0013842015322377153,
      "loss": 0.8083,
      "num_input_tokens_seen": 68290992,
      "step": 117625
    },
    {
      "epoch": 17.5201072386059,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0013833836183807023,
      "loss": 0.8149,
      "num_input_tokens_seen": 68293872,
      "step": 117630
    },
    {
      "epoch": 17.520851951146856,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0013825659345625874,
      "loss": 0.8003,
      "num_input_tokens_seen": 68296816,
      "step": 117635
    },
    {
      "epoch": 17.521596663687816,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0013817484807971774,
      "loss": 0.8033,
      "num_input_tokens_seen": 68299792,
      "step": 117640
    },
    {
      "epoch": 17.522341376228777,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0013809312570982884,
      "loss": 0.7681,
      "num_input_tokens_seen": 68302736,
      "step": 117645
    },
    {
      "epoch": 17.523086088769734,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0013801142634797209,
      "loss": 0.7673,
      "num_input_tokens_seen": 68305680,
      "step": 117650
    },
    {
      "epoch": 17.523830801310694,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0013792974999552836,
      "loss": 0.8163,
      "num_input_tokens_seen": 68308496,
      "step": 117655
    },
    {
      "epoch": 17.52457551385165,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0013784809665387692,
      "loss": 0.8012,
      "num_input_tokens_seen": 68311376,
      "step": 117660
    },
    {
      "epoch": 17.525320226392612,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0013776646632439765,
      "loss": 0.8353,
      "num_input_tokens_seen": 68314096,
      "step": 117665
    },
    {
      "epoch": 17.526064938933573,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0013768485900846944,
      "loss": 0.8067,
      "num_input_tokens_seen": 68317328,
      "step": 117670
    },
    {
      "epoch": 17.52680965147453,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.001376032747074707,
      "loss": 0.7901,
      "num_input_tokens_seen": 68320208,
      "step": 117675
    },
    {
      "epoch": 17.52755436401549,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.001375217134227803,
      "loss": 0.8151,
      "num_input_tokens_seen": 68323504,
      "step": 117680
    },
    {
      "epoch": 17.52829907655645,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0013744017515577567,
      "loss": 0.783,
      "num_input_tokens_seen": 68326576,
      "step": 117685
    },
    {
      "epoch": 17.529043789097408,
      "grad_norm": 0.0703125,
      "learning_rate": 0.001373586599078345,
      "loss": 0.7881,
      "num_input_tokens_seen": 68329232,
      "step": 117690
    },
    {
      "epoch": 17.529788501638368,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0013727716768033359,
      "loss": 0.8017,
      "num_input_tokens_seen": 68332176,
      "step": 117695
    },
    {
      "epoch": 17.530533214179325,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0013719569847465013,
      "loss": 0.7955,
      "num_input_tokens_seen": 68335472,
      "step": 117700
    },
    {
      "epoch": 17.531277926720286,
      "grad_norm": 0.046875,
      "learning_rate": 0.0013711425229216035,
      "loss": 0.7867,
      "num_input_tokens_seen": 68338384,
      "step": 117705
    },
    {
      "epoch": 17.532022639261246,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0013703282913423965,
      "loss": 0.8107,
      "num_input_tokens_seen": 68341328,
      "step": 117710
    },
    {
      "epoch": 17.532767351802203,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0013695142900226442,
      "loss": 0.8089,
      "num_input_tokens_seen": 68344176,
      "step": 117715
    },
    {
      "epoch": 17.533512064343164,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0013687005189760926,
      "loss": 0.7962,
      "num_input_tokens_seen": 68346832,
      "step": 117720
    },
    {
      "epoch": 17.534256776884124,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0013678869782164937,
      "loss": 0.7907,
      "num_input_tokens_seen": 68349936,
      "step": 117725
    },
    {
      "epoch": 17.53500148942508,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.001367073667757585,
      "loss": 0.7955,
      "num_input_tokens_seen": 68353168,
      "step": 117730
    },
    {
      "epoch": 17.53574620196604,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.0013662605876131155,
      "loss": 0.7921,
      "num_input_tokens_seen": 68355856,
      "step": 117735
    },
    {
      "epoch": 17.536490914507,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0013654477377968127,
      "loss": 0.7957,
      "num_input_tokens_seen": 68359120,
      "step": 117740
    },
    {
      "epoch": 17.53723562704796,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0013646351183224169,
      "loss": 0.8023,
      "num_input_tokens_seen": 68362352,
      "step": 117745
    },
    {
      "epoch": 17.53798033958892,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0013638227292036508,
      "loss": 0.7828,
      "num_input_tokens_seen": 68365296,
      "step": 117750
    },
    {
      "epoch": 17.538725052129877,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0013630105704542416,
      "loss": 0.7938,
      "num_input_tokens_seen": 68368240,
      "step": 117755
    },
    {
      "epoch": 17.539469764670837,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0013621986420879085,
      "loss": 0.7773,
      "num_input_tokens_seen": 68371152,
      "step": 117760
    },
    {
      "epoch": 17.540214477211798,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0013613869441183668,
      "loss": 0.8435,
      "num_input_tokens_seen": 68373936,
      "step": 117765
    },
    {
      "epoch": 17.540959189752755,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0013605754765593325,
      "loss": 0.8021,
      "num_input_tokens_seen": 68376848,
      "step": 117770
    },
    {
      "epoch": 17.541703902293715,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.0013597642394245096,
      "loss": 0.791,
      "num_input_tokens_seen": 68379664,
      "step": 117775
    },
    {
      "epoch": 17.542448614834672,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0013589532327276104,
      "loss": 0.8073,
      "num_input_tokens_seen": 68382480,
      "step": 117780
    },
    {
      "epoch": 17.543193327375633,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0013581424564823323,
      "loss": 0.8097,
      "num_input_tokens_seen": 68385136,
      "step": 117785
    },
    {
      "epoch": 17.543938039916593,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0013573319107023695,
      "loss": 0.7956,
      "num_input_tokens_seen": 68387856,
      "step": 117790
    },
    {
      "epoch": 17.54468275245755,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.001356521595401421,
      "loss": 0.8098,
      "num_input_tokens_seen": 68390736,
      "step": 117795
    },
    {
      "epoch": 17.54542746499851,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0013557115105931705,
      "loss": 0.8101,
      "num_input_tokens_seen": 68393680,
      "step": 117800
    },
    {
      "epoch": 17.54617217753947,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.001354901656291309,
      "loss": 0.7857,
      "num_input_tokens_seen": 68396592,
      "step": 117805
    },
    {
      "epoch": 17.54691689008043,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.001354092032509514,
      "loss": 0.7877,
      "num_input_tokens_seen": 68399472,
      "step": 117810
    },
    {
      "epoch": 17.54766160262139,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0013532826392614677,
      "loss": 0.7848,
      "num_input_tokens_seen": 68402832,
      "step": 117815
    },
    {
      "epoch": 17.548406315162346,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0013524734765608408,
      "loss": 0.7826,
      "num_input_tokens_seen": 68405648,
      "step": 117820
    },
    {
      "epoch": 17.549151027703306,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0013516645444213043,
      "loss": 0.7995,
      "num_input_tokens_seen": 68408528,
      "step": 117825
    },
    {
      "epoch": 17.549895740244267,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0013508558428565237,
      "loss": 0.8007,
      "num_input_tokens_seen": 68411440,
      "step": 117830
    },
    {
      "epoch": 17.550640452785224,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.001350047371880158,
      "loss": 0.7929,
      "num_input_tokens_seen": 68414480,
      "step": 117835
    },
    {
      "epoch": 17.551385165326185,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0013492391315058716,
      "loss": 0.8067,
      "num_input_tokens_seen": 68417264,
      "step": 117840
    },
    {
      "epoch": 17.552129877867145,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0013484311217473132,
      "loss": 0.803,
      "num_input_tokens_seen": 68420272,
      "step": 117845
    },
    {
      "epoch": 17.552874590408102,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0013476233426181388,
      "loss": 0.7967,
      "num_input_tokens_seen": 68423312,
      "step": 117850
    },
    {
      "epoch": 17.553619302949063,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0013468157941319908,
      "loss": 0.7917,
      "num_input_tokens_seen": 68426352,
      "step": 117855
    },
    {
      "epoch": 17.55436401549002,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0013460084763025149,
      "loss": 0.8373,
      "num_input_tokens_seen": 68429456,
      "step": 117860
    },
    {
      "epoch": 17.55510872803098,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.00134520138914335,
      "loss": 0.8065,
      "num_input_tokens_seen": 68431952,
      "step": 117865
    },
    {
      "epoch": 17.55585344057194,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0013443945326681256,
      "loss": 0.7898,
      "num_input_tokens_seen": 68435120,
      "step": 117870
    },
    {
      "epoch": 17.556598153112898,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0013435879068904788,
      "loss": 0.7902,
      "num_input_tokens_seen": 68438064,
      "step": 117875
    },
    {
      "epoch": 17.557342865653858,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0013427815118240353,
      "loss": 0.8097,
      "num_input_tokens_seen": 68440880,
      "step": 117880
    },
    {
      "epoch": 17.558087578194815,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0013419753474824146,
      "loss": 0.7828,
      "num_input_tokens_seen": 68444080,
      "step": 117885
    },
    {
      "epoch": 17.558832290735776,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0013411694138792424,
      "loss": 0.7817,
      "num_input_tokens_seen": 68447088,
      "step": 117890
    },
    {
      "epoch": 17.559577003276736,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0013403637110281307,
      "loss": 0.8059,
      "num_input_tokens_seen": 68449936,
      "step": 117895
    },
    {
      "epoch": 17.560321715817693,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0013395582389426875,
      "loss": 0.7954,
      "num_input_tokens_seen": 68453168,
      "step": 117900
    },
    {
      "epoch": 17.561066428358654,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.001338752997636528,
      "loss": 0.8025,
      "num_input_tokens_seen": 68456016,
      "step": 117905
    },
    {
      "epoch": 17.561811140899614,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0013379479871232502,
      "loss": 0.792,
      "num_input_tokens_seen": 68459088,
      "step": 117910
    },
    {
      "epoch": 17.56255585344057,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0013371432074164545,
      "loss": 0.792,
      "num_input_tokens_seen": 68462000,
      "step": 117915
    },
    {
      "epoch": 17.56330056598153,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0013363386585297404,
      "loss": 0.818,
      "num_input_tokens_seen": 68465040,
      "step": 117920
    },
    {
      "epoch": 17.56404527852249,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0013355343404766933,
      "loss": 0.7965,
      "num_input_tokens_seen": 68468240,
      "step": 117925
    },
    {
      "epoch": 17.56478999106345,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0013347302532709109,
      "loss": 0.7856,
      "num_input_tokens_seen": 68471120,
      "step": 117930
    },
    {
      "epoch": 17.56553470360441,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0013339263969259674,
      "loss": 0.7865,
      "num_input_tokens_seen": 68474064,
      "step": 117935
    },
    {
      "epoch": 17.566279416145367,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.0013331227714554516,
      "loss": 0.7989,
      "num_input_tokens_seen": 68476848,
      "step": 117940
    },
    {
      "epoch": 17.567024128686327,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0013323193768729345,
      "loss": 0.8112,
      "num_input_tokens_seen": 68479920,
      "step": 117945
    },
    {
      "epoch": 17.567768841227288,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.001331516213191992,
      "loss": 0.7907,
      "num_input_tokens_seen": 68482512,
      "step": 117950
    },
    {
      "epoch": 17.568513553768245,
      "grad_norm": 0.0625,
      "learning_rate": 0.0013307132804261917,
      "loss": 0.8048,
      "num_input_tokens_seen": 68485392,
      "step": 117955
    },
    {
      "epoch": 17.569258266309205,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0013299105785890923,
      "loss": 0.8011,
      "num_input_tokens_seen": 68488240,
      "step": 117960
    },
    {
      "epoch": 17.570002978850162,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.001329108107694265,
      "loss": 0.7762,
      "num_input_tokens_seen": 68491120,
      "step": 117965
    },
    {
      "epoch": 17.570747691391123,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0013283058677552588,
      "loss": 0.7992,
      "num_input_tokens_seen": 68493968,
      "step": 117970
    },
    {
      "epoch": 17.571492403932083,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0013275038587856314,
      "loss": 0.8032,
      "num_input_tokens_seen": 68496848,
      "step": 117975
    },
    {
      "epoch": 17.57223711647304,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0013267020807989282,
      "loss": 0.7762,
      "num_input_tokens_seen": 68499664,
      "step": 117980
    },
    {
      "epoch": 17.572981829014,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0013259005338086987,
      "loss": 0.785,
      "num_input_tokens_seen": 68502608,
      "step": 117985
    },
    {
      "epoch": 17.57372654155496,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.001325099217828482,
      "loss": 0.8165,
      "num_input_tokens_seen": 68505296,
      "step": 117990
    },
    {
      "epoch": 17.57447125409592,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0013242981328718138,
      "loss": 0.7832,
      "num_input_tokens_seen": 68508336,
      "step": 117995
    },
    {
      "epoch": 17.57521596663688,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0013234972789522303,
      "loss": 0.7918,
      "num_input_tokens_seen": 68512208,
      "step": 118000
    },
    {
      "epoch": 17.575960679177836,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.0013226966560832586,
      "loss": 0.8023,
      "num_input_tokens_seen": 68515216,
      "step": 118005
    },
    {
      "epoch": 17.576705391718797,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.001321896264278428,
      "loss": 0.7743,
      "num_input_tokens_seen": 68518224,
      "step": 118010
    },
    {
      "epoch": 17.577450104259757,
      "grad_norm": 0.046875,
      "learning_rate": 0.0013210961035512597,
      "loss": 0.8346,
      "num_input_tokens_seen": 68521200,
      "step": 118015
    },
    {
      "epoch": 17.578194816800714,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.001320296173915269,
      "loss": 0.8056,
      "num_input_tokens_seen": 68524272,
      "step": 118020
    },
    {
      "epoch": 17.578939529341675,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0013194964753839706,
      "loss": 0.7801,
      "num_input_tokens_seen": 68527120,
      "step": 118025
    },
    {
      "epoch": 17.57968424188263,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.001318697007970873,
      "loss": 0.821,
      "num_input_tokens_seen": 68529808,
      "step": 118030
    },
    {
      "epoch": 17.580428954423592,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0013178977716894863,
      "loss": 0.7968,
      "num_input_tokens_seen": 68532912,
      "step": 118035
    },
    {
      "epoch": 17.581173666964553,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.001317098766553309,
      "loss": 0.8154,
      "num_input_tokens_seen": 68535696,
      "step": 118040
    },
    {
      "epoch": 17.58191837950551,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0013162999925758422,
      "loss": 0.7922,
      "num_input_tokens_seen": 68538416,
      "step": 118045
    },
    {
      "epoch": 17.58266309204647,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0013155014497705786,
      "loss": 0.7955,
      "num_input_tokens_seen": 68541264,
      "step": 118050
    },
    {
      "epoch": 17.58340780458743,
      "grad_norm": 0.072265625,
      "learning_rate": 0.001314703138151012,
      "loss": 0.7791,
      "num_input_tokens_seen": 68544272,
      "step": 118055
    },
    {
      "epoch": 17.584152517128388,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0013139050577306254,
      "loss": 0.813,
      "num_input_tokens_seen": 68547376,
      "step": 118060
    },
    {
      "epoch": 17.584897229669348,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0013131072085229044,
      "loss": 0.8171,
      "num_input_tokens_seen": 68550928,
      "step": 118065
    },
    {
      "epoch": 17.585641942210305,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.001312309590541325,
      "loss": 0.8005,
      "num_input_tokens_seen": 68553680,
      "step": 118070
    },
    {
      "epoch": 17.586386654751266,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0013115122037993632,
      "loss": 0.7945,
      "num_input_tokens_seen": 68556592,
      "step": 118075
    },
    {
      "epoch": 17.587131367292226,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.001310715048310491,
      "loss": 0.7997,
      "num_input_tokens_seen": 68559568,
      "step": 118080
    },
    {
      "epoch": 17.587876079833183,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0013099181240881768,
      "loss": 0.8,
      "num_input_tokens_seen": 68562096,
      "step": 118085
    },
    {
      "epoch": 17.588620792374144,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0013091214311458808,
      "loss": 0.8039,
      "num_input_tokens_seen": 68565104,
      "step": 118090
    },
    {
      "epoch": 17.589365504915104,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.001308324969497061,
      "loss": 0.7976,
      "num_input_tokens_seen": 68567920,
      "step": 118095
    },
    {
      "epoch": 17.59011021745606,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0013075287391551781,
      "loss": 0.8037,
      "num_input_tokens_seen": 68570800,
      "step": 118100
    },
    {
      "epoch": 17.590854929997022,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.001306732740133678,
      "loss": 0.7844,
      "num_input_tokens_seen": 68573616,
      "step": 118105
    },
    {
      "epoch": 17.59159964253798,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0013059369724460134,
      "loss": 0.7899,
      "num_input_tokens_seen": 68576528,
      "step": 118110
    },
    {
      "epoch": 17.59234435507894,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.001305141436105625,
      "loss": 0.8136,
      "num_input_tokens_seen": 68579408,
      "step": 118115
    },
    {
      "epoch": 17.5930890676199,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0013043461311259524,
      "loss": 0.7872,
      "num_input_tokens_seen": 68582128,
      "step": 118120
    },
    {
      "epoch": 17.593833780160857,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0013035510575204344,
      "loss": 0.838,
      "num_input_tokens_seen": 68584752,
      "step": 118125
    },
    {
      "epoch": 17.594578492701817,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0013027562153024974,
      "loss": 0.771,
      "num_input_tokens_seen": 68587600,
      "step": 118130
    },
    {
      "epoch": 17.595323205242778,
      "grad_norm": 0.046875,
      "learning_rate": 0.0013019616044855752,
      "loss": 0.7935,
      "num_input_tokens_seen": 68590800,
      "step": 118135
    },
    {
      "epoch": 17.596067917783735,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0013011672250830874,
      "loss": 0.8117,
      "num_input_tokens_seen": 68593712,
      "step": 118140
    },
    {
      "epoch": 17.596812630324695,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0013003730771084599,
      "loss": 0.7881,
      "num_input_tokens_seen": 68596624,
      "step": 118145
    },
    {
      "epoch": 17.597557342865652,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.001299579160575105,
      "loss": 0.8173,
      "num_input_tokens_seen": 68599728,
      "step": 118150
    },
    {
      "epoch": 17.598302055406613,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0012987854754964356,
      "loss": 0.7879,
      "num_input_tokens_seen": 68602448,
      "step": 118155
    },
    {
      "epoch": 17.599046767947573,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.001297992021885861,
      "loss": 0.8093,
      "num_input_tokens_seen": 68604976,
      "step": 118160
    },
    {
      "epoch": 17.59979148048853,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.001297198799756782,
      "loss": 0.812,
      "num_input_tokens_seen": 68607824,
      "step": 118165
    },
    {
      "epoch": 17.60053619302949,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0012964058091226042,
      "loss": 0.8215,
      "num_input_tokens_seen": 68610512,
      "step": 118170
    },
    {
      "epoch": 17.601280905570448,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0012956130499967194,
      "loss": 0.791,
      "num_input_tokens_seen": 68613328,
      "step": 118175
    },
    {
      "epoch": 17.60202561811141,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.001294820522392528,
      "loss": 0.8172,
      "num_input_tokens_seen": 68616240,
      "step": 118180
    },
    {
      "epoch": 17.60277033065237,
      "grad_norm": 0.044921875,
      "learning_rate": 0.001294028226323411,
      "loss": 0.8283,
      "num_input_tokens_seen": 68619088,
      "step": 118185
    },
    {
      "epoch": 17.603515043193326,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.001293236161802761,
      "loss": 0.788,
      "num_input_tokens_seen": 68621936,
      "step": 118190
    },
    {
      "epoch": 17.604259755734287,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.001292444328843954,
      "loss": 0.7925,
      "num_input_tokens_seen": 68624656,
      "step": 118195
    },
    {
      "epoch": 17.605004468275247,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0012916527274603644,
      "loss": 0.7961,
      "num_input_tokens_seen": 68627536,
      "step": 118200
    },
    {
      "epoch": 17.605749180816204,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0012908613576653749,
      "loss": 0.8115,
      "num_input_tokens_seen": 68630320,
      "step": 118205
    },
    {
      "epoch": 17.606493893357165,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0012900702194723478,
      "loss": 0.7824,
      "num_input_tokens_seen": 68633168,
      "step": 118210
    },
    {
      "epoch": 17.60723860589812,
      "grad_norm": 0.0625,
      "learning_rate": 0.0012892793128946479,
      "loss": 0.783,
      "num_input_tokens_seen": 68636176,
      "step": 118215
    },
    {
      "epoch": 17.607983318439082,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0012884886379456405,
      "loss": 0.787,
      "num_input_tokens_seen": 68639056,
      "step": 118220
    },
    {
      "epoch": 17.608728030980043,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0012876981946386822,
      "loss": 0.7894,
      "num_input_tokens_seen": 68641904,
      "step": 118225
    },
    {
      "epoch": 17.609472743521,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0012869079829871271,
      "loss": 0.7961,
      "num_input_tokens_seen": 68644656,
      "step": 118230
    },
    {
      "epoch": 17.61021745606196,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0012861180030043211,
      "loss": 0.7814,
      "num_input_tokens_seen": 68647728,
      "step": 118235
    },
    {
      "epoch": 17.61096216860292,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.001285328254703617,
      "loss": 0.7959,
      "num_input_tokens_seen": 68650352,
      "step": 118240
    },
    {
      "epoch": 17.611706881143878,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0012845387380983491,
      "loss": 0.8021,
      "num_input_tokens_seen": 68653424,
      "step": 118245
    },
    {
      "epoch": 17.61245159368484,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0012837494532018635,
      "loss": 0.8124,
      "num_input_tokens_seen": 68656176,
      "step": 118250
    },
    {
      "epoch": 17.613196306225795,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0012829604000274874,
      "loss": 0.7818,
      "num_input_tokens_seen": 68659120,
      "step": 118255
    },
    {
      "epoch": 17.613941018766756,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0012821715785885557,
      "loss": 0.803,
      "num_input_tokens_seen": 68662352,
      "step": 118260
    },
    {
      "epoch": 17.614685731307716,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0012813829888983907,
      "loss": 0.7776,
      "num_input_tokens_seen": 68665872,
      "step": 118265
    },
    {
      "epoch": 17.615430443848673,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0012805946309703187,
      "loss": 0.7949,
      "num_input_tokens_seen": 68668688,
      "step": 118270
    },
    {
      "epoch": 17.616175156389634,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0012798065048176587,
      "loss": 0.7973,
      "num_input_tokens_seen": 68671760,
      "step": 118275
    },
    {
      "epoch": 17.616919868930594,
      "grad_norm": 0.028564453125,
      "learning_rate": 0.0012790186104537221,
      "loss": 0.7762,
      "num_input_tokens_seen": 68674416,
      "step": 118280
    },
    {
      "epoch": 17.61766458147155,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0012782309478918212,
      "loss": 0.8451,
      "num_input_tokens_seen": 68677424,
      "step": 118285
    },
    {
      "epoch": 17.618409294012512,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.0012774435171452592,
      "loss": 0.793,
      "num_input_tokens_seen": 68680368,
      "step": 118290
    },
    {
      "epoch": 17.61915400655347,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0012766563182273433,
      "loss": 0.7769,
      "num_input_tokens_seen": 68683344,
      "step": 118295
    },
    {
      "epoch": 17.61989871909443,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0012758693511513712,
      "loss": 0.7972,
      "num_input_tokens_seen": 68686256,
      "step": 118300
    },
    {
      "epoch": 17.62064343163539,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.0012750826159306377,
      "loss": 0.7893,
      "num_input_tokens_seen": 68689008,
      "step": 118305
    },
    {
      "epoch": 17.621388144176347,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0012742961125784369,
      "loss": 0.7837,
      "num_input_tokens_seen": 68691760,
      "step": 118310
    },
    {
      "epoch": 17.622132856717307,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0012735098411080482,
      "loss": 0.799,
      "num_input_tokens_seen": 68694736,
      "step": 118315
    },
    {
      "epoch": 17.622877569258268,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0012727238015327624,
      "loss": 0.8042,
      "num_input_tokens_seen": 68697968,
      "step": 118320
    },
    {
      "epoch": 17.623622281799225,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0012719379938658558,
      "loss": 0.7886,
      "num_input_tokens_seen": 68702000,
      "step": 118325
    },
    {
      "epoch": 17.624366994340185,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0012711524181206045,
      "loss": 0.7845,
      "num_input_tokens_seen": 68704784,
      "step": 118330
    },
    {
      "epoch": 17.625111706881142,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0012703670743102779,
      "loss": 0.7987,
      "num_input_tokens_seen": 68707536,
      "step": 118335
    },
    {
      "epoch": 17.625856419422103,
      "grad_norm": 0.03125,
      "learning_rate": 0.0012695819624481486,
      "loss": 0.8069,
      "num_input_tokens_seen": 68710544,
      "step": 118340
    },
    {
      "epoch": 17.626601131963064,
      "grad_norm": 0.037109375,
      "learning_rate": 0.001268797082547476,
      "loss": 0.7903,
      "num_input_tokens_seen": 68713424,
      "step": 118345
    },
    {
      "epoch": 17.62734584450402,
      "grad_norm": 0.02734375,
      "learning_rate": 0.001268012434621521,
      "loss": 0.7985,
      "num_input_tokens_seen": 68716432,
      "step": 118350
    },
    {
      "epoch": 17.62809055704498,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.00126722801868354,
      "loss": 0.8031,
      "num_input_tokens_seen": 68719408,
      "step": 118355
    },
    {
      "epoch": 17.62883526958594,
      "grad_norm": 0.10546875,
      "learning_rate": 0.001266443834746782,
      "loss": 0.8113,
      "num_input_tokens_seen": 68722224,
      "step": 118360
    },
    {
      "epoch": 17.6295799821269,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0012656598828245003,
      "loss": 0.787,
      "num_input_tokens_seen": 68724912,
      "step": 118365
    },
    {
      "epoch": 17.63032469466786,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.001264876162929932,
      "loss": 0.763,
      "num_input_tokens_seen": 68728048,
      "step": 118370
    },
    {
      "epoch": 17.631069407208816,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0012640926750763237,
      "loss": 0.8164,
      "num_input_tokens_seen": 68730768,
      "step": 118375
    },
    {
      "epoch": 17.631814119749777,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0012633094192769062,
      "loss": 0.7935,
      "num_input_tokens_seen": 68733328,
      "step": 118380
    },
    {
      "epoch": 17.632558832290737,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.001262526395544919,
      "loss": 0.7821,
      "num_input_tokens_seen": 68736624,
      "step": 118385
    },
    {
      "epoch": 17.633303544831694,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0012617436038935847,
      "loss": 0.8005,
      "num_input_tokens_seen": 68739600,
      "step": 118390
    },
    {
      "epoch": 17.634048257372655,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.001260961044336128,
      "loss": 0.8167,
      "num_input_tokens_seen": 68742416,
      "step": 118395
    },
    {
      "epoch": 17.63479296991361,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0012601787168857715,
      "loss": 0.807,
      "num_input_tokens_seen": 68745328,
      "step": 118400
    },
    {
      "epoch": 17.635537682454572,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0012593966215557295,
      "loss": 0.7736,
      "num_input_tokens_seen": 68748240,
      "step": 118405
    },
    {
      "epoch": 17.636282394995533,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00125861475835922,
      "loss": 0.7986,
      "num_input_tokens_seen": 68751184,
      "step": 118410
    },
    {
      "epoch": 17.63702710753649,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0012578331273094456,
      "loss": 0.8203,
      "num_input_tokens_seen": 68754096,
      "step": 118415
    },
    {
      "epoch": 17.63777182007745,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0012570517284196158,
      "loss": 0.7922,
      "num_input_tokens_seen": 68757040,
      "step": 118420
    },
    {
      "epoch": 17.63851653261841,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0012562705617029251,
      "loss": 0.8362,
      "num_input_tokens_seen": 68759888,
      "step": 118425
    },
    {
      "epoch": 17.639261245159368,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.0012554896271725778,
      "loss": 0.7923,
      "num_input_tokens_seen": 68762800,
      "step": 118430
    },
    {
      "epoch": 17.64000595770033,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0012547089248417652,
      "loss": 0.8025,
      "num_input_tokens_seen": 68765776,
      "step": 118435
    },
    {
      "epoch": 17.640750670241285,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0012539284547236718,
      "loss": 0.7937,
      "num_input_tokens_seen": 68768592,
      "step": 118440
    },
    {
      "epoch": 17.641495382782246,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0012531482168314882,
      "loss": 0.799,
      "num_input_tokens_seen": 68771760,
      "step": 118445
    },
    {
      "epoch": 17.642240095323206,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0012523682111783912,
      "loss": 0.8221,
      "num_input_tokens_seen": 68774576,
      "step": 118450
    },
    {
      "epoch": 17.642984807864163,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.001251588437777565,
      "loss": 0.7916,
      "num_input_tokens_seen": 68777552,
      "step": 118455
    },
    {
      "epoch": 17.643729520405124,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0012508088966421737,
      "loss": 0.7889,
      "num_input_tokens_seen": 68780592,
      "step": 118460
    },
    {
      "epoch": 17.644474232946084,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0012500295877853973,
      "loss": 0.8047,
      "num_input_tokens_seen": 68783536,
      "step": 118465
    },
    {
      "epoch": 17.64521894548704,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.001249250511220395,
      "loss": 0.8008,
      "num_input_tokens_seen": 68786480,
      "step": 118470
    },
    {
      "epoch": 17.645963658028002,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0012484716669603263,
      "loss": 0.7979,
      "num_input_tokens_seen": 68789424,
      "step": 118475
    },
    {
      "epoch": 17.64670837056896,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0012476930550183556,
      "loss": 0.7978,
      "num_input_tokens_seen": 68792144,
      "step": 118480
    },
    {
      "epoch": 17.64745308310992,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0012469146754076343,
      "loss": 0.7849,
      "num_input_tokens_seen": 68795088,
      "step": 118485
    },
    {
      "epoch": 17.64819779565088,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0012461365281413116,
      "loss": 0.7971,
      "num_input_tokens_seen": 68798160,
      "step": 118490
    },
    {
      "epoch": 17.648942508191837,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.001245358613232529,
      "loss": 0.7957,
      "num_input_tokens_seen": 68801232,
      "step": 118495
    },
    {
      "epoch": 17.649687220732797,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0012445809306944372,
      "loss": 0.7858,
      "num_input_tokens_seen": 68804176,
      "step": 118500
    },
    {
      "epoch": 17.650431933273758,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0012438034805401677,
      "loss": 0.8106,
      "num_input_tokens_seen": 68806928,
      "step": 118505
    },
    {
      "epoch": 17.651176645814715,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0012430262627828581,
      "loss": 0.7846,
      "num_input_tokens_seen": 68809616,
      "step": 118510
    },
    {
      "epoch": 17.651921358355676,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0012422492774356397,
      "loss": 0.7907,
      "num_input_tokens_seen": 68812496,
      "step": 118515
    },
    {
      "epoch": 17.652666070896633,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0012414725245116338,
      "loss": 0.8039,
      "num_input_tokens_seen": 68815312,
      "step": 118520
    },
    {
      "epoch": 17.653410783437593,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.001240696004023968,
      "loss": 0.7796,
      "num_input_tokens_seen": 68818384,
      "step": 118525
    },
    {
      "epoch": 17.654155495978554,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0012399197159857567,
      "loss": 0.7914,
      "num_input_tokens_seen": 68820976,
      "step": 118530
    },
    {
      "epoch": 17.65490020851951,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0012391436604101197,
      "loss": 0.7841,
      "num_input_tokens_seen": 68823792,
      "step": 118535
    },
    {
      "epoch": 17.65564492106047,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0012383678373101648,
      "loss": 0.7986,
      "num_input_tokens_seen": 68826896,
      "step": 118540
    },
    {
      "epoch": 17.656389633601428,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0012375922466989963,
      "loss": 0.8188,
      "num_input_tokens_seen": 68829808,
      "step": 118545
    },
    {
      "epoch": 17.65713434614239,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0012368168885897206,
      "loss": 0.7977,
      "num_input_tokens_seen": 68832624,
      "step": 118550
    },
    {
      "epoch": 17.65787905868335,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0012360417629954357,
      "loss": 0.7969,
      "num_input_tokens_seen": 68835376,
      "step": 118555
    },
    {
      "epoch": 17.658623771224306,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0012352668699292374,
      "loss": 0.7951,
      "num_input_tokens_seen": 68838800,
      "step": 118560
    },
    {
      "epoch": 17.659368483765267,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.0012344922094042103,
      "loss": 0.7918,
      "num_input_tokens_seen": 68841520,
      "step": 118565
    },
    {
      "epoch": 17.660113196306227,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.001233717781433451,
      "loss": 0.8065,
      "num_input_tokens_seen": 68844432,
      "step": 118570
    },
    {
      "epoch": 17.660857908847184,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0012329435860300353,
      "loss": 0.8049,
      "num_input_tokens_seen": 68847472,
      "step": 118575
    },
    {
      "epoch": 17.661602621388145,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0012321696232070495,
      "loss": 0.7911,
      "num_input_tokens_seen": 68850672,
      "step": 118580
    },
    {
      "epoch": 17.6623473339291,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0012313958929775614,
      "loss": 0.8167,
      "num_input_tokens_seen": 68853584,
      "step": 118585
    },
    {
      "epoch": 17.663092046470062,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0012306223953546475,
      "loss": 0.7894,
      "num_input_tokens_seen": 68856528,
      "step": 118590
    },
    {
      "epoch": 17.663836759011023,
      "grad_norm": 0.025390625,
      "learning_rate": 0.0012298491303513752,
      "loss": 0.7826,
      "num_input_tokens_seen": 68859504,
      "step": 118595
    },
    {
      "epoch": 17.66458147155198,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.001229076097980803,
      "loss": 0.7953,
      "num_input_tokens_seen": 68862416,
      "step": 118600
    },
    {
      "epoch": 17.66532618409294,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0012283032982559965,
      "loss": 0.7904,
      "num_input_tokens_seen": 68865520,
      "step": 118605
    },
    {
      "epoch": 17.6660708966339,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.001227530731190009,
      "loss": 0.809,
      "num_input_tokens_seen": 68868144,
      "step": 118610
    },
    {
      "epoch": 17.666815609174858,
      "grad_norm": 0.40625,
      "learning_rate": 0.0012267583967958916,
      "loss": 0.8615,
      "num_input_tokens_seen": 68870896,
      "step": 118615
    },
    {
      "epoch": 17.66756032171582,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0012259862950866906,
      "loss": 0.8002,
      "num_input_tokens_seen": 68873520,
      "step": 118620
    },
    {
      "epoch": 17.668305034256775,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0012252144260754554,
      "loss": 0.7948,
      "num_input_tokens_seen": 68876304,
      "step": 118625
    },
    {
      "epoch": 17.669049746797736,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0012244427897752208,
      "loss": 0.7907,
      "num_input_tokens_seen": 68879088,
      "step": 118630
    },
    {
      "epoch": 17.669794459338696,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0012236713861990228,
      "loss": 0.8088,
      "num_input_tokens_seen": 68881872,
      "step": 118635
    },
    {
      "epoch": 17.670539171879653,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0012229002153598977,
      "loss": 0.803,
      "num_input_tokens_seen": 68884784,
      "step": 118640
    },
    {
      "epoch": 17.671283884420614,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0012221292772708668,
      "loss": 0.8337,
      "num_input_tokens_seen": 68887728,
      "step": 118645
    },
    {
      "epoch": 17.672028596961574,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0012213585719449633,
      "loss": 0.7999,
      "num_input_tokens_seen": 68890768,
      "step": 118650
    },
    {
      "epoch": 17.67277330950253,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0012205880993951994,
      "loss": 0.7836,
      "num_input_tokens_seen": 68893648,
      "step": 118655
    },
    {
      "epoch": 17.673518022043492,
      "grad_norm": 0.037109375,
      "learning_rate": 0.001219817859634597,
      "loss": 0.7998,
      "num_input_tokens_seen": 68896528,
      "step": 118660
    },
    {
      "epoch": 17.67426273458445,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.0012190478526761638,
      "loss": 0.7955,
      "num_input_tokens_seen": 68899152,
      "step": 118665
    },
    {
      "epoch": 17.67500744712541,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0012182780785329128,
      "loss": 0.803,
      "num_input_tokens_seen": 68902416,
      "step": 118670
    },
    {
      "epoch": 17.67575215966637,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0012175085372178485,
      "loss": 0.771,
      "num_input_tokens_seen": 68905328,
      "step": 118675
    },
    {
      "epoch": 17.676496872207327,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.001216739228743967,
      "loss": 0.7728,
      "num_input_tokens_seen": 68907920,
      "step": 118680
    },
    {
      "epoch": 17.677241584748288,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00121597015312427,
      "loss": 0.8106,
      "num_input_tokens_seen": 68910800,
      "step": 118685
    },
    {
      "epoch": 17.677986297289245,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0012152013103717435,
      "loss": 0.7971,
      "num_input_tokens_seen": 68913616,
      "step": 118690
    },
    {
      "epoch": 17.678731009830205,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0012144327004993836,
      "loss": 0.8331,
      "num_input_tokens_seen": 68916336,
      "step": 118695
    },
    {
      "epoch": 17.679475722371166,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0012136643235201704,
      "loss": 0.7991,
      "num_input_tokens_seen": 68919120,
      "step": 118700
    },
    {
      "epoch": 17.680220434912123,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0012128961794470882,
      "loss": 0.7817,
      "num_input_tokens_seen": 68921968,
      "step": 118705
    },
    {
      "epoch": 17.680965147453083,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0012121282682931117,
      "loss": 0.8065,
      "num_input_tokens_seen": 68924848,
      "step": 118710
    },
    {
      "epoch": 17.681709859994044,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0012113605900712155,
      "loss": 0.784,
      "num_input_tokens_seen": 68927696,
      "step": 118715
    },
    {
      "epoch": 17.682454572535,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0012105931447943673,
      "loss": 0.7943,
      "num_input_tokens_seen": 68930352,
      "step": 118720
    },
    {
      "epoch": 17.68319928507596,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0012098259324755322,
      "loss": 0.7933,
      "num_input_tokens_seen": 68933424,
      "step": 118725
    },
    {
      "epoch": 17.683943997616918,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0012090589531276745,
      "loss": 0.795,
      "num_input_tokens_seen": 68936336,
      "step": 118730
    },
    {
      "epoch": 17.68468871015788,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0012082922067637458,
      "loss": 0.7985,
      "num_input_tokens_seen": 68939216,
      "step": 118735
    },
    {
      "epoch": 17.68543342269884,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0012075256933967071,
      "loss": 0.7787,
      "num_input_tokens_seen": 68942512,
      "step": 118740
    },
    {
      "epoch": 17.686178135239796,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0012067594130395032,
      "loss": 0.799,
      "num_input_tokens_seen": 68945200,
      "step": 118745
    },
    {
      "epoch": 17.686922847780757,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0012059933657050787,
      "loss": 0.8229,
      "num_input_tokens_seen": 68948272,
      "step": 118750
    },
    {
      "epoch": 17.687667560321717,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0012052275514063781,
      "loss": 0.8104,
      "num_input_tokens_seen": 68951472,
      "step": 118755
    },
    {
      "epoch": 17.688412272862674,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0012044619701563346,
      "loss": 0.802,
      "num_input_tokens_seen": 68954384,
      "step": 118760
    },
    {
      "epoch": 17.689156985403635,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.001203696621967888,
      "loss": 0.802,
      "num_input_tokens_seen": 68957040,
      "step": 118765
    },
    {
      "epoch": 17.68990169794459,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0012029315068539624,
      "loss": 0.7957,
      "num_input_tokens_seen": 68959664,
      "step": 118770
    },
    {
      "epoch": 17.690646410485552,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.001202166624827488,
      "loss": 0.8419,
      "num_input_tokens_seen": 68962576,
      "step": 118775
    },
    {
      "epoch": 17.691391123026513,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0012014019759013844,
      "loss": 0.7848,
      "num_input_tokens_seen": 68965712,
      "step": 118780
    },
    {
      "epoch": 17.69213583556747,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.001200637560088571,
      "loss": 0.7785,
      "num_input_tokens_seen": 68968304,
      "step": 118785
    },
    {
      "epoch": 17.69288054810843,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0011998733774019592,
      "loss": 0.811,
      "num_input_tokens_seen": 68971344,
      "step": 118790
    },
    {
      "epoch": 17.69362526064939,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0011991094278544622,
      "loss": 0.8017,
      "num_input_tokens_seen": 68974032,
      "step": 118795
    },
    {
      "epoch": 17.694369973190348,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.0011983457114589863,
      "loss": 0.802,
      "num_input_tokens_seen": 68977040,
      "step": 118800
    },
    {
      "epoch": 17.69511468573131,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0011975822282284292,
      "loss": 0.7903,
      "num_input_tokens_seen": 68979792,
      "step": 118805
    },
    {
      "epoch": 17.695859398272265,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0011968189781756944,
      "loss": 0.8059,
      "num_input_tokens_seen": 68982544,
      "step": 118810
    },
    {
      "epoch": 17.696604110813226,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.0011960559613136744,
      "loss": 0.8058,
      "num_input_tokens_seen": 68985360,
      "step": 118815
    },
    {
      "epoch": 17.697348823354186,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0011952931776552595,
      "loss": 0.8016,
      "num_input_tokens_seen": 68988208,
      "step": 118820
    },
    {
      "epoch": 17.698093535895143,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0011945306272133321,
      "loss": 0.7924,
      "num_input_tokens_seen": 68991280,
      "step": 118825
    },
    {
      "epoch": 17.698838248436104,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0011937683100007822,
      "loss": 0.781,
      "num_input_tokens_seen": 68994288,
      "step": 118830
    },
    {
      "epoch": 17.699582960977065,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0011930062260304846,
      "loss": 0.8037,
      "num_input_tokens_seen": 68997328,
      "step": 118835
    },
    {
      "epoch": 17.70032767351802,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0011922443753153106,
      "loss": 0.7862,
      "num_input_tokens_seen": 69000144,
      "step": 118840
    },
    {
      "epoch": 17.701072386058982,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.001191482757868138,
      "loss": 0.7856,
      "num_input_tokens_seen": 69003152,
      "step": 118845
    },
    {
      "epoch": 17.70181709859994,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0011907213737018267,
      "loss": 0.7948,
      "num_input_tokens_seen": 69005904,
      "step": 118850
    },
    {
      "epoch": 17.7025618111409,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.001189960222829245,
      "loss": 0.7711,
      "num_input_tokens_seen": 69008784,
      "step": 118855
    },
    {
      "epoch": 17.70330652368186,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0011891993052632471,
      "loss": 0.7934,
      "num_input_tokens_seen": 69011376,
      "step": 118860
    },
    {
      "epoch": 17.704051236222817,
      "grad_norm": 0.05859375,
      "learning_rate": 0.001188438621016693,
      "loss": 0.7996,
      "num_input_tokens_seen": 69014512,
      "step": 118865
    },
    {
      "epoch": 17.704795948763778,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0011876781701024308,
      "loss": 0.798,
      "num_input_tokens_seen": 69017648,
      "step": 118870
    },
    {
      "epoch": 17.705540661304738,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.001186917952533305,
      "loss": 0.7794,
      "num_input_tokens_seen": 69020400,
      "step": 118875
    },
    {
      "epoch": 17.706285373845695,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.001186157968322164,
      "loss": 0.7923,
      "num_input_tokens_seen": 69023216,
      "step": 118880
    },
    {
      "epoch": 17.707030086386656,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0011853982174818438,
      "loss": 0.7883,
      "num_input_tokens_seen": 69026192,
      "step": 118885
    },
    {
      "epoch": 17.707774798927613,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0011846387000251813,
      "loss": 0.7778,
      "num_input_tokens_seen": 69028944,
      "step": 118890
    },
    {
      "epoch": 17.708519511468573,
      "grad_norm": 0.095703125,
      "learning_rate": 0.001183879415965004,
      "loss": 0.8112,
      "num_input_tokens_seen": 69031632,
      "step": 118895
    },
    {
      "epoch": 17.709264224009534,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0011831203653141437,
      "loss": 0.7893,
      "num_input_tokens_seen": 69034288,
      "step": 118900
    },
    {
      "epoch": 17.71000893655049,
      "grad_norm": 0.022216796875,
      "learning_rate": 0.00118236154808542,
      "loss": 0.8125,
      "num_input_tokens_seen": 69037136,
      "step": 118905
    },
    {
      "epoch": 17.71075364909145,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0011816029642916575,
      "loss": 0.7913,
      "num_input_tokens_seen": 69040048,
      "step": 118910
    },
    {
      "epoch": 17.711498361632408,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0011808446139456663,
      "loss": 0.7943,
      "num_input_tokens_seen": 69042896,
      "step": 118915
    },
    {
      "epoch": 17.71224307417337,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0011800864970602592,
      "loss": 0.8066,
      "num_input_tokens_seen": 69046032,
      "step": 118920
    },
    {
      "epoch": 17.71298778671433,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0011793286136482478,
      "loss": 0.7758,
      "num_input_tokens_seen": 69048944,
      "step": 118925
    },
    {
      "epoch": 17.713732499255286,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.00117857096372243,
      "loss": 0.7965,
      "num_input_tokens_seen": 69052080,
      "step": 118930
    },
    {
      "epoch": 17.714477211796247,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0011778135472956107,
      "loss": 0.7931,
      "num_input_tokens_seen": 69054960,
      "step": 118935
    },
    {
      "epoch": 17.715221924337207,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.001177056364380583,
      "loss": 0.7992,
      "num_input_tokens_seen": 69058256,
      "step": 118940
    },
    {
      "epoch": 17.715966636878164,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0011762994149901384,
      "loss": 0.7784,
      "num_input_tokens_seen": 69061072,
      "step": 118945
    },
    {
      "epoch": 17.716711349419125,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.0011755426991370633,
      "loss": 0.7929,
      "num_input_tokens_seen": 69064304,
      "step": 118950
    },
    {
      "epoch": 17.717456061960082,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0011747862168341471,
      "loss": 0.7826,
      "num_input_tokens_seen": 69067088,
      "step": 118955
    },
    {
      "epoch": 17.718200774501042,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0011740299680941652,
      "loss": 0.7781,
      "num_input_tokens_seen": 69069872,
      "step": 118960
    },
    {
      "epoch": 17.718945487042003,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0011732739529298935,
      "loss": 0.7999,
      "num_input_tokens_seen": 69072432,
      "step": 118965
    },
    {
      "epoch": 17.71969019958296,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0011725181713541071,
      "loss": 0.8037,
      "num_input_tokens_seen": 69075280,
      "step": 118970
    },
    {
      "epoch": 17.72043491212392,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0011717626233795709,
      "loss": 0.8458,
      "num_input_tokens_seen": 69078192,
      "step": 118975
    },
    {
      "epoch": 17.72117962466488,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0011710073090190525,
      "loss": 0.8064,
      "num_input_tokens_seen": 69081392,
      "step": 118980
    },
    {
      "epoch": 17.721924337205838,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.001170252228285309,
      "loss": 0.8017,
      "num_input_tokens_seen": 69084016,
      "step": 118985
    },
    {
      "epoch": 17.7226690497468,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0011694973811910995,
      "loss": 0.801,
      "num_input_tokens_seen": 69087088,
      "step": 118990
    },
    {
      "epoch": 17.723413762287755,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0011687427677491763,
      "loss": 0.7831,
      "num_input_tokens_seen": 69090064,
      "step": 118995
    },
    {
      "epoch": 17.724158474828716,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0011679883879722834,
      "loss": 0.8016,
      "num_input_tokens_seen": 69092976,
      "step": 119000
    },
    {
      "epoch": 17.724903187369677,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0011672342418731713,
      "loss": 0.7956,
      "num_input_tokens_seen": 69096240,
      "step": 119005
    },
    {
      "epoch": 17.725647899910633,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0011664803294645775,
      "loss": 0.7934,
      "num_input_tokens_seen": 69099248,
      "step": 119010
    },
    {
      "epoch": 17.726392612451594,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0011657266507592373,
      "loss": 0.8057,
      "num_input_tokens_seen": 69102288,
      "step": 119015
    },
    {
      "epoch": 17.727137324992555,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0011649732057698835,
      "loss": 0.8005,
      "num_input_tokens_seen": 69105392,
      "step": 119020
    },
    {
      "epoch": 17.72788203753351,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0011642199945092497,
      "loss": 0.8055,
      "num_input_tokens_seen": 69108016,
      "step": 119025
    },
    {
      "epoch": 17.728626750074472,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0011634670169900535,
      "loss": 0.7989,
      "num_input_tokens_seen": 69110704,
      "step": 119030
    },
    {
      "epoch": 17.72937146261543,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0011627142732250216,
      "loss": 0.7895,
      "num_input_tokens_seen": 69113744,
      "step": 119035
    },
    {
      "epoch": 17.73011617515639,
      "grad_norm": 0.083984375,
      "learning_rate": 0.001161961763226869,
      "loss": 0.7774,
      "num_input_tokens_seen": 69116784,
      "step": 119040
    },
    {
      "epoch": 17.73086088769735,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.001161209487008304,
      "loss": 0.7865,
      "num_input_tokens_seen": 69119856,
      "step": 119045
    },
    {
      "epoch": 17.731605600238307,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0011604574445820443,
      "loss": 0.8092,
      "num_input_tokens_seen": 69123120,
      "step": 119050
    },
    {
      "epoch": 17.732350312779268,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0011597056359607864,
      "loss": 0.7825,
      "num_input_tokens_seen": 69126096,
      "step": 119055
    },
    {
      "epoch": 17.733095025320225,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0011589540611572373,
      "loss": 0.783,
      "num_input_tokens_seen": 69128912,
      "step": 119060
    },
    {
      "epoch": 17.733839737861185,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.00115820272018409,
      "loss": 0.7952,
      "num_input_tokens_seen": 69131920,
      "step": 119065
    },
    {
      "epoch": 17.734584450402146,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0011574516130540423,
      "loss": 0.81,
      "num_input_tokens_seen": 69134608,
      "step": 119070
    },
    {
      "epoch": 17.735329162943103,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0011567007397797811,
      "loss": 0.7901,
      "num_input_tokens_seen": 69137616,
      "step": 119075
    },
    {
      "epoch": 17.736073875484063,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0011559501003739897,
      "loss": 0.7907,
      "num_input_tokens_seen": 69140400,
      "step": 119080
    },
    {
      "epoch": 17.736818588025024,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0011551996948493525,
      "loss": 0.7915,
      "num_input_tokens_seen": 69143248,
      "step": 119085
    },
    {
      "epoch": 17.73756330056598,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0011544495232185413,
      "loss": 0.8009,
      "num_input_tokens_seen": 69146288,
      "step": 119090
    },
    {
      "epoch": 17.73830801310694,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0011536995854942377,
      "loss": 0.8068,
      "num_input_tokens_seen": 69149136,
      "step": 119095
    },
    {
      "epoch": 17.7390527256479,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.0011529498816891031,
      "loss": 0.7903,
      "num_input_tokens_seen": 69151952,
      "step": 119100
    },
    {
      "epoch": 17.73979743818886,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.001152200411815809,
      "loss": 0.7873,
      "num_input_tokens_seen": 69154576,
      "step": 119105
    },
    {
      "epoch": 17.74054215072982,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0011514511758870122,
      "loss": 0.8059,
      "num_input_tokens_seen": 69157680,
      "step": 119110
    },
    {
      "epoch": 17.741286863270776,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0011507021739153756,
      "loss": 0.786,
      "num_input_tokens_seen": 69160528,
      "step": 119115
    },
    {
      "epoch": 17.742031575811737,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0011499534059135508,
      "loss": 0.8028,
      "num_input_tokens_seen": 69163184,
      "step": 119120
    },
    {
      "epoch": 17.742776288352697,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0011492048718941827,
      "loss": 0.8182,
      "num_input_tokens_seen": 69166096,
      "step": 119125
    },
    {
      "epoch": 17.743521000893654,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0011484565718699246,
      "loss": 0.7961,
      "num_input_tokens_seen": 69169072,
      "step": 119130
    },
    {
      "epoch": 17.744265713434615,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0011477085058534113,
      "loss": 0.7942,
      "num_input_tokens_seen": 69171728,
      "step": 119135
    },
    {
      "epoch": 17.745010425975572,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0011469606738572862,
      "loss": 0.7929,
      "num_input_tokens_seen": 69174352,
      "step": 119140
    },
    {
      "epoch": 17.745755138516532,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0011462130758941823,
      "loss": 0.7899,
      "num_input_tokens_seen": 69176816,
      "step": 119145
    },
    {
      "epoch": 17.746499851057493,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0011454657119767263,
      "loss": 0.7866,
      "num_input_tokens_seen": 69179920,
      "step": 119150
    },
    {
      "epoch": 17.74724456359845,
      "grad_norm": 0.056640625,
      "learning_rate": 0.001144718582117548,
      "loss": 0.7724,
      "num_input_tokens_seen": 69182960,
      "step": 119155
    },
    {
      "epoch": 17.74798927613941,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0011439716863292627,
      "loss": 0.7832,
      "num_input_tokens_seen": 69186192,
      "step": 119160
    },
    {
      "epoch": 17.74873398868037,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0011432250246244979,
      "loss": 0.7975,
      "num_input_tokens_seen": 69188752,
      "step": 119165
    },
    {
      "epoch": 17.749478701221328,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.001142478597015859,
      "loss": 0.801,
      "num_input_tokens_seen": 69192048,
      "step": 119170
    },
    {
      "epoch": 17.75022341376229,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0011417324035159625,
      "loss": 0.8072,
      "num_input_tokens_seen": 69194928,
      "step": 119175
    },
    {
      "epoch": 17.750968126303245,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00114098644413741,
      "loss": 0.7895,
      "num_input_tokens_seen": 69198064,
      "step": 119180
    },
    {
      "epoch": 17.751712838844206,
      "grad_norm": 0.060546875,
      "learning_rate": 0.0011402407188928082,
      "loss": 0.8008,
      "num_input_tokens_seen": 69200912,
      "step": 119185
    },
    {
      "epoch": 17.752457551385167,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.0011394952277947517,
      "loss": 0.8334,
      "num_input_tokens_seen": 69203536,
      "step": 119190
    },
    {
      "epoch": 17.753202263926124,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0011387499708558374,
      "loss": 0.7945,
      "num_input_tokens_seen": 69206384,
      "step": 119195
    },
    {
      "epoch": 17.753946976467084,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0011380049480886533,
      "loss": 0.7836,
      "num_input_tokens_seen": 69209040,
      "step": 119200
    },
    {
      "epoch": 17.75469168900804,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0011372601595057861,
      "loss": 0.7924,
      "num_input_tokens_seen": 69212176,
      "step": 119205
    },
    {
      "epoch": 17.755436401549,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.001136515605119821,
      "loss": 0.812,
      "num_input_tokens_seen": 69214800,
      "step": 119210
    },
    {
      "epoch": 17.756181114089962,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.001135771284943336,
      "loss": 0.7823,
      "num_input_tokens_seen": 69217584,
      "step": 119215
    },
    {
      "epoch": 17.75692582663092,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0011350271989889027,
      "loss": 0.8027,
      "num_input_tokens_seen": 69220592,
      "step": 119220
    },
    {
      "epoch": 17.75767053917188,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0011342833472690894,
      "loss": 0.7991,
      "num_input_tokens_seen": 69223536,
      "step": 119225
    },
    {
      "epoch": 17.75841525171284,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0011335397297964711,
      "loss": 0.7967,
      "num_input_tokens_seen": 69226320,
      "step": 119230
    },
    {
      "epoch": 17.759159964253797,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0011327963465836027,
      "loss": 0.8014,
      "num_input_tokens_seen": 69229360,
      "step": 119235
    },
    {
      "epoch": 17.759904676794758,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0011320531976430476,
      "loss": 0.8001,
      "num_input_tokens_seen": 69232368,
      "step": 119240
    },
    {
      "epoch": 17.76064938933572,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0011313102829873605,
      "loss": 0.7989,
      "num_input_tokens_seen": 69235792,
      "step": 119245
    },
    {
      "epoch": 17.761394101876675,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0011305676026290867,
      "loss": 0.791,
      "num_input_tokens_seen": 69238704,
      "step": 119250
    },
    {
      "epoch": 17.762138814417636,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.001129825156580781,
      "loss": 0.7985,
      "num_input_tokens_seen": 69241744,
      "step": 119255
    },
    {
      "epoch": 17.762883526958593,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0011290829448549783,
      "loss": 0.7905,
      "num_input_tokens_seen": 69244528,
      "step": 119260
    },
    {
      "epoch": 17.763628239499553,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0011283409674642253,
      "loss": 0.8012,
      "num_input_tokens_seen": 69247440,
      "step": 119265
    },
    {
      "epoch": 17.764372952040514,
      "grad_norm": 0.38671875,
      "learning_rate": 0.001127599224421052,
      "loss": 0.8286,
      "num_input_tokens_seen": 69250256,
      "step": 119270
    },
    {
      "epoch": 17.76511766458147,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0011268577157379916,
      "loss": 0.7986,
      "num_input_tokens_seen": 69253104,
      "step": 119275
    },
    {
      "epoch": 17.76586237712243,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0011261164414275676,
      "loss": 0.8013,
      "num_input_tokens_seen": 69255792,
      "step": 119280
    },
    {
      "epoch": 17.76660708966339,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0011253754015023065,
      "loss": 0.8108,
      "num_input_tokens_seen": 69258768,
      "step": 119285
    },
    {
      "epoch": 17.76735180220435,
      "grad_norm": 0.0286865234375,
      "learning_rate": 0.0011246345959747283,
      "loss": 0.7936,
      "num_input_tokens_seen": 69261616,
      "step": 119290
    },
    {
      "epoch": 17.76809651474531,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0011238940248573415,
      "loss": 0.8043,
      "num_input_tokens_seen": 69264656,
      "step": 119295
    },
    {
      "epoch": 17.768841227286266,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0011231536881626657,
      "loss": 0.7941,
      "num_input_tokens_seen": 69267920,
      "step": 119300
    },
    {
      "epoch": 17.769585939827227,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.001122413585903203,
      "loss": 0.8003,
      "num_input_tokens_seen": 69270896,
      "step": 119305
    },
    {
      "epoch": 17.770330652368187,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0011216737180914599,
      "loss": 0.7909,
      "num_input_tokens_seen": 69274096,
      "step": 119310
    },
    {
      "epoch": 17.771075364909144,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0011209340847399313,
      "loss": 0.7867,
      "num_input_tokens_seen": 69276880,
      "step": 119315
    },
    {
      "epoch": 17.771820077450105,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.001120194685861119,
      "loss": 0.7827,
      "num_input_tokens_seen": 69279856,
      "step": 119320
    },
    {
      "epoch": 17.772564789991062,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0011194555214675095,
      "loss": 0.7805,
      "num_input_tokens_seen": 69282832,
      "step": 119325
    },
    {
      "epoch": 17.773309502532022,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.0011187165915715895,
      "loss": 0.7576,
      "num_input_tokens_seen": 69285552,
      "step": 119330
    },
    {
      "epoch": 17.774054215072983,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.001117977896185846,
      "loss": 0.8153,
      "num_input_tokens_seen": 69288464,
      "step": 119335
    },
    {
      "epoch": 17.77479892761394,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0011172394353227589,
      "loss": 0.8054,
      "num_input_tokens_seen": 69291280,
      "step": 119340
    },
    {
      "epoch": 17.7755436401549,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0011165012089947995,
      "loss": 0.8017,
      "num_input_tokens_seen": 69294320,
      "step": 119345
    },
    {
      "epoch": 17.77628835269586,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0011157632172144399,
      "loss": 0.792,
      "num_input_tokens_seen": 69297296,
      "step": 119350
    },
    {
      "epoch": 17.777033065236818,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0011150254599941517,
      "loss": 0.8009,
      "num_input_tokens_seen": 69300528,
      "step": 119355
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0011142879373463965,
      "loss": 0.7885,
      "num_input_tokens_seen": 69303344,
      "step": 119360
    },
    {
      "epoch": 17.778522490318736,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0011135506492836328,
      "loss": 0.7984,
      "num_input_tokens_seen": 69306000,
      "step": 119365
    },
    {
      "epoch": 17.779267202859696,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.001112813595818317,
      "loss": 0.8073,
      "num_input_tokens_seen": 69309008,
      "step": 119370
    },
    {
      "epoch": 17.780011915400657,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0011120767769629014,
      "loss": 0.7882,
      "num_input_tokens_seen": 69311888,
      "step": 119375
    },
    {
      "epoch": 17.780756627941614,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0011113401927298338,
      "loss": 0.787,
      "num_input_tokens_seen": 69314768,
      "step": 119380
    },
    {
      "epoch": 17.781501340482574,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0011106038431315563,
      "loss": 0.7764,
      "num_input_tokens_seen": 69317808,
      "step": 119385
    },
    {
      "epoch": 17.782246053023535,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0011098677281805136,
      "loss": 0.8023,
      "num_input_tokens_seen": 69320784,
      "step": 119390
    },
    {
      "epoch": 17.78299076556449,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0011091318478891343,
      "loss": 0.7762,
      "num_input_tokens_seen": 69323472,
      "step": 119395
    },
    {
      "epoch": 17.783735478105452,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0011083962022698584,
      "loss": 0.8082,
      "num_input_tokens_seen": 69326352,
      "step": 119400
    },
    {
      "epoch": 17.78448019064641,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0011076607913351093,
      "loss": 0.8238,
      "num_input_tokens_seen": 69329136,
      "step": 119405
    },
    {
      "epoch": 17.78522490318737,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0011069256150973105,
      "loss": 0.8007,
      "num_input_tokens_seen": 69331952,
      "step": 119410
    },
    {
      "epoch": 17.78596961572833,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0011061906735688853,
      "loss": 0.8286,
      "num_input_tokens_seen": 69334768,
      "step": 119415
    },
    {
      "epoch": 17.786714328269287,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0011054559667622438,
      "loss": 0.8106,
      "num_input_tokens_seen": 69337552,
      "step": 119420
    },
    {
      "epoch": 17.787459040810248,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0011047214946898042,
      "loss": 0.789,
      "num_input_tokens_seen": 69340656,
      "step": 119425
    },
    {
      "epoch": 17.788203753351205,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.0011039872573639703,
      "loss": 0.8079,
      "num_input_tokens_seen": 69343536,
      "step": 119430
    },
    {
      "epoch": 17.788948465892165,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0011032532547971502,
      "loss": 0.8323,
      "num_input_tokens_seen": 69346480,
      "step": 119435
    },
    {
      "epoch": 17.789693178433126,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.001102519487001744,
      "loss": 0.7938,
      "num_input_tokens_seen": 69349584,
      "step": 119440
    },
    {
      "epoch": 17.790437890974083,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.001101785953990142,
      "loss": 0.8231,
      "num_input_tokens_seen": 69352688,
      "step": 119445
    },
    {
      "epoch": 17.791182603515043,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0011010526557747441,
      "loss": 0.7999,
      "num_input_tokens_seen": 69355504,
      "step": 119450
    },
    {
      "epoch": 17.791927316056004,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.0011003195923679338,
      "loss": 0.7852,
      "num_input_tokens_seen": 69358192,
      "step": 119455
    },
    {
      "epoch": 17.79267202859696,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0010995867637820977,
      "loss": 0.8306,
      "num_input_tokens_seen": 69360880,
      "step": 119460
    },
    {
      "epoch": 17.79341674113792,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0010988541700296145,
      "loss": 0.7797,
      "num_input_tokens_seen": 69363696,
      "step": 119465
    },
    {
      "epoch": 17.79416145367888,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0010981218111228623,
      "loss": 0.8107,
      "num_input_tokens_seen": 69366512,
      "step": 119470
    },
    {
      "epoch": 17.79490616621984,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.001097389687074215,
      "loss": 0.815,
      "num_input_tokens_seen": 69369296,
      "step": 119475
    },
    {
      "epoch": 17.7956508787608,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0010966577978960372,
      "loss": 0.7957,
      "num_input_tokens_seen": 69372080,
      "step": 119480
    },
    {
      "epoch": 17.796395591301756,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0010959261436006961,
      "loss": 0.7746,
      "num_input_tokens_seen": 69374672,
      "step": 119485
    },
    {
      "epoch": 17.797140303842717,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0010951947242005483,
      "loss": 0.7828,
      "num_input_tokens_seen": 69377584,
      "step": 119490
    },
    {
      "epoch": 17.797885016383677,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.001094463539707957,
      "loss": 0.8097,
      "num_input_tokens_seen": 69380240,
      "step": 119495
    },
    {
      "epoch": 17.798629728924634,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.001093732590135268,
      "loss": 0.7784,
      "num_input_tokens_seen": 69382992,
      "step": 119500
    },
    {
      "epoch": 17.799374441465595,
      "grad_norm": 0.0625,
      "learning_rate": 0.0010930018754948356,
      "loss": 0.8041,
      "num_input_tokens_seen": 69385584,
      "step": 119505
    },
    {
      "epoch": 17.800119154006552,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0010922713957989987,
      "loss": 0.7717,
      "num_input_tokens_seen": 69388720,
      "step": 119510
    },
    {
      "epoch": 17.800863866547513,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0010915411510601057,
      "loss": 0.7865,
      "num_input_tokens_seen": 69391920,
      "step": 119515
    },
    {
      "epoch": 17.801608579088473,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0010908111412904885,
      "loss": 0.8015,
      "num_input_tokens_seen": 69394576,
      "step": 119520
    },
    {
      "epoch": 17.80235329162943,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0010900813665024771,
      "loss": 0.8039,
      "num_input_tokens_seen": 69397424,
      "step": 119525
    },
    {
      "epoch": 17.80309800417039,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0010893518267084067,
      "loss": 0.7924,
      "num_input_tokens_seen": 69400400,
      "step": 119530
    },
    {
      "epoch": 17.80384271671135,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0010886225219205957,
      "loss": 0.7796,
      "num_input_tokens_seen": 69403152,
      "step": 119535
    },
    {
      "epoch": 17.804587429252308,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0010878934521513727,
      "loss": 0.7798,
      "num_input_tokens_seen": 69406064,
      "step": 119540
    },
    {
      "epoch": 17.80533214179327,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0010871646174130478,
      "loss": 0.7921,
      "num_input_tokens_seen": 69408720,
      "step": 119545
    },
    {
      "epoch": 17.806076854334226,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0010864360177179377,
      "loss": 0.7942,
      "num_input_tokens_seen": 69411600,
      "step": 119550
    },
    {
      "epoch": 17.806821566875186,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0010857076530783477,
      "loss": 0.8057,
      "num_input_tokens_seen": 69414416,
      "step": 119555
    },
    {
      "epoch": 17.807566279416147,
      "grad_norm": 0.048828125,
      "learning_rate": 0.001084979523506588,
      "loss": 0.7799,
      "num_input_tokens_seen": 69417424,
      "step": 119560
    },
    {
      "epoch": 17.808310991957104,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.001084251629014955,
      "loss": 0.7955,
      "num_input_tokens_seen": 69420688,
      "step": 119565
    },
    {
      "epoch": 17.809055704498064,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.001083523969615746,
      "loss": 0.8095,
      "num_input_tokens_seen": 69423376,
      "step": 119570
    },
    {
      "epoch": 17.80980041703902,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0010827965453212578,
      "loss": 0.8052,
      "num_input_tokens_seen": 69426128,
      "step": 119575
    },
    {
      "epoch": 17.81054512957998,
      "grad_norm": 0.02490234375,
      "learning_rate": 0.0010820693561437737,
      "loss": 0.8013,
      "num_input_tokens_seen": 69429040,
      "step": 119580
    },
    {
      "epoch": 17.811289842120942,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0010813424020955859,
      "loss": 0.8061,
      "num_input_tokens_seen": 69431856,
      "step": 119585
    },
    {
      "epoch": 17.8120345546619,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.001080615683188969,
      "loss": 0.7943,
      "num_input_tokens_seen": 69434768,
      "step": 119590
    },
    {
      "epoch": 17.81277926720286,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0010798891994362053,
      "loss": 0.7843,
      "num_input_tokens_seen": 69437456,
      "step": 119595
    },
    {
      "epoch": 17.81352397974382,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0010791629508495664,
      "loss": 0.8244,
      "num_input_tokens_seen": 69440432,
      "step": 119600
    },
    {
      "epoch": 17.814268692284777,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.001078436937441321,
      "loss": 0.7744,
      "num_input_tokens_seen": 69443472,
      "step": 119605
    },
    {
      "epoch": 17.815013404825738,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0010777111592237308,
      "loss": 0.7922,
      "num_input_tokens_seen": 69446736,
      "step": 119610
    },
    {
      "epoch": 17.815758117366695,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0010769856162090646,
      "loss": 0.8017,
      "num_input_tokens_seen": 69449520,
      "step": 119615
    },
    {
      "epoch": 17.816502829907655,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0010762603084095755,
      "loss": 0.8128,
      "num_input_tokens_seen": 69452432,
      "step": 119620
    },
    {
      "epoch": 17.817247542448616,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.001075535235837514,
      "loss": 0.7789,
      "num_input_tokens_seen": 69455152,
      "step": 119625
    },
    {
      "epoch": 17.817992254989573,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0010748103985051339,
      "loss": 0.7967,
      "num_input_tokens_seen": 69457936,
      "step": 119630
    },
    {
      "epoch": 17.818736967530533,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.001074085796424678,
      "loss": 0.785,
      "num_input_tokens_seen": 69460912,
      "step": 119635
    },
    {
      "epoch": 17.819481680071494,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0010733614296083922,
      "loss": 0.7976,
      "num_input_tokens_seen": 69463472,
      "step": 119640
    },
    {
      "epoch": 17.82022639261245,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0010726372980685095,
      "loss": 0.7985,
      "num_input_tokens_seen": 69466384,
      "step": 119645
    },
    {
      "epoch": 17.82097110515341,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.001071913401817262,
      "loss": 0.7947,
      "num_input_tokens_seen": 69469392,
      "step": 119650
    },
    {
      "epoch": 17.82171581769437,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0010711897408668834,
      "loss": 0.8,
      "num_input_tokens_seen": 69472272,
      "step": 119655
    },
    {
      "epoch": 17.82246053023533,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0010704663152295956,
      "loss": 0.7937,
      "num_input_tokens_seen": 69475696,
      "step": 119660
    },
    {
      "epoch": 17.82320524277629,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0010697431249176252,
      "loss": 0.793,
      "num_input_tokens_seen": 69478704,
      "step": 119665
    },
    {
      "epoch": 17.823949955317246,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0010690201699431862,
      "loss": 0.7935,
      "num_input_tokens_seen": 69481648,
      "step": 119670
    },
    {
      "epoch": 17.824694667858207,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.0010682974503184934,
      "loss": 0.7906,
      "num_input_tokens_seen": 69484400,
      "step": 119675
    },
    {
      "epoch": 17.825439380399168,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0010675749660557553,
      "loss": 0.8039,
      "num_input_tokens_seen": 69487312,
      "step": 119680
    },
    {
      "epoch": 17.826184092940125,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0010668527171671742,
      "loss": 0.8101,
      "num_input_tokens_seen": 69490256,
      "step": 119685
    },
    {
      "epoch": 17.826928805481085,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0010661307036649585,
      "loss": 0.8123,
      "num_input_tokens_seen": 69493072,
      "step": 119690
    },
    {
      "epoch": 17.827673518022042,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0010654089255613002,
      "loss": 0.8012,
      "num_input_tokens_seen": 69495696,
      "step": 119695
    },
    {
      "epoch": 17.828418230563003,
      "grad_norm": 0.2197265625,
      "learning_rate": 0.0010646873828683995,
      "loss": 0.8326,
      "num_input_tokens_seen": 69498576,
      "step": 119700
    },
    {
      "epoch": 17.829162943103963,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.001063966075598438,
      "loss": 0.7948,
      "num_input_tokens_seen": 69501648,
      "step": 119705
    },
    {
      "epoch": 17.82990765564492,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.0010632450037636099,
      "loss": 0.7903,
      "num_input_tokens_seen": 69504688,
      "step": 119710
    },
    {
      "epoch": 17.83065236818588,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0010625241673760883,
      "loss": 0.7991,
      "num_input_tokens_seen": 69507536,
      "step": 119715
    },
    {
      "epoch": 17.83139708072684,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0010618035664480584,
      "loss": 0.7962,
      "num_input_tokens_seen": 69510544,
      "step": 119720
    },
    {
      "epoch": 17.832141793267798,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0010610832009916927,
      "loss": 0.7974,
      "num_input_tokens_seen": 69513456,
      "step": 119725
    },
    {
      "epoch": 17.83288650580876,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0010603630710191558,
      "loss": 0.8124,
      "num_input_tokens_seen": 69516304,
      "step": 119730
    },
    {
      "epoch": 17.833631218349716,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.00105964317654262,
      "loss": 0.7937,
      "num_input_tokens_seen": 69519088,
      "step": 119735
    },
    {
      "epoch": 17.834375930890676,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.001058923517574244,
      "loss": 0.805,
      "num_input_tokens_seen": 69521936,
      "step": 119740
    },
    {
      "epoch": 17.835120643431637,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.001058204094126186,
      "loss": 0.7948,
      "num_input_tokens_seen": 69525104,
      "step": 119745
    },
    {
      "epoch": 17.835865355972594,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0010574849062105983,
      "loss": 0.7955,
      "num_input_tokens_seen": 69528112,
      "step": 119750
    },
    {
      "epoch": 17.836610068513554,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0010567659538396345,
      "loss": 0.7884,
      "num_input_tokens_seen": 69530608,
      "step": 119755
    },
    {
      "epoch": 17.837354781054515,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0010560472370254382,
      "loss": 0.8023,
      "num_input_tokens_seen": 69533680,
      "step": 119760
    },
    {
      "epoch": 17.83809949359547,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0010553287557801493,
      "loss": 0.8093,
      "num_input_tokens_seen": 69536400,
      "step": 119765
    },
    {
      "epoch": 17.838844206136432,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.001054610510115912,
      "loss": 0.7839,
      "num_input_tokens_seen": 69539216,
      "step": 119770
    },
    {
      "epoch": 17.83958891867739,
      "grad_norm": 0.078125,
      "learning_rate": 0.0010538925000448528,
      "loss": 0.8251,
      "num_input_tokens_seen": 69542096,
      "step": 119775
    },
    {
      "epoch": 17.84033363121835,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.001053174725579109,
      "loss": 0.7915,
      "num_input_tokens_seen": 69544912,
      "step": 119780
    },
    {
      "epoch": 17.84107834375931,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0010524571867308007,
      "loss": 0.8059,
      "num_input_tokens_seen": 69547760,
      "step": 119785
    },
    {
      "epoch": 17.841823056300267,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.001051739883512055,
      "loss": 0.8094,
      "num_input_tokens_seen": 69550640,
      "step": 119790
    },
    {
      "epoch": 17.842567768841228,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0010510228159349854,
      "loss": 0.7988,
      "num_input_tokens_seen": 69553520,
      "step": 119795
    },
    {
      "epoch": 17.843312481382185,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.0010503059840117107,
      "loss": 0.805,
      "num_input_tokens_seen": 69556560,
      "step": 119800
    },
    {
      "epoch": 17.844057193923145,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0010495893877543376,
      "loss": 0.8069,
      "num_input_tokens_seen": 69559472,
      "step": 119805
    },
    {
      "epoch": 17.844801906464106,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0010488730271749752,
      "loss": 0.8118,
      "num_input_tokens_seen": 69562416,
      "step": 119810
    },
    {
      "epoch": 17.845546619005063,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.0010481569022857218,
      "loss": 0.7748,
      "num_input_tokens_seen": 69565328,
      "step": 119815
    },
    {
      "epoch": 17.846291331546023,
      "grad_norm": 0.060546875,
      "learning_rate": 0.001047441013098676,
      "loss": 0.8027,
      "num_input_tokens_seen": 69568208,
      "step": 119820
    },
    {
      "epoch": 17.847036044086984,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.001046725359625935,
      "loss": 0.7984,
      "num_input_tokens_seen": 69571344,
      "step": 119825
    },
    {
      "epoch": 17.84778075662794,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.001046009941879586,
      "loss": 0.7911,
      "num_input_tokens_seen": 69574288,
      "step": 119830
    },
    {
      "epoch": 17.8485254691689,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0010452947598717187,
      "loss": 0.8098,
      "num_input_tokens_seen": 69577296,
      "step": 119835
    },
    {
      "epoch": 17.84927018170986,
      "grad_norm": 0.078125,
      "learning_rate": 0.0010445798136144123,
      "loss": 0.797,
      "num_input_tokens_seen": 69580144,
      "step": 119840
    },
    {
      "epoch": 17.85001489425082,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.001043865103119747,
      "loss": 0.8031,
      "num_input_tokens_seen": 69582992,
      "step": 119845
    },
    {
      "epoch": 17.85075960679178,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0010431506283997982,
      "loss": 0.8048,
      "num_input_tokens_seen": 69586032,
      "step": 119850
    },
    {
      "epoch": 17.851504319332737,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.0010424363894666311,
      "loss": 0.7907,
      "num_input_tokens_seen": 69589040,
      "step": 119855
    },
    {
      "epoch": 17.852249031873697,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0010417223863323176,
      "loss": 0.7938,
      "num_input_tokens_seen": 69591920,
      "step": 119860
    },
    {
      "epoch": 17.852993744414658,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.001041008619008915,
      "loss": 0.7928,
      "num_input_tokens_seen": 69594640,
      "step": 119865
    },
    {
      "epoch": 17.853738456955615,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0010402950875084871,
      "loss": 0.7919,
      "num_input_tokens_seen": 69597616,
      "step": 119870
    },
    {
      "epoch": 17.854483169496575,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0010395817918430871,
      "loss": 0.794,
      "num_input_tokens_seen": 69600368,
      "step": 119875
    },
    {
      "epoch": 17.855227882037532,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0010388687320247625,
      "loss": 0.8135,
      "num_input_tokens_seen": 69603312,
      "step": 119880
    },
    {
      "epoch": 17.855972594578493,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0010381559080655616,
      "loss": 0.821,
      "num_input_tokens_seen": 69606064,
      "step": 119885
    },
    {
      "epoch": 17.856717307119453,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.001037443319977525,
      "loss": 0.8056,
      "num_input_tokens_seen": 69608976,
      "step": 119890
    },
    {
      "epoch": 17.85746201966041,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0010367309677726932,
      "loss": 0.7945,
      "num_input_tokens_seen": 69611984,
      "step": 119895
    },
    {
      "epoch": 17.85820673220137,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0010360188514630996,
      "loss": 0.8036,
      "num_input_tokens_seen": 69614704,
      "step": 119900
    },
    {
      "epoch": 17.85895144474233,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0010353069710607764,
      "loss": 0.7988,
      "num_input_tokens_seen": 69617488,
      "step": 119905
    },
    {
      "epoch": 17.859696157283288,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0010345953265777457,
      "loss": 0.7851,
      "num_input_tokens_seen": 69620368,
      "step": 119910
    },
    {
      "epoch": 17.86044086982425,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0010338839180260362,
      "loss": 0.7868,
      "num_input_tokens_seen": 69623472,
      "step": 119915
    },
    {
      "epoch": 17.861185582365206,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0010331727454176614,
      "loss": 0.8027,
      "num_input_tokens_seen": 69626192,
      "step": 119920
    },
    {
      "epoch": 17.861930294906166,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0010324618087646403,
      "loss": 0.8055,
      "num_input_tokens_seen": 69629200,
      "step": 119925
    },
    {
      "epoch": 17.862675007447127,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0010317511080789816,
      "loss": 0.7951,
      "num_input_tokens_seen": 69632112,
      "step": 119930
    },
    {
      "epoch": 17.863419719988084,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0010310406433726887,
      "loss": 0.7828,
      "num_input_tokens_seen": 69635088,
      "step": 119935
    },
    {
      "epoch": 17.864164432529044,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.0010303304146577674,
      "loss": 0.7965,
      "num_input_tokens_seen": 69638160,
      "step": 119940
    },
    {
      "epoch": 17.86490914507,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0010296204219462162,
      "loss": 0.8046,
      "num_input_tokens_seen": 69640752,
      "step": 119945
    },
    {
      "epoch": 17.865653857610962,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.001028910665250029,
      "loss": 0.7899,
      "num_input_tokens_seen": 69643824,
      "step": 119950
    },
    {
      "epoch": 17.866398570151922,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0010282011445811945,
      "loss": 0.7906,
      "num_input_tokens_seen": 69646768,
      "step": 119955
    },
    {
      "epoch": 17.86714328269288,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0010274918599517046,
      "loss": 0.7791,
      "num_input_tokens_seen": 69649552,
      "step": 119960
    },
    {
      "epoch": 17.86788799523384,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.0010267828113735367,
      "loss": 0.794,
      "num_input_tokens_seen": 69652400,
      "step": 119965
    },
    {
      "epoch": 17.8686327077748,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0010260739988586709,
      "loss": 0.7986,
      "num_input_tokens_seen": 69655120,
      "step": 119970
    },
    {
      "epoch": 17.869377420315757,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0010253654224190844,
      "loss": 0.8078,
      "num_input_tokens_seen": 69658480,
      "step": 119975
    },
    {
      "epoch": 17.870122132856718,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0010246570820667427,
      "loss": 0.7919,
      "num_input_tokens_seen": 69661552,
      "step": 119980
    },
    {
      "epoch": 17.870866845397675,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0010239489778136196,
      "loss": 0.7789,
      "num_input_tokens_seen": 69664368,
      "step": 119985
    },
    {
      "epoch": 17.871611557938635,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0010232411096716704,
      "loss": 0.8048,
      "num_input_tokens_seen": 69667248,
      "step": 119990
    },
    {
      "epoch": 17.872356270479596,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0010225334776528606,
      "loss": 0.7894,
      "num_input_tokens_seen": 69670160,
      "step": 119995
    },
    {
      "epoch": 17.873100983020553,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0010218260817691405,
      "loss": 0.8172,
      "num_input_tokens_seen": 69673008,
      "step": 120000
    },
    {
      "epoch": 17.873845695561513,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0010211189220324622,
      "loss": 0.8309,
      "num_input_tokens_seen": 69675888,
      "step": 120005
    },
    {
      "epoch": 17.874590408102474,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0010204119984547715,
      "loss": 0.8227,
      "num_input_tokens_seen": 69678608,
      "step": 120010
    },
    {
      "epoch": 17.87533512064343,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0010197053110480119,
      "loss": 0.7756,
      "num_input_tokens_seen": 69681648,
      "step": 120015
    },
    {
      "epoch": 17.87607983318439,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0010189988598241222,
      "loss": 0.7667,
      "num_input_tokens_seen": 69684560,
      "step": 120020
    },
    {
      "epoch": 17.87682454572535,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0010182926447950362,
      "loss": 0.8307,
      "num_input_tokens_seen": 69687696,
      "step": 120025
    },
    {
      "epoch": 17.87756925826631,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.001017586665972686,
      "loss": 0.784,
      "num_input_tokens_seen": 69690448,
      "step": 120030
    },
    {
      "epoch": 17.87831397080727,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0010168809233689973,
      "loss": 0.7899,
      "num_input_tokens_seen": 69693424,
      "step": 120035
    },
    {
      "epoch": 17.879058683348227,
      "grad_norm": 0.044921875,
      "learning_rate": 0.001016175416995897,
      "loss": 0.7867,
      "num_input_tokens_seen": 69696336,
      "step": 120040
    },
    {
      "epoch": 17.879803395889187,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0010154701468652987,
      "loss": 0.7825,
      "num_input_tokens_seen": 69699376,
      "step": 120045
    },
    {
      "epoch": 17.880548108430148,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0010147651129891167,
      "loss": 0.7949,
      "num_input_tokens_seen": 69702128,
      "step": 120050
    },
    {
      "epoch": 17.881292820971105,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0010140603153792676,
      "loss": 0.8002,
      "num_input_tokens_seen": 69704880,
      "step": 120055
    },
    {
      "epoch": 17.882037533512065,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0010133557540476506,
      "loss": 0.7884,
      "num_input_tokens_seen": 69707760,
      "step": 120060
    },
    {
      "epoch": 17.882782246053022,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0010126514290061761,
      "loss": 0.8116,
      "num_input_tokens_seen": 69710672,
      "step": 120065
    },
    {
      "epoch": 17.883526958593983,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.001011947340266741,
      "loss": 0.7984,
      "num_input_tokens_seen": 69713712,
      "step": 120070
    },
    {
      "epoch": 17.884271671134943,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0010112434878412364,
      "loss": 0.8026,
      "num_input_tokens_seen": 69716560,
      "step": 120075
    },
    {
      "epoch": 17.8850163836759,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0010105398717415552,
      "loss": 0.7847,
      "num_input_tokens_seen": 69719408,
      "step": 120080
    },
    {
      "epoch": 17.88576109621686,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0010098364919795854,
      "loss": 0.8129,
      "num_input_tokens_seen": 69722096,
      "step": 120085
    },
    {
      "epoch": 17.886505808757818,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0010091333485672104,
      "loss": 0.7927,
      "num_input_tokens_seen": 69725200,
      "step": 120090
    },
    {
      "epoch": 17.88725052129878,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0010084304415163042,
      "loss": 0.7903,
      "num_input_tokens_seen": 69728368,
      "step": 120095
    },
    {
      "epoch": 17.88799523383974,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0010077277708387489,
      "loss": 0.7995,
      "num_input_tokens_seen": 69731280,
      "step": 120100
    },
    {
      "epoch": 17.888739946380696,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0010070253365464083,
      "loss": 0.8056,
      "num_input_tokens_seen": 69734352,
      "step": 120105
    },
    {
      "epoch": 17.889484658921656,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0010063231386511563,
      "loss": 0.8359,
      "num_input_tokens_seen": 69737200,
      "step": 120110
    },
    {
      "epoch": 17.890229371462617,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0010056211771648483,
      "loss": 0.8118,
      "num_input_tokens_seen": 69740112,
      "step": 120115
    },
    {
      "epoch": 17.890974084003574,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0010049194520993514,
      "loss": 0.7998,
      "num_input_tokens_seen": 69743024,
      "step": 120120
    },
    {
      "epoch": 17.891718796544534,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0010042179634665149,
      "loss": 0.8062,
      "num_input_tokens_seen": 69745840,
      "step": 120125
    },
    {
      "epoch": 17.89246350908549,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0010035167112781905,
      "loss": 0.8011,
      "num_input_tokens_seen": 69748752,
      "step": 120130
    },
    {
      "epoch": 17.893208221626452,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0010028156955462274,
      "loss": 0.7775,
      "num_input_tokens_seen": 69751760,
      "step": 120135
    },
    {
      "epoch": 17.893952934167412,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.001002114916282466,
      "loss": 0.8015,
      "num_input_tokens_seen": 69754672,
      "step": 120140
    },
    {
      "epoch": 17.89469764670837,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0010014143734987484,
      "loss": 0.7935,
      "num_input_tokens_seen": 69757584,
      "step": 120145
    },
    {
      "epoch": 17.89544235924933,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0010007140672069037,
      "loss": 0.7931,
      "num_input_tokens_seen": 69760528,
      "step": 120150
    },
    {
      "epoch": 17.89618707179029,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0010000139974187688,
      "loss": 0.7677,
      "num_input_tokens_seen": 69763216,
      "step": 120155
    },
    {
      "epoch": 17.896931784331247,
      "grad_norm": 0.0625,
      "learning_rate": 0.0009993141641461661,
      "loss": 0.7923,
      "num_input_tokens_seen": 69766160,
      "step": 120160
    },
    {
      "epoch": 17.897676496872208,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0009986145674009245,
      "loss": 0.7893,
      "num_input_tokens_seen": 69768944,
      "step": 120165
    },
    {
      "epoch": 17.898421209413165,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0009979152071948577,
      "loss": 0.7889,
      "num_input_tokens_seen": 69772144,
      "step": 120170
    },
    {
      "epoch": 17.899165921954125,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0009972160835397814,
      "loss": 0.7941,
      "num_input_tokens_seen": 69774992,
      "step": 120175
    },
    {
      "epoch": 17.899910634495086,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0009965171964475095,
      "loss": 0.8176,
      "num_input_tokens_seen": 69777936,
      "step": 120180
    },
    {
      "epoch": 17.900655347036043,
      "grad_norm": 0.07421875,
      "learning_rate": 0.000995818545929844,
      "loss": 0.7966,
      "num_input_tokens_seen": 69780688,
      "step": 120185
    },
    {
      "epoch": 17.901400059577004,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.000995120131998594,
      "loss": 0.793,
      "num_input_tokens_seen": 69783536,
      "step": 120190
    },
    {
      "epoch": 17.902144772117964,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0009944219546655535,
      "loss": 0.7936,
      "num_input_tokens_seen": 69786512,
      "step": 120195
    },
    {
      "epoch": 17.90288948465892,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.0009937240139425213,
      "loss": 0.8316,
      "num_input_tokens_seen": 69789360,
      "step": 120200
    },
    {
      "epoch": 17.90363419719988,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.000993026309841286,
      "loss": 0.812,
      "num_input_tokens_seen": 69792304,
      "step": 120205
    },
    {
      "epoch": 17.90437890974084,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.000992328842373637,
      "loss": 0.8193,
      "num_input_tokens_seen": 69795184,
      "step": 120210
    },
    {
      "epoch": 17.9051236222818,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.0009916316115513545,
      "loss": 0.7887,
      "num_input_tokens_seen": 69798000,
      "step": 120215
    },
    {
      "epoch": 17.90586833482276,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0009909346173862144,
      "loss": 0.8177,
      "num_input_tokens_seen": 69801072,
      "step": 120220
    },
    {
      "epoch": 17.906613047363717,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0009902378598900003,
      "loss": 0.7925,
      "num_input_tokens_seen": 69803568,
      "step": 120225
    },
    {
      "epoch": 17.907357759904677,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0009895413390744762,
      "loss": 0.7966,
      "num_input_tokens_seen": 69806544,
      "step": 120230
    },
    {
      "epoch": 17.908102472445638,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0009888450549514127,
      "loss": 0.7844,
      "num_input_tokens_seen": 69809616,
      "step": 120235
    },
    {
      "epoch": 17.908847184986595,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.0009881490075325705,
      "loss": 0.7784,
      "num_input_tokens_seen": 69812432,
      "step": 120240
    },
    {
      "epoch": 17.909591897527555,
      "grad_norm": 0.234375,
      "learning_rate": 0.00098745319682971,
      "loss": 0.8177,
      "num_input_tokens_seen": 69815344,
      "step": 120245
    },
    {
      "epoch": 17.910336610068512,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0009867576228545887,
      "loss": 0.7876,
      "num_input_tokens_seen": 69818224,
      "step": 120250
    },
    {
      "epoch": 17.911081322609473,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.0009860622856189504,
      "loss": 0.7811,
      "num_input_tokens_seen": 69821104,
      "step": 120255
    },
    {
      "epoch": 17.911826035150433,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0009853671851345507,
      "loss": 0.7938,
      "num_input_tokens_seen": 69823792,
      "step": 120260
    },
    {
      "epoch": 17.91257074769139,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0009846723214131286,
      "loss": 0.796,
      "num_input_tokens_seen": 69826384,
      "step": 120265
    },
    {
      "epoch": 17.91331546023235,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0009839776944664197,
      "loss": 0.8103,
      "num_input_tokens_seen": 69829200,
      "step": 120270
    },
    {
      "epoch": 17.91406017277331,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0009832833043061644,
      "loss": 0.8166,
      "num_input_tokens_seen": 69832048,
      "step": 120275
    },
    {
      "epoch": 17.91480488531427,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0009825891509440937,
      "loss": 0.7756,
      "num_input_tokens_seen": 69834800,
      "step": 120280
    },
    {
      "epoch": 17.91554959785523,
      "grad_norm": 0.060546875,
      "learning_rate": 0.0009818952343919312,
      "loss": 0.8152,
      "num_input_tokens_seen": 69837488,
      "step": 120285
    },
    {
      "epoch": 17.916294310396186,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0009812015546613994,
      "loss": 0.8083,
      "num_input_tokens_seen": 69840368,
      "step": 120290
    },
    {
      "epoch": 17.917039022937146,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0009805081117642207,
      "loss": 0.791,
      "num_input_tokens_seen": 69843280,
      "step": 120295
    },
    {
      "epoch": 17.917783735478107,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.000979814905712107,
      "loss": 0.8082,
      "num_input_tokens_seen": 69846224,
      "step": 120300
    },
    {
      "epoch": 17.918528448019064,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0009791219365167742,
      "loss": 0.8018,
      "num_input_tokens_seen": 69849008,
      "step": 120305
    },
    {
      "epoch": 17.919273160560024,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0009784292041899228,
      "loss": 0.8138,
      "num_input_tokens_seen": 69851792,
      "step": 120310
    },
    {
      "epoch": 17.92001787310098,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0009777367087432604,
      "loss": 0.7976,
      "num_input_tokens_seen": 69854672,
      "step": 120315
    },
    {
      "epoch": 17.920762585641942,
      "grad_norm": 0.052734375,
      "learning_rate": 0.000977044450188484,
      "loss": 0.7976,
      "num_input_tokens_seen": 69857744,
      "step": 120320
    },
    {
      "epoch": 17.921507298182902,
      "grad_norm": 0.189453125,
      "learning_rate": 0.0009763524285372909,
      "loss": 0.8028,
      "num_input_tokens_seen": 69861008,
      "step": 120325
    },
    {
      "epoch": 17.92225201072386,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0009756606438013703,
      "loss": 0.8067,
      "num_input_tokens_seen": 69863728,
      "step": 120330
    },
    {
      "epoch": 17.92299672326482,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0009749690959924112,
      "loss": 0.7877,
      "num_input_tokens_seen": 69866608,
      "step": 120335
    },
    {
      "epoch": 17.92374143580578,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.000974277785122094,
      "loss": 0.8135,
      "num_input_tokens_seen": 69869008,
      "step": 120340
    },
    {
      "epoch": 17.924486148346737,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0009735867112020945,
      "loss": 0.798,
      "num_input_tokens_seen": 69871760,
      "step": 120345
    },
    {
      "epoch": 17.925230860887698,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.0009728958742440952,
      "loss": 0.7852,
      "num_input_tokens_seen": 69874608,
      "step": 120350
    },
    {
      "epoch": 17.925975573428655,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0009722052742597615,
      "loss": 0.8108,
      "num_input_tokens_seen": 69877136,
      "step": 120355
    },
    {
      "epoch": 17.926720285969616,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0009715149112607657,
      "loss": 0.8038,
      "num_input_tokens_seen": 69879952,
      "step": 120360
    },
    {
      "epoch": 17.927464998510576,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0009708247852587654,
      "loss": 0.8035,
      "num_input_tokens_seen": 69882672,
      "step": 120365
    },
    {
      "epoch": 17.928209711051533,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.000970134896265421,
      "loss": 0.8235,
      "num_input_tokens_seen": 69885520,
      "step": 120370
    },
    {
      "epoch": 17.928954423592494,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00096944524429239,
      "loss": 0.7872,
      "num_input_tokens_seen": 69888560,
      "step": 120375
    },
    {
      "epoch": 17.929699136133454,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0009687558293513198,
      "loss": 0.8391,
      "num_input_tokens_seen": 69891472,
      "step": 120380
    },
    {
      "epoch": 17.93044384867441,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.000968066651453861,
      "loss": 0.8224,
      "num_input_tokens_seen": 69894672,
      "step": 120385
    },
    {
      "epoch": 17.93118856121537,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0009673777106116526,
      "loss": 0.7784,
      "num_input_tokens_seen": 69897456,
      "step": 120390
    },
    {
      "epoch": 17.93193327375633,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.000966689006836337,
      "loss": 0.7882,
      "num_input_tokens_seen": 69900720,
      "step": 120395
    },
    {
      "epoch": 17.93267798629729,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0009660005401395483,
      "loss": 0.7941,
      "num_input_tokens_seen": 69903536,
      "step": 120400
    },
    {
      "epoch": 17.93342269883825,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0009653123105329169,
      "loss": 0.8056,
      "num_input_tokens_seen": 69906288,
      "step": 120405
    },
    {
      "epoch": 17.934167411379207,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0009646243180280705,
      "loss": 0.7834,
      "num_input_tokens_seen": 69909136,
      "step": 120410
    },
    {
      "epoch": 17.934912123920167,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.000963936562636628,
      "loss": 0.802,
      "num_input_tokens_seen": 69911888,
      "step": 120415
    },
    {
      "epoch": 17.935656836461128,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0009632490443702152,
      "loss": 0.8049,
      "num_input_tokens_seen": 69914864,
      "step": 120420
    },
    {
      "epoch": 17.936401549002085,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0009625617632404409,
      "loss": 0.7946,
      "num_input_tokens_seen": 69918064,
      "step": 120425
    },
    {
      "epoch": 17.937146261543045,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.000961874719258921,
      "loss": 0.7829,
      "num_input_tokens_seen": 69920752,
      "step": 120430
    },
    {
      "epoch": 17.937890974084002,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0009611879124372562,
      "loss": 0.8128,
      "num_input_tokens_seen": 69923760,
      "step": 120435
    },
    {
      "epoch": 17.938635686624963,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0009605013427870573,
      "loss": 0.7985,
      "num_input_tokens_seen": 69926640,
      "step": 120440
    },
    {
      "epoch": 17.939380399165923,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0009598150103199149,
      "loss": 0.8016,
      "num_input_tokens_seen": 69929488,
      "step": 120445
    },
    {
      "epoch": 17.94012511170688,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.000959128915047433,
      "loss": 0.7994,
      "num_input_tokens_seen": 69932528,
      "step": 120450
    },
    {
      "epoch": 17.94086982424784,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0009584430569811958,
      "loss": 0.8112,
      "num_input_tokens_seen": 69935184,
      "step": 120455
    },
    {
      "epoch": 17.941614536788798,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.0009577574361327889,
      "loss": 0.784,
      "num_input_tokens_seen": 69938160,
      "step": 120460
    },
    {
      "epoch": 17.94235924932976,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0009570720525138032,
      "loss": 0.8268,
      "num_input_tokens_seen": 69940848,
      "step": 120465
    },
    {
      "epoch": 17.94310396187072,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.000956386906135811,
      "loss": 0.7949,
      "num_input_tokens_seen": 69943696,
      "step": 120470
    },
    {
      "epoch": 17.943848674411676,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.000955701997010388,
      "loss": 0.7893,
      "num_input_tokens_seen": 69946544,
      "step": 120475
    },
    {
      "epoch": 17.944593386952636,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0009550173251491033,
      "loss": 0.779,
      "num_input_tokens_seen": 69949712,
      "step": 120480
    },
    {
      "epoch": 17.945338099493597,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0009543328905635295,
      "loss": 0.7892,
      "num_input_tokens_seen": 69952560,
      "step": 120485
    },
    {
      "epoch": 17.946082812034554,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0009536486932652271,
      "loss": 0.8193,
      "num_input_tokens_seen": 69955472,
      "step": 120490
    },
    {
      "epoch": 17.946827524575514,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0009529647332657504,
      "loss": 0.7934,
      "num_input_tokens_seen": 69958480,
      "step": 120495
    },
    {
      "epoch": 17.94757223711647,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0009522810105766599,
      "loss": 0.8128,
      "num_input_tokens_seen": 69961552,
      "step": 120500
    },
    {
      "epoch": 17.948316949657432,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0009515975252095033,
      "loss": 0.8041,
      "num_input_tokens_seen": 69964496,
      "step": 120505
    },
    {
      "epoch": 17.949061662198392,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0009509142771758293,
      "loss": 0.7851,
      "num_input_tokens_seen": 69967248,
      "step": 120510
    },
    {
      "epoch": 17.94980637473935,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0009502312664871775,
      "loss": 0.7849,
      "num_input_tokens_seen": 69970160,
      "step": 120515
    },
    {
      "epoch": 17.95055108728031,
      "grad_norm": 0.02734375,
      "learning_rate": 0.0009495484931550918,
      "loss": 0.7873,
      "num_input_tokens_seen": 69972976,
      "step": 120520
    },
    {
      "epoch": 17.95129579982127,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0009488659571911012,
      "loss": 0.8062,
      "num_input_tokens_seen": 69975920,
      "step": 120525
    },
    {
      "epoch": 17.952040512362228,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0009481836586067415,
      "loss": 0.7836,
      "num_input_tokens_seen": 69978832,
      "step": 120530
    },
    {
      "epoch": 17.952785224903188,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0009475015974135387,
      "loss": 0.7951,
      "num_input_tokens_seen": 69981872,
      "step": 120535
    },
    {
      "epoch": 17.953529937444145,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0009468197736230116,
      "loss": 0.8132,
      "num_input_tokens_seen": 69984656,
      "step": 120540
    },
    {
      "epoch": 17.954274649985106,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0009461381872466828,
      "loss": 0.778,
      "num_input_tokens_seen": 69987568,
      "step": 120545
    },
    {
      "epoch": 17.955019362526066,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0009454568382960631,
      "loss": 0.7959,
      "num_input_tokens_seen": 69990352,
      "step": 120550
    },
    {
      "epoch": 17.955764075067023,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0009447757267826684,
      "loss": 0.789,
      "num_input_tokens_seen": 69993136,
      "step": 120555
    },
    {
      "epoch": 17.956508787607984,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.000944094852718001,
      "loss": 0.7905,
      "num_input_tokens_seen": 69995984,
      "step": 120560
    },
    {
      "epoch": 17.957253500148944,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0009434142161135667,
      "loss": 0.8124,
      "num_input_tokens_seen": 69998736,
      "step": 120565
    },
    {
      "epoch": 17.9579982126899,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.000942733816980863,
      "loss": 0.8103,
      "num_input_tokens_seen": 70001680,
      "step": 120570
    },
    {
      "epoch": 17.95874292523086,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.0009420536553313824,
      "loss": 0.8114,
      "num_input_tokens_seen": 70004400,
      "step": 120575
    },
    {
      "epoch": 17.95948763777182,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.000941373731176619,
      "loss": 0.8046,
      "num_input_tokens_seen": 70007216,
      "step": 120580
    },
    {
      "epoch": 17.96023235031278,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0009406940445280553,
      "loss": 0.8171,
      "num_input_tokens_seen": 70010384,
      "step": 120585
    },
    {
      "epoch": 17.96097706285374,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0009400145953971789,
      "loss": 0.8085,
      "num_input_tokens_seen": 70013232,
      "step": 120590
    },
    {
      "epoch": 17.961721775394697,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0009393353837954671,
      "loss": 0.7862,
      "num_input_tokens_seen": 70016080,
      "step": 120595
    },
    {
      "epoch": 17.962466487935657,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0009386564097343908,
      "loss": 0.7819,
      "num_input_tokens_seen": 70018928,
      "step": 120600
    },
    {
      "epoch": 17.963211200476614,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0009379776732254241,
      "loss": 0.8021,
      "num_input_tokens_seen": 70022288,
      "step": 120605
    },
    {
      "epoch": 17.963955913017575,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0009372991742800346,
      "loss": 0.803,
      "num_input_tokens_seen": 70025040,
      "step": 120610
    },
    {
      "epoch": 17.964700625558535,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0009366209129096814,
      "loss": 0.7997,
      "num_input_tokens_seen": 70028208,
      "step": 120615
    },
    {
      "epoch": 17.965445338099492,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.000935942889125822,
      "loss": 0.7877,
      "num_input_tokens_seen": 70031088,
      "step": 120620
    },
    {
      "epoch": 17.966190050640453,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0009352651029399172,
      "loss": 0.7912,
      "num_input_tokens_seen": 70034096,
      "step": 120625
    },
    {
      "epoch": 17.966934763181413,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0009345875543634096,
      "loss": 0.8302,
      "num_input_tokens_seen": 70036976,
      "step": 120630
    },
    {
      "epoch": 17.96767947572237,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0009339102434077534,
      "loss": 0.7701,
      "num_input_tokens_seen": 70039856,
      "step": 120635
    },
    {
      "epoch": 17.96842418826333,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0009332331700843843,
      "loss": 0.7895,
      "num_input_tokens_seen": 70042736,
      "step": 120640
    },
    {
      "epoch": 17.969168900804288,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0009325563344047466,
      "loss": 0.8017,
      "num_input_tokens_seen": 70045936,
      "step": 120645
    },
    {
      "epoch": 17.96991361334525,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0009318797363802728,
      "loss": 0.7845,
      "num_input_tokens_seen": 70049296,
      "step": 120650
    },
    {
      "epoch": 17.97065832588621,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0009312033760223903,
      "loss": 0.7797,
      "num_input_tokens_seen": 70052304,
      "step": 120655
    },
    {
      "epoch": 17.971403038427166,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0009305272533425285,
      "loss": 0.7598,
      "num_input_tokens_seen": 70055216,
      "step": 120660
    },
    {
      "epoch": 17.972147750968126,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.0009298513683521114,
      "loss": 0.788,
      "num_input_tokens_seen": 70058384,
      "step": 120665
    },
    {
      "epoch": 17.972892463509087,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0009291757210625534,
      "loss": 0.7973,
      "num_input_tokens_seen": 70061200,
      "step": 120670
    },
    {
      "epoch": 17.973637176050044,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0009285003114852702,
      "loss": 0.8108,
      "num_input_tokens_seen": 70064080,
      "step": 120675
    },
    {
      "epoch": 17.974381888591004,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0009278251396316727,
      "loss": 0.8007,
      "num_input_tokens_seen": 70067088,
      "step": 120680
    },
    {
      "epoch": 17.97512660113196,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0009271502055131669,
      "loss": 0.8013,
      "num_input_tokens_seen": 70070384,
      "step": 120685
    },
    {
      "epoch": 17.975871313672922,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0009264755091411569,
      "loss": 0.7973,
      "num_input_tokens_seen": 70073072,
      "step": 120690
    },
    {
      "epoch": 17.976616026213883,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0009258010505270386,
      "loss": 0.8127,
      "num_input_tokens_seen": 70076368,
      "step": 120695
    },
    {
      "epoch": 17.97736073875484,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0009251268296822063,
      "loss": 0.7789,
      "num_input_tokens_seen": 70079216,
      "step": 120700
    },
    {
      "epoch": 17.9781054512958,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0009244528466180523,
      "loss": 0.7932,
      "num_input_tokens_seen": 70081840,
      "step": 120705
    },
    {
      "epoch": 17.97885016383676,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0009237791013459595,
      "loss": 0.7941,
      "num_input_tokens_seen": 70084944,
      "step": 120710
    },
    {
      "epoch": 17.979594876377718,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0009231055938773153,
      "loss": 0.8061,
      "num_input_tokens_seen": 70087792,
      "step": 120715
    },
    {
      "epoch": 17.980339588918678,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0009224323242234922,
      "loss": 0.7985,
      "num_input_tokens_seen": 70090544,
      "step": 120720
    },
    {
      "epoch": 17.981084301459635,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0009217592923958695,
      "loss": 0.8161,
      "num_input_tokens_seen": 70093488,
      "step": 120725
    },
    {
      "epoch": 17.981829014000596,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0009210864984058164,
      "loss": 0.7916,
      "num_input_tokens_seen": 70096272,
      "step": 120730
    },
    {
      "epoch": 17.982573726541556,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0009204139422646957,
      "loss": 0.7997,
      "num_input_tokens_seen": 70098992,
      "step": 120735
    },
    {
      "epoch": 17.983318439082513,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0009197416239838729,
      "loss": 0.8065,
      "num_input_tokens_seen": 70101936,
      "step": 120740
    },
    {
      "epoch": 17.984063151623474,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0009190695435747026,
      "loss": 0.7839,
      "num_input_tokens_seen": 70104976,
      "step": 120745
    },
    {
      "epoch": 17.984807864164434,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0009183977010485422,
      "loss": 0.7947,
      "num_input_tokens_seen": 70107824,
      "step": 120750
    },
    {
      "epoch": 17.98555257670539,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0009177260964167394,
      "loss": 0.7888,
      "num_input_tokens_seen": 70110736,
      "step": 120755
    },
    {
      "epoch": 17.98629728924635,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0009170547296906433,
      "loss": 0.8081,
      "num_input_tokens_seen": 70114000,
      "step": 120760
    },
    {
      "epoch": 17.98704200178731,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0009163836008815934,
      "loss": 0.7937,
      "num_input_tokens_seen": 70117136,
      "step": 120765
    },
    {
      "epoch": 17.98778671432827,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0009157127100009288,
      "loss": 0.7845,
      "num_input_tokens_seen": 70120080,
      "step": 120770
    },
    {
      "epoch": 17.98853142686923,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0009150420570599853,
      "loss": 0.8039,
      "num_input_tokens_seen": 70123088,
      "step": 120775
    },
    {
      "epoch": 17.989276139410187,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0009143716420700875,
      "loss": 0.7943,
      "num_input_tokens_seen": 70125936,
      "step": 120780
    },
    {
      "epoch": 17.990020851951147,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0009137014650425679,
      "loss": 0.8034,
      "num_input_tokens_seen": 70128656,
      "step": 120785
    },
    {
      "epoch": 17.990765564492108,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.0009130315259887423,
      "loss": 0.7857,
      "num_input_tokens_seen": 70131568,
      "step": 120790
    },
    {
      "epoch": 17.991510277033065,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0009123618249199333,
      "loss": 0.7936,
      "num_input_tokens_seen": 70134608,
      "step": 120795
    },
    {
      "epoch": 17.992254989574025,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0009116923618474554,
      "loss": 0.797,
      "num_input_tokens_seen": 70137648,
      "step": 120800
    },
    {
      "epoch": 17.992999702114982,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0009110231367826144,
      "loss": 0.7976,
      "num_input_tokens_seen": 70140624,
      "step": 120805
    },
    {
      "epoch": 17.993744414655943,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0009103541497367195,
      "loss": 0.7787,
      "num_input_tokens_seen": 70143248,
      "step": 120810
    },
    {
      "epoch": 17.994489127196903,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0009096854007210669,
      "loss": 0.8062,
      "num_input_tokens_seen": 70146384,
      "step": 120815
    },
    {
      "epoch": 17.99523383973786,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0009090168897469624,
      "loss": 0.812,
      "num_input_tokens_seen": 70149488,
      "step": 120820
    },
    {
      "epoch": 17.99597855227882,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0009083486168256937,
      "loss": 0.7907,
      "num_input_tokens_seen": 70152208,
      "step": 120825
    },
    {
      "epoch": 17.996723264819778,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0009076805819685568,
      "loss": 0.7877,
      "num_input_tokens_seen": 70155152,
      "step": 120830
    },
    {
      "epoch": 17.99746797736074,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0009070127851868292,
      "loss": 0.8062,
      "num_input_tokens_seen": 70158160,
      "step": 120835
    },
    {
      "epoch": 17.9982126899017,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0009063452264918004,
      "loss": 0.7905,
      "num_input_tokens_seen": 70161072,
      "step": 120840
    },
    {
      "epoch": 17.998957402442656,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.0009056779058947412,
      "loss": 0.8015,
      "num_input_tokens_seen": 70163856,
      "step": 120845
    },
    {
      "epoch": 17.999702114983616,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0009050108234069326,
      "loss": 0.81,
      "num_input_tokens_seen": 70166640,
      "step": 120850
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.7993520498275757,
      "eval_runtime": 70.495,
      "eval_samples_per_second": 42.329,
      "eval_steps_per_second": 10.582,
      "num_input_tokens_seen": 70167432,
      "step": 120852
    },
    {
      "epoch": 18.000446827524577,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0009043439790396406,
      "loss": 0.793,
      "num_input_tokens_seen": 70169320,
      "step": 120855
    },
    {
      "epoch": 18.001191540065534,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.000903677372804128,
      "loss": 0.8064,
      "num_input_tokens_seen": 70172264,
      "step": 120860
    },
    {
      "epoch": 18.001936252606495,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0009030110047116624,
      "loss": 0.8034,
      "num_input_tokens_seen": 70174984,
      "step": 120865
    },
    {
      "epoch": 18.00268096514745,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.000902344874773498,
      "loss": 0.787,
      "num_input_tokens_seen": 70177832,
      "step": 120870
    },
    {
      "epoch": 18.003425677688412,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0009016789830008892,
      "loss": 0.7993,
      "num_input_tokens_seen": 70180872,
      "step": 120875
    },
    {
      "epoch": 18.004170390229373,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.000901013329405082,
      "loss": 0.7974,
      "num_input_tokens_seen": 70183784,
      "step": 120880
    },
    {
      "epoch": 18.00491510277033,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0009003479139973275,
      "loss": 0.804,
      "num_input_tokens_seen": 70186536,
      "step": 120885
    },
    {
      "epoch": 18.00565981531129,
      "grad_norm": 0.060546875,
      "learning_rate": 0.0008996827367888649,
      "loss": 0.8347,
      "num_input_tokens_seen": 70190024,
      "step": 120890
    },
    {
      "epoch": 18.00640452785225,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0008990177977909286,
      "loss": 0.7969,
      "num_input_tokens_seen": 70192680,
      "step": 120895
    },
    {
      "epoch": 18.007149240393208,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0008983530970147579,
      "loss": 0.7943,
      "num_input_tokens_seen": 70195784,
      "step": 120900
    },
    {
      "epoch": 18.007893952934168,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0008976886344715773,
      "loss": 0.7889,
      "num_input_tokens_seen": 70198728,
      "step": 120905
    },
    {
      "epoch": 18.008638665475125,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.000897024410172616,
      "loss": 0.812,
      "num_input_tokens_seen": 70201544,
      "step": 120910
    },
    {
      "epoch": 18.009383378016086,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0008963604241290918,
      "loss": 0.8042,
      "num_input_tokens_seen": 70204232,
      "step": 120915
    },
    {
      "epoch": 18.010128090557046,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0008956966763522256,
      "loss": 0.779,
      "num_input_tokens_seen": 70206856,
      "step": 120920
    },
    {
      "epoch": 18.010872803098003,
      "grad_norm": 0.09375,
      "learning_rate": 0.0008950331668532302,
      "loss": 0.816,
      "num_input_tokens_seen": 70209512,
      "step": 120925
    },
    {
      "epoch": 18.011617515638964,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0008943698956433098,
      "loss": 0.797,
      "num_input_tokens_seen": 70212488,
      "step": 120930
    },
    {
      "epoch": 18.012362228179924,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0008937068627336774,
      "loss": 0.7913,
      "num_input_tokens_seen": 70216008,
      "step": 120935
    },
    {
      "epoch": 18.01310694072088,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0008930440681355306,
      "loss": 0.7797,
      "num_input_tokens_seen": 70218920,
      "step": 120940
    },
    {
      "epoch": 18.013851653261842,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0008923815118600653,
      "loss": 0.7846,
      "num_input_tokens_seen": 70221704,
      "step": 120945
    },
    {
      "epoch": 18.0145963658028,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0008917191939184759,
      "loss": 0.798,
      "num_input_tokens_seen": 70224552,
      "step": 120950
    },
    {
      "epoch": 18.01534107834376,
      "grad_norm": 0.0390625,
      "learning_rate": 0.000891057114321952,
      "loss": 0.7828,
      "num_input_tokens_seen": 70227240,
      "step": 120955
    },
    {
      "epoch": 18.01608579088472,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0008903952730816761,
      "loss": 0.8104,
      "num_input_tokens_seen": 70230088,
      "step": 120960
    },
    {
      "epoch": 18.016830503425677,
      "grad_norm": 0.041015625,
      "learning_rate": 0.000889733670208836,
      "loss": 0.8117,
      "num_input_tokens_seen": 70232904,
      "step": 120965
    },
    {
      "epoch": 18.017575215966637,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0008890723057146027,
      "loss": 0.7906,
      "num_input_tokens_seen": 70235432,
      "step": 120970
    },
    {
      "epoch": 18.018319928507594,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.000888411179610149,
      "loss": 0.7914,
      "num_input_tokens_seen": 70238120,
      "step": 120975
    },
    {
      "epoch": 18.019064641048555,
      "grad_norm": 0.109375,
      "learning_rate": 0.0008877502919066493,
      "loss": 0.7946,
      "num_input_tokens_seen": 70241032,
      "step": 120980
    },
    {
      "epoch": 18.019809353589515,
      "grad_norm": 0.05078125,
      "learning_rate": 0.000887089642615263,
      "loss": 0.7992,
      "num_input_tokens_seen": 70243912,
      "step": 120985
    },
    {
      "epoch": 18.020554066130472,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0008864292317471544,
      "loss": 0.8061,
      "num_input_tokens_seen": 70246792,
      "step": 120990
    },
    {
      "epoch": 18.021298778671433,
      "grad_norm": 0.375,
      "learning_rate": 0.0008857690593134798,
      "loss": 0.8256,
      "num_input_tokens_seen": 70249576,
      "step": 120995
    },
    {
      "epoch": 18.022043491212393,
      "grad_norm": 0.0274658203125,
      "learning_rate": 0.0008851091253253934,
      "loss": 0.8124,
      "num_input_tokens_seen": 70252520,
      "step": 121000
    },
    {
      "epoch": 18.02278820375335,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0008844494297940397,
      "loss": 0.7867,
      "num_input_tokens_seen": 70255336,
      "step": 121005
    },
    {
      "epoch": 18.02353291629431,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0008837899727305681,
      "loss": 0.8084,
      "num_input_tokens_seen": 70258440,
      "step": 121010
    },
    {
      "epoch": 18.024277628835268,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.000883130754146118,
      "loss": 0.8052,
      "num_input_tokens_seen": 70261128,
      "step": 121015
    },
    {
      "epoch": 18.02502234137623,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.000882471774051824,
      "loss": 0.8001,
      "num_input_tokens_seen": 70263976,
      "step": 121020
    },
    {
      "epoch": 18.02576705391719,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0008818130324588236,
      "loss": 0.8193,
      "num_input_tokens_seen": 70266920,
      "step": 121025
    },
    {
      "epoch": 18.026511766458146,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0008811545293782397,
      "loss": 0.7904,
      "num_input_tokens_seen": 70269832,
      "step": 121030
    },
    {
      "epoch": 18.027256478999107,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0008804962648212033,
      "loss": 0.8124,
      "num_input_tokens_seen": 70272808,
      "step": 121035
    },
    {
      "epoch": 18.028001191540067,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.000879838238798829,
      "loss": 0.7845,
      "num_input_tokens_seen": 70275560,
      "step": 121040
    },
    {
      "epoch": 18.028745904081024,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0008791804513222395,
      "loss": 0.8067,
      "num_input_tokens_seen": 70278312,
      "step": 121045
    },
    {
      "epoch": 18.029490616621985,
      "grad_norm": 0.046875,
      "learning_rate": 0.0008785229024025426,
      "loss": 0.8091,
      "num_input_tokens_seen": 70281480,
      "step": 121050
    },
    {
      "epoch": 18.03023532916294,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.000877865592050851,
      "loss": 0.81,
      "num_input_tokens_seen": 70284392,
      "step": 121055
    },
    {
      "epoch": 18.030980041703902,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0008772085202782659,
      "loss": 0.803,
      "num_input_tokens_seen": 70287336,
      "step": 121060
    },
    {
      "epoch": 18.031724754244863,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.00087655168709589,
      "loss": 0.7699,
      "num_input_tokens_seen": 70290408,
      "step": 121065
    },
    {
      "epoch": 18.03246946678582,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0008758950925148195,
      "loss": 0.7913,
      "num_input_tokens_seen": 70293192,
      "step": 121070
    },
    {
      "epoch": 18.03321417932678,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0008752387365461439,
      "loss": 0.8008,
      "num_input_tokens_seen": 70295880,
      "step": 121075
    },
    {
      "epoch": 18.03395889186774,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0008745826192009559,
      "loss": 0.8062,
      "num_input_tokens_seen": 70298632,
      "step": 121080
    },
    {
      "epoch": 18.034703604408698,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.000873926740490335,
      "loss": 0.7908,
      "num_input_tokens_seen": 70301608,
      "step": 121085
    },
    {
      "epoch": 18.035448316949658,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0008732711004253674,
      "loss": 0.8268,
      "num_input_tokens_seen": 70304616,
      "step": 121090
    },
    {
      "epoch": 18.036193029490615,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.0008726156990171274,
      "loss": 0.7973,
      "num_input_tokens_seen": 70307496,
      "step": 121095
    },
    {
      "epoch": 18.036937742031576,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0008719605362766847,
      "loss": 0.8052,
      "num_input_tokens_seen": 70310760,
      "step": 121100
    },
    {
      "epoch": 18.037682454572536,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0008713056122151119,
      "loss": 0.8253,
      "num_input_tokens_seen": 70313704,
      "step": 121105
    },
    {
      "epoch": 18.038427167113493,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0008706509268434687,
      "loss": 0.8025,
      "num_input_tokens_seen": 70316840,
      "step": 121110
    },
    {
      "epoch": 18.039171879654454,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0008699964801728193,
      "loss": 0.8089,
      "num_input_tokens_seen": 70319624,
      "step": 121115
    },
    {
      "epoch": 18.039916592195414,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0008693422722142152,
      "loss": 0.8005,
      "num_input_tokens_seen": 70322408,
      "step": 121120
    },
    {
      "epoch": 18.04066130473637,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0008686883029787156,
      "loss": 0.7878,
      "num_input_tokens_seen": 70325224,
      "step": 121125
    },
    {
      "epoch": 18.041406017277332,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0008680345724773635,
      "loss": 0.8071,
      "num_input_tokens_seen": 70327944,
      "step": 121130
    },
    {
      "epoch": 18.04215072981829,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0008673810807212051,
      "loss": 0.8066,
      "num_input_tokens_seen": 70330504,
      "step": 121135
    },
    {
      "epoch": 18.04289544235925,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0008667278277212797,
      "loss": 0.7741,
      "num_input_tokens_seen": 70334696,
      "step": 121140
    },
    {
      "epoch": 18.04364015490021,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0008660748134886203,
      "loss": 0.7914,
      "num_input_tokens_seen": 70337448,
      "step": 121145
    },
    {
      "epoch": 18.044384867441167,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0008654220380342648,
      "loss": 0.8066,
      "num_input_tokens_seen": 70340360,
      "step": 121150
    },
    {
      "epoch": 18.045129579982127,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0008647695013692358,
      "loss": 0.8244,
      "num_input_tokens_seen": 70343592,
      "step": 121155
    },
    {
      "epoch": 18.045874292523084,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0008641172035045613,
      "loss": 0.7996,
      "num_input_tokens_seen": 70346504,
      "step": 121160
    },
    {
      "epoch": 18.046619005064045,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0008634651444512575,
      "loss": 0.7821,
      "num_input_tokens_seen": 70349544,
      "step": 121165
    },
    {
      "epoch": 18.047363717605005,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0008628133242203456,
      "loss": 0.8074,
      "num_input_tokens_seen": 70352360,
      "step": 121170
    },
    {
      "epoch": 18.048108430145962,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0008621617428228334,
      "loss": 0.7857,
      "num_input_tokens_seen": 70355528,
      "step": 121175
    },
    {
      "epoch": 18.048853142686923,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0008615104002697271,
      "loss": 0.8026,
      "num_input_tokens_seen": 70358280,
      "step": 121180
    },
    {
      "epoch": 18.049597855227884,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0008608592965720346,
      "loss": 0.7902,
      "num_input_tokens_seen": 70361288,
      "step": 121185
    },
    {
      "epoch": 18.05034256776884,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0008602084317407521,
      "loss": 0.7845,
      "num_input_tokens_seen": 70364296,
      "step": 121190
    },
    {
      "epoch": 18.0510872803098,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0008595578057868807,
      "loss": 0.8034,
      "num_input_tokens_seen": 70367240,
      "step": 121195
    },
    {
      "epoch": 18.051831992850758,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0008589074187214068,
      "loss": 0.7922,
      "num_input_tokens_seen": 70370056,
      "step": 121200
    },
    {
      "epoch": 18.05257670539172,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0008582572705553199,
      "loss": 0.8026,
      "num_input_tokens_seen": 70373064,
      "step": 121205
    },
    {
      "epoch": 18.05332141793268,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.0008576073612996009,
      "loss": 0.7928,
      "num_input_tokens_seen": 70375816,
      "step": 121210
    },
    {
      "epoch": 18.054066130473636,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0008569576909652349,
      "loss": 0.7954,
      "num_input_tokens_seen": 70378920,
      "step": 121215
    },
    {
      "epoch": 18.054810843014597,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0008563082595631944,
      "loss": 0.7943,
      "num_input_tokens_seen": 70382344,
      "step": 121220
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0008556590671044489,
      "loss": 0.7989,
      "num_input_tokens_seen": 70385288,
      "step": 121225
    },
    {
      "epoch": 18.056300268096514,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0008550101135999699,
      "loss": 0.7854,
      "num_input_tokens_seen": 70388008,
      "step": 121230
    },
    {
      "epoch": 18.057044980637475,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0008543613990607151,
      "loss": 0.8031,
      "num_input_tokens_seen": 70390952,
      "step": 121235
    },
    {
      "epoch": 18.05778969317843,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0008537129234976509,
      "loss": 0.8091,
      "num_input_tokens_seen": 70393704,
      "step": 121240
    },
    {
      "epoch": 18.058534405719392,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0008530646869217267,
      "loss": 0.7936,
      "num_input_tokens_seen": 70396808,
      "step": 121245
    },
    {
      "epoch": 18.059279118260353,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0008524166893438988,
      "loss": 0.805,
      "num_input_tokens_seen": 70399560,
      "step": 121250
    },
    {
      "epoch": 18.06002383080131,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.000851768930775112,
      "loss": 0.7917,
      "num_input_tokens_seen": 70402568,
      "step": 121255
    },
    {
      "epoch": 18.06076854334227,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0008511214112263071,
      "loss": 0.7921,
      "num_input_tokens_seen": 70405512,
      "step": 121260
    },
    {
      "epoch": 18.06151325588323,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0008504741307084273,
      "loss": 0.7908,
      "num_input_tokens_seen": 70408264,
      "step": 121265
    },
    {
      "epoch": 18.062257968424188,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.0008498270892324072,
      "loss": 0.795,
      "num_input_tokens_seen": 70411112,
      "step": 121270
    },
    {
      "epoch": 18.06300268096515,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0008491802868091763,
      "loss": 0.8104,
      "num_input_tokens_seen": 70413768,
      "step": 121275
    },
    {
      "epoch": 18.063747393506105,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0008485337234496593,
      "loss": 0.7864,
      "num_input_tokens_seen": 70416744,
      "step": 121280
    },
    {
      "epoch": 18.064492106047066,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0008478873991647856,
      "loss": 0.8099,
      "num_input_tokens_seen": 70419528,
      "step": 121285
    },
    {
      "epoch": 18.065236818588026,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0008472413139654666,
      "loss": 0.7845,
      "num_input_tokens_seen": 70422536,
      "step": 121290
    },
    {
      "epoch": 18.065981531128983,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0008465954678626253,
      "loss": 0.7974,
      "num_input_tokens_seen": 70425704,
      "step": 121295
    },
    {
      "epoch": 18.066726243669944,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0008459498608671678,
      "loss": 0.8071,
      "num_input_tokens_seen": 70428872,
      "step": 121300
    },
    {
      "epoch": 18.0674709562109,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0008453044929900006,
      "loss": 0.7954,
      "num_input_tokens_seen": 70431560,
      "step": 121305
    },
    {
      "epoch": 18.06821566875186,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0008446593642420296,
      "loss": 0.7928,
      "num_input_tokens_seen": 70434408,
      "step": 121310
    },
    {
      "epoch": 18.068960381292822,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0008440144746341499,
      "loss": 0.811,
      "num_input_tokens_seen": 70437320,
      "step": 121315
    },
    {
      "epoch": 18.06970509383378,
      "grad_norm": 0.031005859375,
      "learning_rate": 0.0008433698241772608,
      "loss": 0.7977,
      "num_input_tokens_seen": 70440296,
      "step": 121320
    },
    {
      "epoch": 18.07044980637474,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0008427254128822487,
      "loss": 0.7763,
      "num_input_tokens_seen": 70443240,
      "step": 121325
    },
    {
      "epoch": 18.0711945189157,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0008420812407600031,
      "loss": 0.8031,
      "num_input_tokens_seen": 70445960,
      "step": 121330
    },
    {
      "epoch": 18.071939231456657,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0008414373078214054,
      "loss": 0.8185,
      "num_input_tokens_seen": 70449320,
      "step": 121335
    },
    {
      "epoch": 18.072683943997617,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.000840793614077332,
      "loss": 0.8172,
      "num_input_tokens_seen": 70452168,
      "step": 121340
    },
    {
      "epoch": 18.073428656538574,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0008401501595386606,
      "loss": 0.7912,
      "num_input_tokens_seen": 70455080,
      "step": 121345
    },
    {
      "epoch": 18.074173369079535,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0008395069442162611,
      "loss": 0.8081,
      "num_input_tokens_seen": 70457896,
      "step": 121350
    },
    {
      "epoch": 18.074918081620496,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0008388639681209997,
      "loss": 0.8023,
      "num_input_tokens_seen": 70460904,
      "step": 121355
    },
    {
      "epoch": 18.075662794161452,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0008382212312637376,
      "loss": 0.8093,
      "num_input_tokens_seen": 70463816,
      "step": 121360
    },
    {
      "epoch": 18.076407506702413,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0008375787336553364,
      "loss": 0.7771,
      "num_input_tokens_seen": 70466856,
      "step": 121365
    },
    {
      "epoch": 18.077152219243374,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0008369364753066472,
      "loss": 0.7941,
      "num_input_tokens_seen": 70469800,
      "step": 121370
    },
    {
      "epoch": 18.07789693178433,
      "grad_norm": 0.048828125,
      "learning_rate": 0.000836294456228523,
      "loss": 0.7811,
      "num_input_tokens_seen": 70472840,
      "step": 121375
    },
    {
      "epoch": 18.07864164432529,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0008356526764318083,
      "loss": 0.8004,
      "num_input_tokens_seen": 70475368,
      "step": 121380
    },
    {
      "epoch": 18.079386356866248,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0008350111359273431,
      "loss": 0.8058,
      "num_input_tokens_seen": 70478440,
      "step": 121385
    },
    {
      "epoch": 18.08013106940721,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.0008343698347259704,
      "loss": 0.8014,
      "num_input_tokens_seen": 70481288,
      "step": 121390
    },
    {
      "epoch": 18.08087578194817,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0008337287728385228,
      "loss": 0.7735,
      "num_input_tokens_seen": 70484168,
      "step": 121395
    },
    {
      "epoch": 18.081620494489126,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0008330879502758304,
      "loss": 0.7897,
      "num_input_tokens_seen": 70487112,
      "step": 121400
    },
    {
      "epoch": 18.082365207030087,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0008324473670487142,
      "loss": 0.802,
      "num_input_tokens_seen": 70489832,
      "step": 121405
    },
    {
      "epoch": 18.083109919571047,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0008318070231680041,
      "loss": 0.795,
      "num_input_tokens_seen": 70492840,
      "step": 121410
    },
    {
      "epoch": 18.083854632112004,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0008311669186445147,
      "loss": 0.7945,
      "num_input_tokens_seen": 70495688,
      "step": 121415
    },
    {
      "epoch": 18.084599344652965,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0008305270534890557,
      "loss": 0.7876,
      "num_input_tokens_seen": 70498568,
      "step": 121420
    },
    {
      "epoch": 18.08534405719392,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0008298874277124452,
      "loss": 0.7904,
      "num_input_tokens_seen": 70501384,
      "step": 121425
    },
    {
      "epoch": 18.086088769734882,
      "grad_norm": 0.048828125,
      "learning_rate": 0.000829248041325481,
      "loss": 0.8036,
      "num_input_tokens_seen": 70504072,
      "step": 121430
    },
    {
      "epoch": 18.086833482275843,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0008286088943389713,
      "loss": 0.8052,
      "num_input_tokens_seen": 70507048,
      "step": 121435
    },
    {
      "epoch": 18.0875781948168,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0008279699867637091,
      "loss": 0.781,
      "num_input_tokens_seen": 70509896,
      "step": 121440
    },
    {
      "epoch": 18.08832290735776,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0008273313186104924,
      "loss": 0.8076,
      "num_input_tokens_seen": 70513128,
      "step": 121445
    },
    {
      "epoch": 18.08906761989872,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.000826692889890106,
      "loss": 0.781,
      "num_input_tokens_seen": 70516264,
      "step": 121450
    },
    {
      "epoch": 18.089812332439678,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.0008260547006133411,
      "loss": 0.7835,
      "num_input_tokens_seen": 70519240,
      "step": 121455
    },
    {
      "epoch": 18.09055704498064,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.000825416750790976,
      "loss": 0.7798,
      "num_input_tokens_seen": 70522344,
      "step": 121460
    },
    {
      "epoch": 18.091301757521595,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0008247790404337885,
      "loss": 0.804,
      "num_input_tokens_seen": 70525000,
      "step": 121465
    },
    {
      "epoch": 18.092046470062556,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0008241415695525516,
      "loss": 0.7851,
      "num_input_tokens_seen": 70527976,
      "step": 121470
    },
    {
      "epoch": 18.092791182603516,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0008235043381580337,
      "loss": 0.8025,
      "num_input_tokens_seen": 70530792,
      "step": 121475
    },
    {
      "epoch": 18.093535895144473,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0008228673462610025,
      "loss": 0.7814,
      "num_input_tokens_seen": 70533736,
      "step": 121480
    },
    {
      "epoch": 18.094280607685434,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.000822230593872218,
      "loss": 0.7854,
      "num_input_tokens_seen": 70536520,
      "step": 121485
    },
    {
      "epoch": 18.09502532022639,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.0008215940810024397,
      "loss": 0.79,
      "num_input_tokens_seen": 70539368,
      "step": 121490
    },
    {
      "epoch": 18.09577003276735,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0008209578076624174,
      "loss": 0.8093,
      "num_input_tokens_seen": 70542344,
      "step": 121495
    },
    {
      "epoch": 18.096514745308312,
      "grad_norm": 0.373046875,
      "learning_rate": 0.0008203217738629026,
      "loss": 0.8164,
      "num_input_tokens_seen": 70545384,
      "step": 121500
    },
    {
      "epoch": 18.09725945784927,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.00081968597961464,
      "loss": 0.8016,
      "num_input_tokens_seen": 70548616,
      "step": 121505
    },
    {
      "epoch": 18.09800417039023,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0008190504249283692,
      "loss": 0.8021,
      "num_input_tokens_seen": 70552040,
      "step": 121510
    },
    {
      "epoch": 18.09874888293119,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.00081841510981483,
      "loss": 0.7985,
      "num_input_tokens_seen": 70555240,
      "step": 121515
    },
    {
      "epoch": 18.099493595472147,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0008177800342847524,
      "loss": 0.8036,
      "num_input_tokens_seen": 70557768,
      "step": 121520
    },
    {
      "epoch": 18.100238308013108,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0008171451983488691,
      "loss": 0.8058,
      "num_input_tokens_seen": 70560744,
      "step": 121525
    },
    {
      "epoch": 18.100983020554064,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0008165106020179035,
      "loss": 0.8213,
      "num_input_tokens_seen": 70563720,
      "step": 121530
    },
    {
      "epoch": 18.101727733095025,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0008158762453025752,
      "loss": 0.8032,
      "num_input_tokens_seen": 70566664,
      "step": 121535
    },
    {
      "epoch": 18.102472445635986,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0008152421282136024,
      "loss": 0.7817,
      "num_input_tokens_seen": 70569416,
      "step": 121540
    },
    {
      "epoch": 18.103217158176943,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0008146082507616931,
      "loss": 0.804,
      "num_input_tokens_seen": 70572296,
      "step": 121545
    },
    {
      "epoch": 18.103961870717903,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.0008139746129575637,
      "loss": 0.7706,
      "num_input_tokens_seen": 70575080,
      "step": 121550
    },
    {
      "epoch": 18.104706583258864,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0008133412148119123,
      "loss": 0.8163,
      "num_input_tokens_seen": 70577832,
      "step": 121555
    },
    {
      "epoch": 18.10545129579982,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0008127080563354455,
      "loss": 0.7882,
      "num_input_tokens_seen": 70580680,
      "step": 121560
    },
    {
      "epoch": 18.10619600834078,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0008120751375388546,
      "loss": 0.7895,
      "num_input_tokens_seen": 70583432,
      "step": 121565
    },
    {
      "epoch": 18.106940720881738,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0008114424584328344,
      "loss": 0.82,
      "num_input_tokens_seen": 70586408,
      "step": 121570
    },
    {
      "epoch": 18.1076854334227,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.0008108100190280748,
      "loss": 0.8062,
      "num_input_tokens_seen": 70589032,
      "step": 121575
    },
    {
      "epoch": 18.10843014596366,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.0008101778193352571,
      "loss": 0.7962,
      "num_input_tokens_seen": 70591528,
      "step": 121580
    },
    {
      "epoch": 18.109174858504616,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.0008095458593650645,
      "loss": 0.7973,
      "num_input_tokens_seen": 70594280,
      "step": 121585
    },
    {
      "epoch": 18.109919571045577,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0008089141391281701,
      "loss": 0.8011,
      "num_input_tokens_seen": 70597672,
      "step": 121590
    },
    {
      "epoch": 18.110664283586537,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0008082826586352504,
      "loss": 0.8044,
      "num_input_tokens_seen": 70600616,
      "step": 121595
    },
    {
      "epoch": 18.111408996127494,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0008076514178969717,
      "loss": 0.8023,
      "num_input_tokens_seen": 70603720,
      "step": 121600
    },
    {
      "epoch": 18.112153708668455,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0008070204169239975,
      "loss": 0.8031,
      "num_input_tokens_seen": 70606472,
      "step": 121605
    },
    {
      "epoch": 18.11289842120941,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0008063896557269856,
      "loss": 0.7921,
      "num_input_tokens_seen": 70609448,
      "step": 121610
    },
    {
      "epoch": 18.113643133750372,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0008057591343165976,
      "loss": 0.8021,
      "num_input_tokens_seen": 70612232,
      "step": 121615
    },
    {
      "epoch": 18.114387846291333,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0008051288527034816,
      "loss": 0.8178,
      "num_input_tokens_seen": 70615336,
      "step": 121620
    },
    {
      "epoch": 18.11513255883229,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0008044988108982859,
      "loss": 0.7994,
      "num_input_tokens_seen": 70618184,
      "step": 121625
    },
    {
      "epoch": 18.11587727137325,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0008038690089116568,
      "loss": 0.798,
      "num_input_tokens_seen": 70621064,
      "step": 121630
    },
    {
      "epoch": 18.11662198391421,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0008032394467542292,
      "loss": 0.7935,
      "num_input_tokens_seen": 70623944,
      "step": 121635
    },
    {
      "epoch": 18.117366696455168,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0008026101244366463,
      "loss": 0.7979,
      "num_input_tokens_seen": 70626696,
      "step": 121640
    },
    {
      "epoch": 18.11811140899613,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0008019810419695328,
      "loss": 0.8065,
      "num_input_tokens_seen": 70629672,
      "step": 121645
    },
    {
      "epoch": 18.118856121537085,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0008013521993635202,
      "loss": 0.8321,
      "num_input_tokens_seen": 70632456,
      "step": 121650
    },
    {
      "epoch": 18.119600834078046,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.0008007235966292336,
      "loss": 0.7904,
      "num_input_tokens_seen": 70635240,
      "step": 121655
    },
    {
      "epoch": 18.120345546619006,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0008000952337772893,
      "loss": 0.8115,
      "num_input_tokens_seen": 70638248,
      "step": 121660
    },
    {
      "epoch": 18.121090259159963,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0007994671108183037,
      "loss": 0.7829,
      "num_input_tokens_seen": 70641320,
      "step": 121665
    },
    {
      "epoch": 18.121834971700924,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0007988392277628852,
      "loss": 0.7757,
      "num_input_tokens_seen": 70644168,
      "step": 121670
    },
    {
      "epoch": 18.12257968424188,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0007982115846216486,
      "loss": 0.8022,
      "num_input_tokens_seen": 70646984,
      "step": 121675
    },
    {
      "epoch": 18.12332439678284,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0007975841814051904,
      "loss": 0.8082,
      "num_input_tokens_seen": 70649832,
      "step": 121680
    },
    {
      "epoch": 18.124069109323802,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0007969570181241153,
      "loss": 0.7657,
      "num_input_tokens_seen": 70652584,
      "step": 121685
    },
    {
      "epoch": 18.12481382186476,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0007963300947890134,
      "loss": 0.7917,
      "num_input_tokens_seen": 70655176,
      "step": 121690
    },
    {
      "epoch": 18.12555853440572,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.000795703411410481,
      "loss": 0.7999,
      "num_input_tokens_seen": 70657800,
      "step": 121695
    },
    {
      "epoch": 18.12630324694668,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0007950769679991015,
      "loss": 0.7984,
      "num_input_tokens_seen": 70660712,
      "step": 121700
    },
    {
      "epoch": 18.127047959487637,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0007944507645654597,
      "loss": 0.7926,
      "num_input_tokens_seen": 70663528,
      "step": 121705
    },
    {
      "epoch": 18.127792672028598,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0007938248011201354,
      "loss": 0.7912,
      "num_input_tokens_seen": 70666376,
      "step": 121710
    },
    {
      "epoch": 18.128537384569555,
      "grad_norm": 0.0244140625,
      "learning_rate": 0.0007931990776737002,
      "loss": 0.7897,
      "num_input_tokens_seen": 70669288,
      "step": 121715
    },
    {
      "epoch": 18.129282097110515,
      "grad_norm": 0.0625,
      "learning_rate": 0.000792573594236729,
      "loss": 0.8104,
      "num_input_tokens_seen": 70672232,
      "step": 121720
    },
    {
      "epoch": 18.130026809651476,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0007919483508197883,
      "loss": 0.7839,
      "num_input_tokens_seen": 70675304,
      "step": 121725
    },
    {
      "epoch": 18.130771522192433,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0007913233474334397,
      "loss": 0.7879,
      "num_input_tokens_seen": 70677992,
      "step": 121730
    },
    {
      "epoch": 18.131516234733393,
      "grad_norm": 0.06640625,
      "learning_rate": 0.000790698584088238,
      "loss": 0.7816,
      "num_input_tokens_seen": 70681096,
      "step": 121735
    },
    {
      "epoch": 18.132260947274354,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0007900740607947465,
      "loss": 0.7971,
      "num_input_tokens_seen": 70683816,
      "step": 121740
    },
    {
      "epoch": 18.13300565981531,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.00078944977756351,
      "loss": 0.7873,
      "num_input_tokens_seen": 70686664,
      "step": 121745
    },
    {
      "epoch": 18.13375037235627,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0007888257344050753,
      "loss": 0.7814,
      "num_input_tokens_seen": 70689480,
      "step": 121750
    },
    {
      "epoch": 18.134495084897228,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0007882019313299887,
      "loss": 0.8025,
      "num_input_tokens_seen": 70692360,
      "step": 121755
    },
    {
      "epoch": 18.13523979743819,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0007875783683487836,
      "loss": 0.7756,
      "num_input_tokens_seen": 70695080,
      "step": 121760
    },
    {
      "epoch": 18.13598450997915,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0007869550454719981,
      "loss": 0.8419,
      "num_input_tokens_seen": 70697960,
      "step": 121765
    },
    {
      "epoch": 18.136729222520106,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0007863319627101605,
      "loss": 0.7745,
      "num_input_tokens_seen": 70700808,
      "step": 121770
    },
    {
      "epoch": 18.137473935061067,
      "grad_norm": 0.046875,
      "learning_rate": 0.0007857091200738008,
      "loss": 0.7905,
      "num_input_tokens_seen": 70703560,
      "step": 121775
    },
    {
      "epoch": 18.138218647602027,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0007850865175734389,
      "loss": 0.8249,
      "num_input_tokens_seen": 70706376,
      "step": 121780
    },
    {
      "epoch": 18.138963360142984,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0007844641552195897,
      "loss": 0.7884,
      "num_input_tokens_seen": 70709160,
      "step": 121785
    },
    {
      "epoch": 18.139708072683945,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.000783842033022773,
      "loss": 0.773,
      "num_input_tokens_seen": 70712168,
      "step": 121790
    },
    {
      "epoch": 18.140452785224902,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0007832201509934972,
      "loss": 0.7819,
      "num_input_tokens_seen": 70715048,
      "step": 121795
    },
    {
      "epoch": 18.141197497765862,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0007825985091422671,
      "loss": 0.7876,
      "num_input_tokens_seen": 70717960,
      "step": 121800
    },
    {
      "epoch": 18.141942210306823,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0007819771074795827,
      "loss": 0.7985,
      "num_input_tokens_seen": 70720776,
      "step": 121805
    },
    {
      "epoch": 18.14268692284778,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0007813559460159475,
      "loss": 0.7744,
      "num_input_tokens_seen": 70723816,
      "step": 121810
    },
    {
      "epoch": 18.14343163538874,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0007807350247618511,
      "loss": 0.7833,
      "num_input_tokens_seen": 70726760,
      "step": 121815
    },
    {
      "epoch": 18.1441763479297,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0007801143437277852,
      "loss": 0.7961,
      "num_input_tokens_seen": 70729992,
      "step": 121820
    },
    {
      "epoch": 18.144921060470658,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0007794939029242348,
      "loss": 0.8,
      "num_input_tokens_seen": 70732744,
      "step": 121825
    },
    {
      "epoch": 18.14566577301162,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0007788737023616815,
      "loss": 0.804,
      "num_input_tokens_seen": 70735784,
      "step": 121830
    },
    {
      "epoch": 18.146410485552575,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0007782537420506036,
      "loss": 0.7976,
      "num_input_tokens_seen": 70738792,
      "step": 121835
    },
    {
      "epoch": 18.147155198093536,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0007776340220014726,
      "loss": 0.7933,
      "num_input_tokens_seen": 70741800,
      "step": 121840
    },
    {
      "epoch": 18.147899910634496,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0007770145422247621,
      "loss": 0.7968,
      "num_input_tokens_seen": 70744776,
      "step": 121845
    },
    {
      "epoch": 18.148644623175453,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0007763953027309333,
      "loss": 0.7905,
      "num_input_tokens_seen": 70748072,
      "step": 121850
    },
    {
      "epoch": 18.149389335716414,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0007757763035304532,
      "loss": 0.8007,
      "num_input_tokens_seen": 70750984,
      "step": 121855
    },
    {
      "epoch": 18.15013404825737,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0007751575446337733,
      "loss": 0.8126,
      "num_input_tokens_seen": 70753896,
      "step": 121860
    },
    {
      "epoch": 18.15087876079833,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0007745390260513502,
      "loss": 0.7873,
      "num_input_tokens_seen": 70756968,
      "step": 121865
    },
    {
      "epoch": 18.151623473339292,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0007739207477936321,
      "loss": 0.8033,
      "num_input_tokens_seen": 70759912,
      "step": 121870
    },
    {
      "epoch": 18.15236818588025,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0007733027098710627,
      "loss": 0.7854,
      "num_input_tokens_seen": 70763016,
      "step": 121875
    },
    {
      "epoch": 18.15311289842121,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0007726849122940848,
      "loss": 0.7975,
      "num_input_tokens_seen": 70766184,
      "step": 121880
    },
    {
      "epoch": 18.15385761096217,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0007720673550731355,
      "loss": 0.7855,
      "num_input_tokens_seen": 70768968,
      "step": 121885
    },
    {
      "epoch": 18.154602323503127,
      "grad_norm": 0.06640625,
      "learning_rate": 0.0007714500382186479,
      "loss": 0.8075,
      "num_input_tokens_seen": 70771784,
      "step": 121890
    },
    {
      "epoch": 18.155347036044088,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0007708329617410502,
      "loss": 0.7961,
      "num_input_tokens_seen": 70774440,
      "step": 121895
    },
    {
      "epoch": 18.156091748585045,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0007702161256507694,
      "loss": 0.8188,
      "num_input_tokens_seen": 70777352,
      "step": 121900
    },
    {
      "epoch": 18.156836461126005,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0007695995299582236,
      "loss": 0.8126,
      "num_input_tokens_seen": 70780168,
      "step": 121905
    },
    {
      "epoch": 18.157581173666966,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0007689831746738295,
      "loss": 0.7963,
      "num_input_tokens_seen": 70782792,
      "step": 121910
    },
    {
      "epoch": 18.158325886207923,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0007683670598080022,
      "loss": 0.8005,
      "num_input_tokens_seen": 70785704,
      "step": 121915
    },
    {
      "epoch": 18.159070598748883,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0007677511853711482,
      "loss": 0.7988,
      "num_input_tokens_seen": 70788552,
      "step": 121920
    },
    {
      "epoch": 18.159815311289844,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0007671355513736743,
      "loss": 0.7906,
      "num_input_tokens_seen": 70791048,
      "step": 121925
    },
    {
      "epoch": 18.1605600238308,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0007665201578259789,
      "loss": 0.8025,
      "num_input_tokens_seen": 70793896,
      "step": 121930
    },
    {
      "epoch": 18.16130473637176,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0007659050047384602,
      "loss": 0.7825,
      "num_input_tokens_seen": 70797000,
      "step": 121935
    },
    {
      "epoch": 18.162049448912718,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0007652900921215099,
      "loss": 0.7987,
      "num_input_tokens_seen": 70799880,
      "step": 121940
    },
    {
      "epoch": 18.16279416145368,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0007646754199855132,
      "loss": 0.7806,
      "num_input_tokens_seen": 70802856,
      "step": 121945
    },
    {
      "epoch": 18.16353887399464,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0007640609883408583,
      "loss": 0.7979,
      "num_input_tokens_seen": 70805928,
      "step": 121950
    },
    {
      "epoch": 18.164283586535596,
      "grad_norm": 0.02734375,
      "learning_rate": 0.0007634467971979236,
      "loss": 0.8036,
      "num_input_tokens_seen": 70808712,
      "step": 121955
    },
    {
      "epoch": 18.165028299076557,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0007628328465670874,
      "loss": 0.8145,
      "num_input_tokens_seen": 70811784,
      "step": 121960
    },
    {
      "epoch": 18.165773011617517,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0007622191364587183,
      "loss": 0.7996,
      "num_input_tokens_seen": 70815016,
      "step": 121965
    },
    {
      "epoch": 18.166517724158474,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0007616056668831877,
      "loss": 0.8186,
      "num_input_tokens_seen": 70818216,
      "step": 121970
    },
    {
      "epoch": 18.167262436699435,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0007609924378508575,
      "loss": 0.8183,
      "num_input_tokens_seen": 70821160,
      "step": 121975
    },
    {
      "epoch": 18.168007149240392,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0007603794493720894,
      "loss": 0.8003,
      "num_input_tokens_seen": 70824232,
      "step": 121980
    },
    {
      "epoch": 18.168751861781352,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0007597667014572367,
      "loss": 0.7978,
      "num_input_tokens_seen": 70827240,
      "step": 121985
    },
    {
      "epoch": 18.169496574322313,
      "grad_norm": 0.365234375,
      "learning_rate": 0.0007591541941166546,
      "loss": 0.8277,
      "num_input_tokens_seen": 70830344,
      "step": 121990
    },
    {
      "epoch": 18.17024128686327,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.000758541927360688,
      "loss": 0.7959,
      "num_input_tokens_seen": 70833224,
      "step": 121995
    },
    {
      "epoch": 18.17098599940423,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.0007579299011996788,
      "loss": 0.7976,
      "num_input_tokens_seen": 70835880,
      "step": 122000
    },
    {
      "epoch": 18.171730711945187,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0007573181156439701,
      "loss": 0.7926,
      "num_input_tokens_seen": 70839016,
      "step": 122005
    },
    {
      "epoch": 18.172475424486148,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0007567065707038956,
      "loss": 0.7939,
      "num_input_tokens_seen": 70842120,
      "step": 122010
    },
    {
      "epoch": 18.17322013702711,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0007560952663897885,
      "loss": 0.8031,
      "num_input_tokens_seen": 70845672,
      "step": 122015
    },
    {
      "epoch": 18.173964849568065,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0007554842027119756,
      "loss": 0.7931,
      "num_input_tokens_seen": 70848808,
      "step": 122020
    },
    {
      "epoch": 18.174709562109026,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.0007548733796807771,
      "loss": 0.7838,
      "num_input_tokens_seen": 70851560,
      "step": 122025
    },
    {
      "epoch": 18.175454274649987,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0007542627973065163,
      "loss": 0.7966,
      "num_input_tokens_seen": 70854472,
      "step": 122030
    },
    {
      "epoch": 18.176198987190944,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0007536524555995066,
      "loss": 0.8137,
      "num_input_tokens_seen": 70857224,
      "step": 122035
    },
    {
      "epoch": 18.176943699731904,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0007530423545700598,
      "loss": 0.7921,
      "num_input_tokens_seen": 70859784,
      "step": 122040
    },
    {
      "epoch": 18.17768841227286,
      "grad_norm": 0.095703125,
      "learning_rate": 0.000752432494228481,
      "loss": 0.7883,
      "num_input_tokens_seen": 70862792,
      "step": 122045
    },
    {
      "epoch": 18.17843312481382,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0007518228745850769,
      "loss": 0.793,
      "num_input_tokens_seen": 70865544,
      "step": 122050
    },
    {
      "epoch": 18.179177837354782,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.0007512134956501442,
      "loss": 0.8375,
      "num_input_tokens_seen": 70868328,
      "step": 122055
    },
    {
      "epoch": 18.17992254989574,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0007506043574339765,
      "loss": 0.7905,
      "num_input_tokens_seen": 70871464,
      "step": 122060
    },
    {
      "epoch": 18.1806672624367,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0007499954599468672,
      "loss": 0.8025,
      "num_input_tokens_seen": 70874440,
      "step": 122065
    },
    {
      "epoch": 18.18141197497766,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.0007493868031990979,
      "loss": 0.7877,
      "num_input_tokens_seen": 70877416,
      "step": 122070
    },
    {
      "epoch": 18.182156687518617,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.000748778387200959,
      "loss": 0.7963,
      "num_input_tokens_seen": 70880680,
      "step": 122075
    },
    {
      "epoch": 18.182901400059578,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0007481702119627203,
      "loss": 0.7845,
      "num_input_tokens_seen": 70883464,
      "step": 122080
    },
    {
      "epoch": 18.183646112600535,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0007475622774946655,
      "loss": 0.7918,
      "num_input_tokens_seen": 70886536,
      "step": 122085
    },
    {
      "epoch": 18.184390825141495,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0007469545838070579,
      "loss": 0.8063,
      "num_input_tokens_seen": 70889704,
      "step": 122090
    },
    {
      "epoch": 18.185135537682456,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0007463471309101676,
      "loss": 0.8023,
      "num_input_tokens_seen": 70892776,
      "step": 122095
    },
    {
      "epoch": 18.185880250223413,
      "grad_norm": 0.024169921875,
      "learning_rate": 0.0007457399188142566,
      "loss": 0.813,
      "num_input_tokens_seen": 70895752,
      "step": 122100
    },
    {
      "epoch": 18.186624962764373,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0007451329475295798,
      "loss": 0.7963,
      "num_input_tokens_seen": 70898632,
      "step": 122105
    },
    {
      "epoch": 18.187369675305334,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0007445262170663957,
      "loss": 0.8018,
      "num_input_tokens_seen": 70901384,
      "step": 122110
    },
    {
      "epoch": 18.18811438784629,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0007439197274349496,
      "loss": 0.7867,
      "num_input_tokens_seen": 70904232,
      "step": 122115
    },
    {
      "epoch": 18.18885910038725,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0007433134786454948,
      "loss": 0.7848,
      "num_input_tokens_seen": 70907016,
      "step": 122120
    },
    {
      "epoch": 18.18960381292821,
      "grad_norm": 0.02880859375,
      "learning_rate": 0.0007427074707082664,
      "loss": 0.7884,
      "num_input_tokens_seen": 70910024,
      "step": 122125
    },
    {
      "epoch": 18.19034852546917,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0007421017036335065,
      "loss": 0.7769,
      "num_input_tokens_seen": 70913000,
      "step": 122130
    },
    {
      "epoch": 18.19109323801013,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0007414961774314449,
      "loss": 0.796,
      "num_input_tokens_seen": 70915688,
      "step": 122135
    },
    {
      "epoch": 18.191837950551086,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.000740890892112317,
      "loss": 0.8308,
      "num_input_tokens_seen": 70918376,
      "step": 122140
    },
    {
      "epoch": 18.192582663092047,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0007402858476863427,
      "loss": 0.7883,
      "num_input_tokens_seen": 70921448,
      "step": 122145
    },
    {
      "epoch": 18.193327375633007,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0007396810441637457,
      "loss": 0.7946,
      "num_input_tokens_seen": 70924360,
      "step": 122150
    },
    {
      "epoch": 18.194072088173964,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0007390764815547446,
      "loss": 0.7872,
      "num_input_tokens_seen": 70927016,
      "step": 122155
    },
    {
      "epoch": 18.194816800714925,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.000738472159869551,
      "loss": 0.78,
      "num_input_tokens_seen": 70929864,
      "step": 122160
    },
    {
      "epoch": 18.195561513255882,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0007378680791183783,
      "loss": 0.8096,
      "num_input_tokens_seen": 70932488,
      "step": 122165
    },
    {
      "epoch": 18.196306225796842,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.000737264239311427,
      "loss": 0.7852,
      "num_input_tokens_seen": 70935560,
      "step": 122170
    },
    {
      "epoch": 18.197050938337803,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0007366606404589004,
      "loss": 0.8253,
      "num_input_tokens_seen": 70938408,
      "step": 122175
    },
    {
      "epoch": 18.19779565087876,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0007360572825709971,
      "loss": 0.7921,
      "num_input_tokens_seen": 70941192,
      "step": 122180
    },
    {
      "epoch": 18.19854036341972,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0007354541656579072,
      "loss": 0.7856,
      "num_input_tokens_seen": 70943944,
      "step": 122185
    },
    {
      "epoch": 18.199285075960677,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0007348512897298226,
      "loss": 0.7964,
      "num_input_tokens_seen": 70946568,
      "step": 122190
    },
    {
      "epoch": 18.200029788501638,
      "grad_norm": 0.02978515625,
      "learning_rate": 0.0007342486547969268,
      "loss": 0.8049,
      "num_input_tokens_seen": 70949160,
      "step": 122195
    },
    {
      "epoch": 18.2007745010426,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0007336462608694016,
      "loss": 0.7994,
      "num_input_tokens_seen": 70951816,
      "step": 122200
    },
    {
      "epoch": 18.201519213583556,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0007330441079574206,
      "loss": 0.802,
      "num_input_tokens_seen": 70954856,
      "step": 122205
    },
    {
      "epoch": 18.202263926124516,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0007324421960711608,
      "loss": 0.7934,
      "num_input_tokens_seen": 70957640,
      "step": 122210
    },
    {
      "epoch": 18.203008638665477,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0007318405252207887,
      "loss": 0.8037,
      "num_input_tokens_seen": 70960424,
      "step": 122215
    },
    {
      "epoch": 18.203753351206434,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0007312390954164699,
      "loss": 0.7951,
      "num_input_tokens_seen": 70963592,
      "step": 122220
    },
    {
      "epoch": 18.204498063747394,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0007306379066683643,
      "loss": 0.802,
      "num_input_tokens_seen": 70966344,
      "step": 122225
    },
    {
      "epoch": 18.20524277628835,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0007300369589866273,
      "loss": 0.7898,
      "num_input_tokens_seen": 70969224,
      "step": 122230
    },
    {
      "epoch": 18.20598748882931,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0007294362523814158,
      "loss": 0.8031,
      "num_input_tokens_seen": 70972072,
      "step": 122235
    },
    {
      "epoch": 18.206732201370272,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0007288357868628714,
      "loss": 0.7798,
      "num_input_tokens_seen": 70974824,
      "step": 122240
    },
    {
      "epoch": 18.20747691391123,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.000728235562441143,
      "loss": 0.794,
      "num_input_tokens_seen": 70977672,
      "step": 122245
    },
    {
      "epoch": 18.20822162645219,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0007276355791263688,
      "loss": 0.7942,
      "num_input_tokens_seen": 70980424,
      "step": 122250
    },
    {
      "epoch": 18.20896633899315,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0007270358369286877,
      "loss": 0.7976,
      "num_input_tokens_seen": 70983240,
      "step": 122255
    },
    {
      "epoch": 18.209711051534107,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0007264363358582298,
      "loss": 0.7978,
      "num_input_tokens_seen": 70986408,
      "step": 122260
    },
    {
      "epoch": 18.210455764075068,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.000725837075925122,
      "loss": 0.7752,
      "num_input_tokens_seen": 70989224,
      "step": 122265
    },
    {
      "epoch": 18.211200476616025,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0007252380571394895,
      "loss": 0.8167,
      "num_input_tokens_seen": 70992264,
      "step": 122270
    },
    {
      "epoch": 18.211945189156985,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0007246392795114509,
      "loss": 0.8009,
      "num_input_tokens_seen": 70995176,
      "step": 122275
    },
    {
      "epoch": 18.212689901697946,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.000724040743051123,
      "loss": 0.8026,
      "num_input_tokens_seen": 70997768,
      "step": 122280
    },
    {
      "epoch": 18.213434614238903,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.000723442447768618,
      "loss": 0.8117,
      "num_input_tokens_seen": 71000712,
      "step": 122285
    },
    {
      "epoch": 18.214179326779863,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0007228443936740425,
      "loss": 0.7859,
      "num_input_tokens_seen": 71003368,
      "step": 122290
    },
    {
      "epoch": 18.214924039320824,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0007222465807774985,
      "loss": 0.7784,
      "num_input_tokens_seen": 71006664,
      "step": 122295
    },
    {
      "epoch": 18.21566875186178,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0007216490090890898,
      "loss": 0.804,
      "num_input_tokens_seen": 71009544,
      "step": 122300
    },
    {
      "epoch": 18.21641346440274,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0007210516786189097,
      "loss": 0.7805,
      "num_input_tokens_seen": 71012232,
      "step": 122305
    },
    {
      "epoch": 18.2171581769437,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.000720454589377047,
      "loss": 0.7919,
      "num_input_tokens_seen": 71015304,
      "step": 122310
    },
    {
      "epoch": 18.21790288948466,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0007198577413735918,
      "loss": 0.8196,
      "num_input_tokens_seen": 71018248,
      "step": 122315
    },
    {
      "epoch": 18.21864760202562,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0007192611346186278,
      "loss": 0.7877,
      "num_input_tokens_seen": 71021192,
      "step": 122320
    },
    {
      "epoch": 18.219392314566576,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.000718664769122232,
      "loss": 0.8124,
      "num_input_tokens_seen": 71023816,
      "step": 122325
    },
    {
      "epoch": 18.220137027107537,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0007180686448944779,
      "loss": 0.7964,
      "num_input_tokens_seen": 71026600,
      "step": 122330
    },
    {
      "epoch": 18.220881739648497,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.0007174727619454407,
      "loss": 0.8123,
      "num_input_tokens_seen": 71029480,
      "step": 122335
    },
    {
      "epoch": 18.221626452189454,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0007168771202851825,
      "loss": 0.7886,
      "num_input_tokens_seen": 71032328,
      "step": 122340
    },
    {
      "epoch": 18.222371164730415,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0007162817199237703,
      "loss": 0.8023,
      "num_input_tokens_seen": 71035176,
      "step": 122345
    },
    {
      "epoch": 18.223115877271372,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0007156865608712625,
      "loss": 0.8,
      "num_input_tokens_seen": 71037992,
      "step": 122350
    },
    {
      "epoch": 18.223860589812332,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0007150916431377097,
      "loss": 0.7912,
      "num_input_tokens_seen": 71040872,
      "step": 122355
    },
    {
      "epoch": 18.224605302353293,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0007144969667331668,
      "loss": 0.8177,
      "num_input_tokens_seen": 71043912,
      "step": 122360
    },
    {
      "epoch": 18.22535001489425,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0007139025316676761,
      "loss": 0.8157,
      "num_input_tokens_seen": 71046888,
      "step": 122365
    },
    {
      "epoch": 18.22609472743521,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.000713308337951286,
      "loss": 0.795,
      "num_input_tokens_seen": 71049544,
      "step": 122370
    },
    {
      "epoch": 18.226839439976168,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0007127143855940287,
      "loss": 0.7944,
      "num_input_tokens_seen": 71052328,
      "step": 122375
    },
    {
      "epoch": 18.227584152517128,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0007121206746059411,
      "loss": 0.7896,
      "num_input_tokens_seen": 71055176,
      "step": 122380
    },
    {
      "epoch": 18.22832886505809,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.000711527204997055,
      "loss": 0.7924,
      "num_input_tokens_seen": 71058216,
      "step": 122385
    },
    {
      "epoch": 18.229073577599046,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0007109339767773942,
      "loss": 0.794,
      "num_input_tokens_seen": 71060936,
      "step": 122390
    },
    {
      "epoch": 18.229818290140006,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0007103409899569807,
      "loss": 0.7733,
      "num_input_tokens_seen": 71063592,
      "step": 122395
    },
    {
      "epoch": 18.230563002680967,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0007097482445458313,
      "loss": 0.7917,
      "num_input_tokens_seen": 71066504,
      "step": 122400
    },
    {
      "epoch": 18.231307715221924,
      "grad_norm": 0.09375,
      "learning_rate": 0.0007091557405539634,
      "loss": 0.7944,
      "num_input_tokens_seen": 71069384,
      "step": 122405
    },
    {
      "epoch": 18.232052427762884,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.0007085634779913819,
      "loss": 0.7899,
      "num_input_tokens_seen": 71072392,
      "step": 122410
    },
    {
      "epoch": 18.23279714030384,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0007079714568680989,
      "loss": 0.7974,
      "num_input_tokens_seen": 71075496,
      "step": 122415
    },
    {
      "epoch": 18.2335418528448,
      "grad_norm": 0.03125,
      "learning_rate": 0.0007073796771941081,
      "loss": 0.8099,
      "num_input_tokens_seen": 71078216,
      "step": 122420
    },
    {
      "epoch": 18.234286565385762,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0007067881389794151,
      "loss": 0.7981,
      "num_input_tokens_seen": 71081576,
      "step": 122425
    },
    {
      "epoch": 18.23503127792672,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0007061968422340098,
      "loss": 0.7824,
      "num_input_tokens_seen": 71084264,
      "step": 122430
    },
    {
      "epoch": 18.23577599046768,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0007056057869678778,
      "loss": 0.7921,
      "num_input_tokens_seen": 71086984,
      "step": 122435
    },
    {
      "epoch": 18.23652070300864,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.000705014973191011,
      "loss": 0.7935,
      "num_input_tokens_seen": 71089736,
      "step": 122440
    },
    {
      "epoch": 18.237265415549597,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0007044244009133849,
      "loss": 0.7944,
      "num_input_tokens_seen": 71092776,
      "step": 122445
    },
    {
      "epoch": 18.238010128090558,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0007038340701449813,
      "loss": 0.7833,
      "num_input_tokens_seen": 71095592,
      "step": 122450
    },
    {
      "epoch": 18.238754840631515,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0007032439808957707,
      "loss": 0.7902,
      "num_input_tokens_seen": 71098856,
      "step": 122455
    },
    {
      "epoch": 18.239499553172475,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0007026541331757219,
      "loss": 0.7987,
      "num_input_tokens_seen": 71101800,
      "step": 122460
    },
    {
      "epoch": 18.240244265713436,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0007020645269948017,
      "loss": 0.8045,
      "num_input_tokens_seen": 71104840,
      "step": 122465
    },
    {
      "epoch": 18.240988978254393,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0007014751623629673,
      "loss": 0.7863,
      "num_input_tokens_seen": 71107816,
      "step": 122470
    },
    {
      "epoch": 18.241733690795353,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0007008860392901789,
      "loss": 0.7721,
      "num_input_tokens_seen": 71110984,
      "step": 122475
    },
    {
      "epoch": 18.242478403336314,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0007002971577863869,
      "loss": 0.807,
      "num_input_tokens_seen": 71113800,
      "step": 122480
    },
    {
      "epoch": 18.24322311587727,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0006997085178615419,
      "loss": 0.7847,
      "num_input_tokens_seen": 71116776,
      "step": 122485
    },
    {
      "epoch": 18.24396782841823,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0006991201195255858,
      "loss": 0.8072,
      "num_input_tokens_seen": 71119848,
      "step": 122490
    },
    {
      "epoch": 18.24471254095919,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.0006985319627884623,
      "loss": 0.8087,
      "num_input_tokens_seen": 71122504,
      "step": 122495
    },
    {
      "epoch": 18.24545725350015,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0006979440476601051,
      "loss": 0.8023,
      "num_input_tokens_seen": 71125224,
      "step": 122500
    },
    {
      "epoch": 18.24620196604111,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0006973563741504479,
      "loss": 0.8098,
      "num_input_tokens_seen": 71127656,
      "step": 122505
    },
    {
      "epoch": 18.246946678582066,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0006967689422694178,
      "loss": 0.7896,
      "num_input_tokens_seen": 71130568,
      "step": 122510
    },
    {
      "epoch": 18.247691391123027,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0006961817520269386,
      "loss": 0.7784,
      "num_input_tokens_seen": 71133736,
      "step": 122515
    },
    {
      "epoch": 18.248436103663984,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0006955948034329323,
      "loss": 0.7862,
      "num_input_tokens_seen": 71136968,
      "step": 122520
    },
    {
      "epoch": 18.249180816204944,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0006950080964973143,
      "loss": 0.7857,
      "num_input_tokens_seen": 71139752,
      "step": 122525
    },
    {
      "epoch": 18.249925528745905,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0006944216312299933,
      "loss": 0.7861,
      "num_input_tokens_seen": 71142440,
      "step": 122530
    },
    {
      "epoch": 18.250670241286862,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0006938354076408798,
      "loss": 0.7807,
      "num_input_tokens_seen": 71145224,
      "step": 122535
    },
    {
      "epoch": 18.251414953827823,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0006932494257398774,
      "loss": 0.8178,
      "num_input_tokens_seen": 71147944,
      "step": 122540
    },
    {
      "epoch": 18.252159666368783,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0006926636855368867,
      "loss": 0.7825,
      "num_input_tokens_seen": 71151016,
      "step": 122545
    },
    {
      "epoch": 18.25290437890974,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.000692078187041798,
      "loss": 0.8,
      "num_input_tokens_seen": 71154184,
      "step": 122550
    },
    {
      "epoch": 18.2536490914507,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0006914929302645101,
      "loss": 0.8047,
      "num_input_tokens_seen": 71157064,
      "step": 122555
    },
    {
      "epoch": 18.254393803991658,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0006909079152149034,
      "loss": 0.7947,
      "num_input_tokens_seen": 71160008,
      "step": 122560
    },
    {
      "epoch": 18.255138516532618,
      "grad_norm": 0.02685546875,
      "learning_rate": 0.0006903231419028666,
      "loss": 0.8024,
      "num_input_tokens_seen": 71162664,
      "step": 122565
    },
    {
      "epoch": 18.25588322907358,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0006897386103382752,
      "loss": 0.8057,
      "num_input_tokens_seen": 71165640,
      "step": 122570
    },
    {
      "epoch": 18.256627941614536,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0006891543205310063,
      "loss": 0.7894,
      "num_input_tokens_seen": 71168328,
      "step": 122575
    },
    {
      "epoch": 18.257372654155496,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0006885702724909287,
      "loss": 0.7841,
      "num_input_tokens_seen": 71171304,
      "step": 122580
    },
    {
      "epoch": 18.258117366696457,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0006879864662279145,
      "loss": 0.8044,
      "num_input_tokens_seen": 71174408,
      "step": 122585
    },
    {
      "epoch": 18.258862079237414,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0006874029017518207,
      "loss": 0.791,
      "num_input_tokens_seen": 71177224,
      "step": 122590
    },
    {
      "epoch": 18.259606791778374,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0006868195790725095,
      "loss": 0.7908,
      "num_input_tokens_seen": 71180424,
      "step": 122595
    },
    {
      "epoch": 18.26035150431933,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0006862364981998331,
      "loss": 0.8017,
      "num_input_tokens_seen": 71183144,
      "step": 122600
    },
    {
      "epoch": 18.26109621686029,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0006856536591436435,
      "loss": 0.803,
      "num_input_tokens_seen": 71186568,
      "step": 122605
    },
    {
      "epoch": 18.261840929401252,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0006850710619137862,
      "loss": 0.7937,
      "num_input_tokens_seen": 71189576,
      "step": 122610
    },
    {
      "epoch": 18.26258564194221,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0006844887065201033,
      "loss": 0.7955,
      "num_input_tokens_seen": 71192392,
      "step": 122615
    },
    {
      "epoch": 18.26333035448317,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.000683906592972437,
      "loss": 0.7863,
      "num_input_tokens_seen": 71195400,
      "step": 122620
    },
    {
      "epoch": 18.26407506702413,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0006833247212806176,
      "loss": 0.7926,
      "num_input_tokens_seen": 71198344,
      "step": 122625
    },
    {
      "epoch": 18.264819779565087,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0006827430914544741,
      "loss": 0.8075,
      "num_input_tokens_seen": 71201192,
      "step": 122630
    },
    {
      "epoch": 18.265564492106048,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0006821617035038368,
      "loss": 0.7876,
      "num_input_tokens_seen": 71204040,
      "step": 122635
    },
    {
      "epoch": 18.266309204647005,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0006815805574385247,
      "loss": 0.7988,
      "num_input_tokens_seen": 71206792,
      "step": 122640
    },
    {
      "epoch": 18.267053917187965,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0006809996532683566,
      "loss": 0.777,
      "num_input_tokens_seen": 71210088,
      "step": 122645
    },
    {
      "epoch": 18.267798629728926,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0006804189910031477,
      "loss": 0.7896,
      "num_input_tokens_seen": 71213000,
      "step": 122650
    },
    {
      "epoch": 18.268543342269883,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0006798385706527054,
      "loss": 0.8099,
      "num_input_tokens_seen": 71215848,
      "step": 122655
    },
    {
      "epoch": 18.269288054810843,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0006792583922268336,
      "loss": 0.7853,
      "num_input_tokens_seen": 71218792,
      "step": 122660
    },
    {
      "epoch": 18.270032767351804,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0006786784557353393,
      "loss": 0.8015,
      "num_input_tokens_seen": 71221608,
      "step": 122665
    },
    {
      "epoch": 18.27077747989276,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0006780987611880163,
      "loss": 0.81,
      "num_input_tokens_seen": 71224520,
      "step": 122670
    },
    {
      "epoch": 18.27152219243372,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0006775193085946568,
      "loss": 0.7957,
      "num_input_tokens_seen": 71227624,
      "step": 122675
    },
    {
      "epoch": 18.27226690497468,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0006769400979650547,
      "loss": 0.8114,
      "num_input_tokens_seen": 71230408,
      "step": 122680
    },
    {
      "epoch": 18.27301161751564,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0006763611293089889,
      "loss": 0.7951,
      "num_input_tokens_seen": 71233128,
      "step": 122685
    },
    {
      "epoch": 18.2737563300566,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0006757824026362463,
      "loss": 0.7897,
      "num_input_tokens_seen": 71235848,
      "step": 122690
    },
    {
      "epoch": 18.274501042597556,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.000675203917956601,
      "loss": 0.8078,
      "num_input_tokens_seen": 71238984,
      "step": 122695
    },
    {
      "epoch": 18.275245755138517,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0006746256752798268,
      "loss": 0.8226,
      "num_input_tokens_seen": 71241928,
      "step": 122700
    },
    {
      "epoch": 18.275990467679474,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0006740476746156926,
      "loss": 0.7987,
      "num_input_tokens_seen": 71245192,
      "step": 122705
    },
    {
      "epoch": 18.276735180220435,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0006734699159739604,
      "loss": 0.7928,
      "num_input_tokens_seen": 71247848,
      "step": 122710
    },
    {
      "epoch": 18.277479892761395,
      "grad_norm": 0.03515625,
      "learning_rate": 0.000672892399364396,
      "loss": 0.7888,
      "num_input_tokens_seen": 71250504,
      "step": 122715
    },
    {
      "epoch": 18.278224605302352,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0006723151247967529,
      "loss": 0.7976,
      "num_input_tokens_seen": 71253416,
      "step": 122720
    },
    {
      "epoch": 18.278969317843313,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0006717380922807836,
      "loss": 0.7795,
      "num_input_tokens_seen": 71256328,
      "step": 122725
    },
    {
      "epoch": 18.279714030384273,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0006711613018262336,
      "loss": 0.7946,
      "num_input_tokens_seen": 71259176,
      "step": 122730
    },
    {
      "epoch": 18.28045874292523,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0006705847534428532,
      "loss": 0.7777,
      "num_input_tokens_seen": 71262088,
      "step": 122735
    },
    {
      "epoch": 18.28120345546619,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0006700084471403766,
      "loss": 0.7883,
      "num_input_tokens_seen": 71265192,
      "step": 122740
    },
    {
      "epoch": 18.281948168007148,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0006694323829285458,
      "loss": 0.8065,
      "num_input_tokens_seen": 71267944,
      "step": 122745
    },
    {
      "epoch": 18.282692880548108,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0006688565608170916,
      "loss": 0.7926,
      "num_input_tokens_seen": 71270984,
      "step": 122750
    },
    {
      "epoch": 18.28343759308907,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.0006682809808157359,
      "loss": 0.7951,
      "num_input_tokens_seen": 71273736,
      "step": 122755
    },
    {
      "epoch": 18.284182305630026,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0006677056429342093,
      "loss": 0.7909,
      "num_input_tokens_seen": 71276808,
      "step": 122760
    },
    {
      "epoch": 18.284927018170986,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0006671305471822275,
      "loss": 0.7899,
      "num_input_tokens_seen": 71279496,
      "step": 122765
    },
    {
      "epoch": 18.285671730711947,
      "grad_norm": 0.0625,
      "learning_rate": 0.0006665556935695093,
      "loss": 0.7776,
      "num_input_tokens_seen": 71282408,
      "step": 122770
    },
    {
      "epoch": 18.286416443252904,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0006659810821057638,
      "loss": 0.804,
      "num_input_tokens_seen": 71285480,
      "step": 122775
    },
    {
      "epoch": 18.287161155793864,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.0006654067128007013,
      "loss": 0.7976,
      "num_input_tokens_seen": 71288264,
      "step": 122780
    },
    {
      "epoch": 18.28790586833482,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0006648325856640225,
      "loss": 0.7977,
      "num_input_tokens_seen": 71291112,
      "step": 122785
    },
    {
      "epoch": 18.28865058087578,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0006642587007054296,
      "loss": 0.7948,
      "num_input_tokens_seen": 71294152,
      "step": 122790
    },
    {
      "epoch": 18.289395293416742,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0006636850579346132,
      "loss": 0.798,
      "num_input_tokens_seen": 71297160,
      "step": 122795
    },
    {
      "epoch": 18.2901400059577,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0006631116573612672,
      "loss": 0.7949,
      "num_input_tokens_seen": 71300264,
      "step": 122800
    },
    {
      "epoch": 18.29088471849866,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.0006625384989950788,
      "loss": 0.7941,
      "num_input_tokens_seen": 71303464,
      "step": 122805
    },
    {
      "epoch": 18.29162943103962,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0006619655828457287,
      "loss": 0.7851,
      "num_input_tokens_seen": 71306344,
      "step": 122810
    },
    {
      "epoch": 18.292374143580577,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0006613929089228992,
      "loss": 0.7687,
      "num_input_tokens_seen": 71309128,
      "step": 122815
    },
    {
      "epoch": 18.293118856121538,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0006608204772362624,
      "loss": 0.791,
      "num_input_tokens_seen": 71311944,
      "step": 122820
    },
    {
      "epoch": 18.293863568662495,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0006602482877954924,
      "loss": 0.7871,
      "num_input_tokens_seen": 71314920,
      "step": 122825
    },
    {
      "epoch": 18.294608281203455,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0006596763406102513,
      "loss": 0.8102,
      "num_input_tokens_seen": 71317704,
      "step": 122830
    },
    {
      "epoch": 18.295352993744416,
      "grad_norm": 0.048828125,
      "learning_rate": 0.000659104635690203,
      "loss": 0.8162,
      "num_input_tokens_seen": 71320744,
      "step": 122835
    },
    {
      "epoch": 18.296097706285373,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0006585331730450083,
      "loss": 0.7857,
      "num_input_tokens_seen": 71323976,
      "step": 122840
    },
    {
      "epoch": 18.296842418826333,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0006579619526843178,
      "loss": 0.8008,
      "num_input_tokens_seen": 71326664,
      "step": 122845
    },
    {
      "epoch": 18.297587131367294,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0006573909746177852,
      "loss": 0.7858,
      "num_input_tokens_seen": 71329864,
      "step": 122850
    },
    {
      "epoch": 18.29833184390825,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0006568202388550548,
      "loss": 0.8038,
      "num_input_tokens_seen": 71332968,
      "step": 122855
    },
    {
      "epoch": 18.29907655644921,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0006562497454057686,
      "loss": 0.7673,
      "num_input_tokens_seen": 71336136,
      "step": 122860
    },
    {
      "epoch": 18.29982126899017,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0006556794942795624,
      "loss": 0.7947,
      "num_input_tokens_seen": 71339080,
      "step": 122865
    },
    {
      "epoch": 18.30056598153113,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.000655109485486075,
      "loss": 0.8017,
      "num_input_tokens_seen": 71341896,
      "step": 122870
    },
    {
      "epoch": 18.30131069407209,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0006545397190349339,
      "loss": 0.7991,
      "num_input_tokens_seen": 71344648,
      "step": 122875
    },
    {
      "epoch": 18.302055406613047,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0006539701949357612,
      "loss": 0.7923,
      "num_input_tokens_seen": 71347496,
      "step": 122880
    },
    {
      "epoch": 18.302800119154007,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0006534009131981827,
      "loss": 0.8026,
      "num_input_tokens_seen": 71350408,
      "step": 122885
    },
    {
      "epoch": 18.303544831694964,
      "grad_norm": 0.060546875,
      "learning_rate": 0.000652831873831814,
      "loss": 0.7995,
      "num_input_tokens_seen": 71353064,
      "step": 122890
    },
    {
      "epoch": 18.304289544235925,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0006522630768462706,
      "loss": 0.7942,
      "num_input_tokens_seen": 71355848,
      "step": 122895
    },
    {
      "epoch": 18.305034256776885,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0006516945222511566,
      "loss": 0.7784,
      "num_input_tokens_seen": 71358536,
      "step": 122900
    },
    {
      "epoch": 18.305778969317842,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0006511262100560844,
      "loss": 0.8036,
      "num_input_tokens_seen": 71361384,
      "step": 122905
    },
    {
      "epoch": 18.306523681858803,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0006505581402706512,
      "loss": 0.7988,
      "num_input_tokens_seen": 71364520,
      "step": 122910
    },
    {
      "epoch": 18.307268394399763,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.000649990312904451,
      "loss": 0.7994,
      "num_input_tokens_seen": 71367304,
      "step": 122915
    },
    {
      "epoch": 18.30801310694072,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.0006494227279670827,
      "loss": 0.8084,
      "num_input_tokens_seen": 71370120,
      "step": 122920
    },
    {
      "epoch": 18.30875781948168,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0006488553854681323,
      "loss": 0.7842,
      "num_input_tokens_seen": 71372648,
      "step": 122925
    },
    {
      "epoch": 18.309502532022638,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0006482882854171834,
      "loss": 0.8028,
      "num_input_tokens_seen": 71375624,
      "step": 122930
    },
    {
      "epoch": 18.310247244563598,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0006477214278238169,
      "loss": 0.8114,
      "num_input_tokens_seen": 71378472,
      "step": 122935
    },
    {
      "epoch": 18.31099195710456,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0006471548126976101,
      "loss": 0.8215,
      "num_input_tokens_seen": 71381352,
      "step": 122940
    },
    {
      "epoch": 18.311736669645516,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0006465884400481353,
      "loss": 0.7878,
      "num_input_tokens_seen": 71384392,
      "step": 122945
    },
    {
      "epoch": 18.312481382186476,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0006460223098849616,
      "loss": 0.8002,
      "num_input_tokens_seen": 71387400,
      "step": 122950
    },
    {
      "epoch": 18.313226094727437,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0006454564222176512,
      "loss": 0.7788,
      "num_input_tokens_seen": 71390600,
      "step": 122955
    },
    {
      "epoch": 18.313970807268394,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0006448907770557649,
      "loss": 0.7845,
      "num_input_tokens_seen": 71393832,
      "step": 122960
    },
    {
      "epoch": 18.314715519809354,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.00064432537440886,
      "loss": 0.7958,
      "num_input_tokens_seen": 71397032,
      "step": 122965
    },
    {
      "epoch": 18.31546023235031,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0006437602142864856,
      "loss": 0.7973,
      "num_input_tokens_seen": 71399848,
      "step": 122970
    },
    {
      "epoch": 18.316204944891272,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0006431952966981924,
      "loss": 0.7966,
      "num_input_tokens_seen": 71402920,
      "step": 122975
    },
    {
      "epoch": 18.316949657432232,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0006426306216535244,
      "loss": 0.7974,
      "num_input_tokens_seen": 71405960,
      "step": 122980
    },
    {
      "epoch": 18.31769436997319,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0006420661891620205,
      "loss": 0.7993,
      "num_input_tokens_seen": 71408584,
      "step": 122985
    },
    {
      "epoch": 18.31843908251415,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0006415019992332116,
      "loss": 0.7893,
      "num_input_tokens_seen": 71411208,
      "step": 122990
    },
    {
      "epoch": 18.31918379505511,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0006409380518766366,
      "loss": 0.7878,
      "num_input_tokens_seen": 71413800,
      "step": 122995
    },
    {
      "epoch": 18.319928507596067,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0006403743471018197,
      "loss": 0.7872,
      "num_input_tokens_seen": 71416712,
      "step": 123000
    },
    {
      "epoch": 18.320673220137028,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0006398108849182798,
      "loss": 0.7955,
      "num_input_tokens_seen": 71419912,
      "step": 123005
    },
    {
      "epoch": 18.321417932677985,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0006392476653355428,
      "loss": 0.7629,
      "num_input_tokens_seen": 71422728,
      "step": 123010
    },
    {
      "epoch": 18.322162645218945,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0006386846883631191,
      "loss": 0.7866,
      "num_input_tokens_seen": 71425608,
      "step": 123015
    },
    {
      "epoch": 18.322907357759906,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0006381219540105248,
      "loss": 0.7978,
      "num_input_tokens_seen": 71428680,
      "step": 123020
    },
    {
      "epoch": 18.323652070300863,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0006375594622872588,
      "loss": 0.7907,
      "num_input_tokens_seen": 71431496,
      "step": 123025
    },
    {
      "epoch": 18.324396782841823,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0006369972132028317,
      "loss": 0.8169,
      "num_input_tokens_seen": 71434280,
      "step": 123030
    },
    {
      "epoch": 18.32514149538278,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.0006364352067667394,
      "loss": 0.8025,
      "num_input_tokens_seen": 71437224,
      "step": 123035
    },
    {
      "epoch": 18.32588620792374,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0006358734429884726,
      "loss": 0.7896,
      "num_input_tokens_seen": 71440040,
      "step": 123040
    },
    {
      "epoch": 18.3266309204647,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0006353119218775272,
      "loss": 0.8115,
      "num_input_tokens_seen": 71442824,
      "step": 123045
    },
    {
      "epoch": 18.32737563300566,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0006347506434433869,
      "loss": 0.8128,
      "num_input_tokens_seen": 71445416,
      "step": 123050
    },
    {
      "epoch": 18.32812034554662,
      "grad_norm": 0.0242919921875,
      "learning_rate": 0.0006341896076955361,
      "loss": 0.8085,
      "num_input_tokens_seen": 71448264,
      "step": 123055
    },
    {
      "epoch": 18.32886505808758,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0006336288146434471,
      "loss": 0.8197,
      "num_input_tokens_seen": 71451336,
      "step": 123060
    },
    {
      "epoch": 18.329609770628537,
      "grad_norm": 0.0302734375,
      "learning_rate": 0.0006330682642966006,
      "loss": 0.7899,
      "num_input_tokens_seen": 71454184,
      "step": 123065
    },
    {
      "epoch": 18.330354483169497,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0006325079566644643,
      "loss": 0.7981,
      "num_input_tokens_seen": 71457128,
      "step": 123070
    },
    {
      "epoch": 18.331099195710454,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0006319478917565019,
      "loss": 0.7873,
      "num_input_tokens_seen": 71460040,
      "step": 123075
    },
    {
      "epoch": 18.331843908251415,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0006313880695821777,
      "loss": 0.8003,
      "num_input_tokens_seen": 71463176,
      "step": 123080
    },
    {
      "epoch": 18.332588620792375,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0006308284901509458,
      "loss": 0.7708,
      "num_input_tokens_seen": 71466120,
      "step": 123085
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.000630269153472267,
      "loss": 0.8131,
      "num_input_tokens_seen": 71468904,
      "step": 123090
    },
    {
      "epoch": 18.334078045874293,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.000629710059555582,
      "loss": 0.7944,
      "num_input_tokens_seen": 71471624,
      "step": 123095
    },
    {
      "epoch": 18.334822758415253,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0006291512084103434,
      "loss": 0.8132,
      "num_input_tokens_seen": 71474408,
      "step": 123100
    },
    {
      "epoch": 18.33556747095621,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.0006285926000459868,
      "loss": 0.8003,
      "num_input_tokens_seen": 71477480,
      "step": 123105
    },
    {
      "epoch": 18.33631218349717,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.000628034234471953,
      "loss": 0.798,
      "num_input_tokens_seen": 71480296,
      "step": 123110
    },
    {
      "epoch": 18.337056896038128,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0006274761116976745,
      "loss": 0.7738,
      "num_input_tokens_seen": 71483208,
      "step": 123115
    },
    {
      "epoch": 18.33780160857909,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0006269182317325788,
      "loss": 0.784,
      "num_input_tokens_seen": 71486280,
      "step": 123120
    },
    {
      "epoch": 18.33854632112005,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0006263605945860933,
      "loss": 0.799,
      "num_input_tokens_seen": 71489224,
      "step": 123125
    },
    {
      "epoch": 18.339291033661006,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0006258032002676339,
      "loss": 0.8016,
      "num_input_tokens_seen": 71492264,
      "step": 123130
    },
    {
      "epoch": 18.340035746201966,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0006252460487866212,
      "loss": 0.8016,
      "num_input_tokens_seen": 71494920,
      "step": 123135
    },
    {
      "epoch": 18.340780458742927,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0006246891401524662,
      "loss": 0.8185,
      "num_input_tokens_seen": 71497672,
      "step": 123140
    },
    {
      "epoch": 18.341525171283884,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0006241324743745779,
      "loss": 0.8028,
      "num_input_tokens_seen": 71500744,
      "step": 123145
    },
    {
      "epoch": 18.342269883824844,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.0006235760514623623,
      "loss": 0.8058,
      "num_input_tokens_seen": 71503624,
      "step": 123150
    },
    {
      "epoch": 18.3430145963658,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0006230198714252149,
      "loss": 0.8079,
      "num_input_tokens_seen": 71506472,
      "step": 123155
    },
    {
      "epoch": 18.343759308906762,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0006224639342725369,
      "loss": 0.792,
      "num_input_tokens_seen": 71509576,
      "step": 123160
    },
    {
      "epoch": 18.344504021447722,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0006219082400137154,
      "loss": 0.8006,
      "num_input_tokens_seen": 71512584,
      "step": 123165
    },
    {
      "epoch": 18.34524873398868,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0006213527886581433,
      "loss": 0.7905,
      "num_input_tokens_seen": 71515720,
      "step": 123170
    },
    {
      "epoch": 18.34599344652964,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.000620797580215201,
      "loss": 0.7967,
      "num_input_tokens_seen": 71518696,
      "step": 123175
    },
    {
      "epoch": 18.3467381590706,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0006202426146942697,
      "loss": 0.7954,
      "num_input_tokens_seen": 71521640,
      "step": 123180
    },
    {
      "epoch": 18.347482871611557,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0006196878921047249,
      "loss": 0.7968,
      "num_input_tokens_seen": 71524360,
      "step": 123185
    },
    {
      "epoch": 18.348227584152518,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0006191334124559378,
      "loss": 0.8082,
      "num_input_tokens_seen": 71527176,
      "step": 123190
    },
    {
      "epoch": 18.348972296693475,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0006185791757572756,
      "loss": 0.8007,
      "num_input_tokens_seen": 71530440,
      "step": 123195
    },
    {
      "epoch": 18.349717009234435,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0006180251820181009,
      "loss": 0.7986,
      "num_input_tokens_seen": 71533448,
      "step": 123200
    },
    {
      "epoch": 18.350461721775396,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0006174714312477747,
      "loss": 0.8039,
      "num_input_tokens_seen": 71536168,
      "step": 123205
    },
    {
      "epoch": 18.351206434316353,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0006169179234556476,
      "loss": 0.7865,
      "num_input_tokens_seen": 71538888,
      "step": 123210
    },
    {
      "epoch": 18.351951146857314,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.0006163646586510774,
      "loss": 0.7847,
      "num_input_tokens_seen": 71541736,
      "step": 123215
    },
    {
      "epoch": 18.35269585939827,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0006158116368434047,
      "loss": 0.8162,
      "num_input_tokens_seen": 71544648,
      "step": 123220
    },
    {
      "epoch": 18.35344057193923,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0006152588580419771,
      "loss": 0.808,
      "num_input_tokens_seen": 71547368,
      "step": 123225
    },
    {
      "epoch": 18.35418528448019,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0006147063222561322,
      "loss": 0.7757,
      "num_input_tokens_seen": 71550408,
      "step": 123230
    },
    {
      "epoch": 18.35492999702115,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0006141540294951991,
      "loss": 0.7791,
      "num_input_tokens_seen": 71552968,
      "step": 123235
    },
    {
      "epoch": 18.35567470956211,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0006136019797685172,
      "loss": 0.8113,
      "num_input_tokens_seen": 71555880,
      "step": 123240
    },
    {
      "epoch": 18.35641942210307,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0006130501730854037,
      "loss": 0.8037,
      "num_input_tokens_seen": 71558824,
      "step": 123245
    },
    {
      "epoch": 18.357164134644027,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0006124986094551898,
      "loss": 0.8052,
      "num_input_tokens_seen": 71561448,
      "step": 123250
    },
    {
      "epoch": 18.357908847184987,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0006119472888871863,
      "loss": 0.7992,
      "num_input_tokens_seen": 71564040,
      "step": 123255
    },
    {
      "epoch": 18.358653559725944,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0006113962113907123,
      "loss": 0.7914,
      "num_input_tokens_seen": 71567208,
      "step": 123260
    },
    {
      "epoch": 18.359398272266905,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0006108453769750721,
      "loss": 0.7887,
      "num_input_tokens_seen": 71569992,
      "step": 123265
    },
    {
      "epoch": 18.360142984807865,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.0006102947856495766,
      "loss": 0.7866,
      "num_input_tokens_seen": 71572840,
      "step": 123270
    },
    {
      "epoch": 18.360887697348822,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0006097444374235267,
      "loss": 0.8439,
      "num_input_tokens_seen": 71575720,
      "step": 123275
    },
    {
      "epoch": 18.361632409889783,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0006091943323062166,
      "loss": 0.7811,
      "num_input_tokens_seen": 71578344,
      "step": 123280
    },
    {
      "epoch": 18.362377122430743,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0006086444703069438,
      "loss": 0.7924,
      "num_input_tokens_seen": 71581224,
      "step": 123285
    },
    {
      "epoch": 18.3631218349717,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0006080948514349943,
      "loss": 0.7771,
      "num_input_tokens_seen": 71584232,
      "step": 123290
    },
    {
      "epoch": 18.36386654751266,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0006075454756996573,
      "loss": 0.7735,
      "num_input_tokens_seen": 71587176,
      "step": 123295
    },
    {
      "epoch": 18.364611260053618,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0006069963431102088,
      "loss": 0.7878,
      "num_input_tokens_seen": 71590120,
      "step": 123300
    },
    {
      "epoch": 18.36535597259458,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0006064474536759312,
      "loss": 0.7997,
      "num_input_tokens_seen": 71593256,
      "step": 123305
    },
    {
      "epoch": 18.36610068513554,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0006058988074060939,
      "loss": 0.8019,
      "num_input_tokens_seen": 71596200,
      "step": 123310
    },
    {
      "epoch": 18.366845397676496,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.0006053504043099677,
      "loss": 0.8409,
      "num_input_tokens_seen": 71598952,
      "step": 123315
    },
    {
      "epoch": 18.367590110217456,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0006048022443968154,
      "loss": 0.7957,
      "num_input_tokens_seen": 71601928,
      "step": 123320
    },
    {
      "epoch": 18.368334822758417,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0006042543276758994,
      "loss": 0.7987,
      "num_input_tokens_seen": 71605192,
      "step": 123325
    },
    {
      "epoch": 18.369079535299374,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0006037066541564756,
      "loss": 0.8071,
      "num_input_tokens_seen": 71608168,
      "step": 123330
    },
    {
      "epoch": 18.369824247840334,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0006031592238477934,
      "loss": 0.8056,
      "num_input_tokens_seen": 71611400,
      "step": 123335
    },
    {
      "epoch": 18.37056896038129,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0006026120367591053,
      "loss": 0.8002,
      "num_input_tokens_seen": 71614568,
      "step": 123340
    },
    {
      "epoch": 18.371313672922252,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.000602065092899654,
      "loss": 0.804,
      "num_input_tokens_seen": 71617352,
      "step": 123345
    },
    {
      "epoch": 18.372058385463212,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0006015183922786805,
      "loss": 0.791,
      "num_input_tokens_seen": 71620424,
      "step": 123350
    },
    {
      "epoch": 18.37280309800417,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0006009719349054204,
      "loss": 0.7875,
      "num_input_tokens_seen": 71623336,
      "step": 123355
    },
    {
      "epoch": 18.37354781054513,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0006004257207891017,
      "loss": 0.8022,
      "num_input_tokens_seen": 71626312,
      "step": 123360
    },
    {
      "epoch": 18.37429252308609,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0005998797499389585,
      "loss": 0.8005,
      "num_input_tokens_seen": 71629192,
      "step": 123365
    },
    {
      "epoch": 18.375037235627047,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0005993340223642084,
      "loss": 0.7892,
      "num_input_tokens_seen": 71632200,
      "step": 123370
    },
    {
      "epoch": 18.375781948168008,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.0005987885380740759,
      "loss": 0.8072,
      "num_input_tokens_seen": 71635080,
      "step": 123375
    },
    {
      "epoch": 18.376526660708965,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0005982432970777751,
      "loss": 0.7955,
      "num_input_tokens_seen": 71637896,
      "step": 123380
    },
    {
      "epoch": 18.377271373249926,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.000597698299384517,
      "loss": 0.8097,
      "num_input_tokens_seen": 71640840,
      "step": 123385
    },
    {
      "epoch": 18.378016085790886,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0005971535450035059,
      "loss": 0.7819,
      "num_input_tokens_seen": 71643720,
      "step": 123390
    },
    {
      "epoch": 18.378760798331843,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0005966090339439478,
      "loss": 0.7908,
      "num_input_tokens_seen": 71646632,
      "step": 123395
    },
    {
      "epoch": 18.379505510872804,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.000596064766215042,
      "loss": 0.8327,
      "num_input_tokens_seen": 71649800,
      "step": 123400
    },
    {
      "epoch": 18.38025022341376,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0005955207418259812,
      "loss": 0.8165,
      "num_input_tokens_seen": 71652488,
      "step": 123405
    },
    {
      "epoch": 18.38099493595472,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0005949769607859578,
      "loss": 0.8007,
      "num_input_tokens_seen": 71655336,
      "step": 123410
    },
    {
      "epoch": 18.38173964849568,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0005944334231041564,
      "loss": 0.7807,
      "num_input_tokens_seen": 71658216,
      "step": 123415
    },
    {
      "epoch": 18.38248436103664,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0005938901287897647,
      "loss": 0.7997,
      "num_input_tokens_seen": 71661128,
      "step": 123420
    },
    {
      "epoch": 18.3832290735776,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0005933470778519551,
      "loss": 0.7776,
      "num_input_tokens_seen": 71664232,
      "step": 123425
    },
    {
      "epoch": 18.38397378611856,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0005928042702999053,
      "loss": 0.8001,
      "num_input_tokens_seen": 71667336,
      "step": 123430
    },
    {
      "epoch": 18.384718498659517,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0005922617061427848,
      "loss": 0.7902,
      "num_input_tokens_seen": 71670344,
      "step": 123435
    },
    {
      "epoch": 18.385463211200477,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0005917193853897579,
      "loss": 0.8024,
      "num_input_tokens_seen": 71673064,
      "step": 123440
    },
    {
      "epoch": 18.386207923741434,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0005911773080499888,
      "loss": 0.8092,
      "num_input_tokens_seen": 71676296,
      "step": 123445
    },
    {
      "epoch": 18.386952636282395,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.000590635474132637,
      "loss": 0.8028,
      "num_input_tokens_seen": 71679208,
      "step": 123450
    },
    {
      "epoch": 18.387697348823355,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0005900938836468516,
      "loss": 0.8071,
      "num_input_tokens_seen": 71682344,
      "step": 123455
    },
    {
      "epoch": 18.388442061364312,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.000589552536601784,
      "loss": 0.7959,
      "num_input_tokens_seen": 71685096,
      "step": 123460
    },
    {
      "epoch": 18.389186773905273,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0005890114330065815,
      "loss": 0.7926,
      "num_input_tokens_seen": 71687560,
      "step": 123465
    },
    {
      "epoch": 18.389931486446233,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.000588470572870382,
      "loss": 0.7912,
      "num_input_tokens_seen": 71690536,
      "step": 123470
    },
    {
      "epoch": 18.39067619898719,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0005879299562023266,
      "loss": 0.7993,
      "num_input_tokens_seen": 71693096,
      "step": 123475
    },
    {
      "epoch": 18.39142091152815,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0005873895830115477,
      "loss": 0.7857,
      "num_input_tokens_seen": 71696168,
      "step": 123480
    },
    {
      "epoch": 18.392165624069108,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0005868494533071716,
      "loss": 0.779,
      "num_input_tokens_seen": 71699016,
      "step": 123485
    },
    {
      "epoch": 18.39291033661007,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.0005863095670983259,
      "loss": 0.8052,
      "num_input_tokens_seen": 71702184,
      "step": 123490
    },
    {
      "epoch": 18.39365504915103,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.00058576992439413,
      "loss": 0.7929,
      "num_input_tokens_seen": 71705000,
      "step": 123495
    },
    {
      "epoch": 18.394399761691986,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0005852305252037016,
      "loss": 0.7953,
      "num_input_tokens_seen": 71707848,
      "step": 123500
    },
    {
      "epoch": 18.395144474232946,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0005846913695361533,
      "loss": 0.792,
      "num_input_tokens_seen": 71710600,
      "step": 123505
    },
    {
      "epoch": 18.395889186773907,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.000584152457400593,
      "loss": 0.7917,
      "num_input_tokens_seen": 71713672,
      "step": 123510
    },
    {
      "epoch": 18.396633899314864,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0005836137888061265,
      "loss": 0.8282,
      "num_input_tokens_seen": 71716520,
      "step": 123515
    },
    {
      "epoch": 18.397378611855824,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0005830753637618519,
      "loss": 0.8089,
      "num_input_tokens_seen": 71719560,
      "step": 123520
    },
    {
      "epoch": 18.39812332439678,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0005825371822768682,
      "loss": 0.7918,
      "num_input_tokens_seen": 71722408,
      "step": 123525
    },
    {
      "epoch": 18.398868036937742,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.0005819992443602617,
      "loss": 0.8117,
      "num_input_tokens_seen": 71725096,
      "step": 123530
    },
    {
      "epoch": 18.399612749478703,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0005814615500211268,
      "loss": 0.7837,
      "num_input_tokens_seen": 71727912,
      "step": 123535
    },
    {
      "epoch": 18.40035746201966,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0005809240992685427,
      "loss": 0.7932,
      "num_input_tokens_seen": 71730632,
      "step": 123540
    },
    {
      "epoch": 18.40110217456062,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.000580386892111594,
      "loss": 0.8025,
      "num_input_tokens_seen": 71733288,
      "step": 123545
    },
    {
      "epoch": 18.401846887101577,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.00057984992855935,
      "loss": 0.7977,
      "num_input_tokens_seen": 71736072,
      "step": 123550
    },
    {
      "epoch": 18.402591599642538,
      "grad_norm": 0.046875,
      "learning_rate": 0.0005793132086208885,
      "loss": 0.8011,
      "num_input_tokens_seen": 71739240,
      "step": 123555
    },
    {
      "epoch": 18.403336312183498,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0005787767323052722,
      "loss": 0.7779,
      "num_input_tokens_seen": 71742024,
      "step": 123560
    },
    {
      "epoch": 18.404081024724455,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0005782404996215656,
      "loss": 0.8027,
      "num_input_tokens_seen": 71744680,
      "step": 123565
    },
    {
      "epoch": 18.404825737265416,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0005777045105788297,
      "loss": 0.8004,
      "num_input_tokens_seen": 71747592,
      "step": 123570
    },
    {
      "epoch": 18.405570449806376,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0005771687651861157,
      "loss": 0.8045,
      "num_input_tokens_seen": 71750664,
      "step": 123575
    },
    {
      "epoch": 18.406315162347333,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0005766332634524778,
      "loss": 0.7945,
      "num_input_tokens_seen": 71753416,
      "step": 123580
    },
    {
      "epoch": 18.407059874888294,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0005760980053869624,
      "loss": 0.7993,
      "num_input_tokens_seen": 71756648,
      "step": 123585
    },
    {
      "epoch": 18.40780458742925,
      "grad_norm": 0.33984375,
      "learning_rate": 0.000575562990998612,
      "loss": 0.795,
      "num_input_tokens_seen": 71759432,
      "step": 123590
    },
    {
      "epoch": 18.40854929997021,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0005750282202964646,
      "loss": 0.8028,
      "num_input_tokens_seen": 71762088,
      "step": 123595
    },
    {
      "epoch": 18.40929401251117,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0005744936932895511,
      "loss": 0.7766,
      "num_input_tokens_seen": 71765160,
      "step": 123600
    },
    {
      "epoch": 18.41003872505213,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0005739594099869078,
      "loss": 0.7865,
      "num_input_tokens_seen": 71767912,
      "step": 123605
    },
    {
      "epoch": 18.41078343759309,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0005734253703975573,
      "loss": 0.8147,
      "num_input_tokens_seen": 71771016,
      "step": 123610
    },
    {
      "epoch": 18.41152815013405,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.000572891574530524,
      "loss": 0.7975,
      "num_input_tokens_seen": 71774184,
      "step": 123615
    },
    {
      "epoch": 18.412272862675007,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0005723580223948227,
      "loss": 0.8061,
      "num_input_tokens_seen": 71777256,
      "step": 123620
    },
    {
      "epoch": 18.413017575215967,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0005718247139994708,
      "loss": 0.7806,
      "num_input_tokens_seen": 71780584,
      "step": 123625
    },
    {
      "epoch": 18.413762287756924,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.000571291649353473,
      "loss": 0.7986,
      "num_input_tokens_seen": 71783560,
      "step": 123630
    },
    {
      "epoch": 18.414507000297885,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0005707588284658421,
      "loss": 0.8182,
      "num_input_tokens_seen": 71786632,
      "step": 123635
    },
    {
      "epoch": 18.415251712838845,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0005702262513455741,
      "loss": 0.7822,
      "num_input_tokens_seen": 71789736,
      "step": 123640
    },
    {
      "epoch": 18.415996425379802,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.0005696939180016685,
      "loss": 0.7908,
      "num_input_tokens_seen": 71792584,
      "step": 123645
    },
    {
      "epoch": 18.416741137920763,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.0005691618284431165,
      "loss": 0.791,
      "num_input_tokens_seen": 71795176,
      "step": 123650
    },
    {
      "epoch": 18.417485850461723,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0005686299826789092,
      "loss": 0.7842,
      "num_input_tokens_seen": 71798344,
      "step": 123655
    },
    {
      "epoch": 18.41823056300268,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0005680983807180329,
      "loss": 0.7999,
      "num_input_tokens_seen": 71801064,
      "step": 123660
    },
    {
      "epoch": 18.41897527554364,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0005675670225694618,
      "loss": 0.7949,
      "num_input_tokens_seen": 71804040,
      "step": 123665
    },
    {
      "epoch": 18.419719988084598,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0005670359082421822,
      "loss": 0.7918,
      "num_input_tokens_seen": 71806888,
      "step": 123670
    },
    {
      "epoch": 18.42046470062556,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0005665050377451602,
      "loss": 0.7941,
      "num_input_tokens_seen": 71809800,
      "step": 123675
    },
    {
      "epoch": 18.42120941316652,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0005659744110873638,
      "loss": 0.7968,
      "num_input_tokens_seen": 71812808,
      "step": 123680
    },
    {
      "epoch": 18.421954125707476,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0005654440282777623,
      "loss": 0.7997,
      "num_input_tokens_seen": 71815688,
      "step": 123685
    },
    {
      "epoch": 18.422698838248436,
      "grad_norm": 0.05859375,
      "learning_rate": 0.000564913889325312,
      "loss": 0.8037,
      "num_input_tokens_seen": 71818440,
      "step": 123690
    },
    {
      "epoch": 18.423443550789397,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0005643839942389706,
      "loss": 0.8174,
      "num_input_tokens_seen": 71821512,
      "step": 123695
    },
    {
      "epoch": 18.424188263330354,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0005638543430276893,
      "loss": 0.8029,
      "num_input_tokens_seen": 71824520,
      "step": 123700
    },
    {
      "epoch": 18.424932975871315,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0005633249357004177,
      "loss": 0.7963,
      "num_input_tokens_seen": 71827464,
      "step": 123705
    },
    {
      "epoch": 18.42567768841227,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0005627957722661003,
      "loss": 0.7934,
      "num_input_tokens_seen": 71830184,
      "step": 123710
    },
    {
      "epoch": 18.426422400953232,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0005622668527336732,
      "loss": 0.7775,
      "num_input_tokens_seen": 71832904,
      "step": 123715
    },
    {
      "epoch": 18.427167113494193,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0005617381771120743,
      "loss": 0.7973,
      "num_input_tokens_seen": 71835912,
      "step": 123720
    },
    {
      "epoch": 18.42791182603515,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0005612097454102332,
      "loss": 0.7903,
      "num_input_tokens_seen": 71838792,
      "step": 123725
    },
    {
      "epoch": 18.42865653857611,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0005606815576370794,
      "loss": 0.7857,
      "num_input_tokens_seen": 71841320,
      "step": 123730
    },
    {
      "epoch": 18.42940125111707,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0005601536138015339,
      "loss": 0.7882,
      "num_input_tokens_seen": 71844200,
      "step": 123735
    },
    {
      "epoch": 18.430145963658028,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0005596259139125182,
      "loss": 0.7941,
      "num_input_tokens_seen": 71846984,
      "step": 123740
    },
    {
      "epoch": 18.430890676198988,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0005590984579789449,
      "loss": 0.8008,
      "num_input_tokens_seen": 71849608,
      "step": 123745
    },
    {
      "epoch": 18.431635388739945,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0005585712460097286,
      "loss": 0.8166,
      "num_input_tokens_seen": 71852552,
      "step": 123750
    },
    {
      "epoch": 18.432380101280906,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0005580442780137723,
      "loss": 0.7915,
      "num_input_tokens_seen": 71855880,
      "step": 123755
    },
    {
      "epoch": 18.433124813821866,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0005575175539999788,
      "loss": 0.7958,
      "num_input_tokens_seen": 71858600,
      "step": 123760
    },
    {
      "epoch": 18.433869526362823,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0005569910739772493,
      "loss": 0.7941,
      "num_input_tokens_seen": 71861608,
      "step": 123765
    },
    {
      "epoch": 18.434614238903784,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0005564648379544751,
      "loss": 0.7881,
      "num_input_tokens_seen": 71864392,
      "step": 123770
    },
    {
      "epoch": 18.43535895144474,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0005559388459405473,
      "loss": 0.796,
      "num_input_tokens_seen": 71867208,
      "step": 123775
    },
    {
      "epoch": 18.4361036639857,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0005554130979443538,
      "loss": 0.7805,
      "num_input_tokens_seen": 71870088,
      "step": 123780
    },
    {
      "epoch": 18.43684837652666,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.000554887593974776,
      "loss": 0.7877,
      "num_input_tokens_seen": 71872904,
      "step": 123785
    },
    {
      "epoch": 18.43759308906762,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0005543623340406883,
      "loss": 0.7788,
      "num_input_tokens_seen": 71875656,
      "step": 123790
    },
    {
      "epoch": 18.43833780160858,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0005538373181509687,
      "loss": 0.7816,
      "num_input_tokens_seen": 71878312,
      "step": 123795
    },
    {
      "epoch": 18.43908251414954,
      "grad_norm": 0.05859375,
      "learning_rate": 0.000553312546314485,
      "loss": 0.7905,
      "num_input_tokens_seen": 71881512,
      "step": 123800
    },
    {
      "epoch": 18.439827226690497,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0005527880185401019,
      "loss": 0.8103,
      "num_input_tokens_seen": 71884424,
      "step": 123805
    },
    {
      "epoch": 18.440571939231457,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0005522637348366838,
      "loss": 0.7983,
      "num_input_tokens_seen": 71887240,
      "step": 123810
    },
    {
      "epoch": 18.441316651772414,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0005517396952130837,
      "loss": 0.8104,
      "num_input_tokens_seen": 71890088,
      "step": 123815
    },
    {
      "epoch": 18.442061364313375,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.000551215899678158,
      "loss": 0.7892,
      "num_input_tokens_seen": 71892968,
      "step": 123820
    },
    {
      "epoch": 18.442806076854335,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0005506923482407527,
      "loss": 0.7936,
      "num_input_tokens_seen": 71895752,
      "step": 123825
    },
    {
      "epoch": 18.443550789395292,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0005501690409097177,
      "loss": 0.7965,
      "num_input_tokens_seen": 71898888,
      "step": 123830
    },
    {
      "epoch": 18.444295501936253,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0005496459776938889,
      "loss": 0.7878,
      "num_input_tokens_seen": 71901704,
      "step": 123835
    },
    {
      "epoch": 18.445040214477213,
      "grad_norm": 0.046875,
      "learning_rate": 0.0005491231586021044,
      "loss": 0.7712,
      "num_input_tokens_seen": 71904520,
      "step": 123840
    },
    {
      "epoch": 18.44578492701817,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0005486005836431972,
      "loss": 0.8122,
      "num_input_tokens_seen": 71907336,
      "step": 123845
    },
    {
      "epoch": 18.44652963955913,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0005480782528259969,
      "loss": 0.8023,
      "num_input_tokens_seen": 71910536,
      "step": 123850
    },
    {
      "epoch": 18.447274352100088,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0005475561661593247,
      "loss": 0.7994,
      "num_input_tokens_seen": 71913320,
      "step": 123855
    },
    {
      "epoch": 18.44801906464105,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.000547034323652002,
      "loss": 0.7808,
      "num_input_tokens_seen": 71916552,
      "step": 123860
    },
    {
      "epoch": 18.44876377718201,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0005465127253128465,
      "loss": 0.7917,
      "num_input_tokens_seen": 71919400,
      "step": 123865
    },
    {
      "epoch": 18.449508489722966,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.0005459913711506664,
      "loss": 0.7676,
      "num_input_tokens_seen": 71922216,
      "step": 123870
    },
    {
      "epoch": 18.450253202263927,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0005454702611742729,
      "loss": 0.804,
      "num_input_tokens_seen": 71925064,
      "step": 123875
    },
    {
      "epoch": 18.450997914804887,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0005449493953924689,
      "loss": 0.8013,
      "num_input_tokens_seen": 71928104,
      "step": 123880
    },
    {
      "epoch": 18.451742627345844,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0005444287738140523,
      "loss": 0.8001,
      "num_input_tokens_seen": 71930920,
      "step": 123885
    },
    {
      "epoch": 18.452487339886805,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0005439083964478214,
      "loss": 0.8186,
      "num_input_tokens_seen": 71933768,
      "step": 123890
    },
    {
      "epoch": 18.45323205242776,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0005433882633025638,
      "loss": 0.7899,
      "num_input_tokens_seen": 71936872,
      "step": 123895
    },
    {
      "epoch": 18.453976764968722,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0005428683743870694,
      "loss": 0.7934,
      "num_input_tokens_seen": 71939880,
      "step": 123900
    },
    {
      "epoch": 18.454721477509683,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0005423487297101193,
      "loss": 0.7908,
      "num_input_tokens_seen": 71942664,
      "step": 123905
    },
    {
      "epoch": 18.45546619005064,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0005418293292804931,
      "loss": 0.8246,
      "num_input_tokens_seen": 71945544,
      "step": 123910
    },
    {
      "epoch": 18.4562109025916,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0005413101731069675,
      "loss": 0.8131,
      "num_input_tokens_seen": 71948200,
      "step": 123915
    },
    {
      "epoch": 18.456955615132557,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0005407912611983118,
      "loss": 0.8037,
      "num_input_tokens_seen": 71951240,
      "step": 123920
    },
    {
      "epoch": 18.457700327673518,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0005402725935632907,
      "loss": 0.8035,
      "num_input_tokens_seen": 71953832,
      "step": 123925
    },
    {
      "epoch": 18.458445040214478,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0005397541702106673,
      "loss": 0.7833,
      "num_input_tokens_seen": 71956456,
      "step": 123930
    },
    {
      "epoch": 18.459189752755435,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0005392359911492011,
      "loss": 0.803,
      "num_input_tokens_seen": 71959496,
      "step": 123935
    },
    {
      "epoch": 18.459934465296396,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0005387180563876453,
      "loss": 0.8046,
      "num_input_tokens_seen": 71962472,
      "step": 123940
    },
    {
      "epoch": 18.460679177837356,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0005382003659347511,
      "loss": 0.7862,
      "num_input_tokens_seen": 71965320,
      "step": 123945
    },
    {
      "epoch": 18.461423890378313,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.0005376829197992617,
      "loss": 0.8111,
      "num_input_tokens_seen": 71968040,
      "step": 123950
    },
    {
      "epoch": 18.462168602919274,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0005371657179899214,
      "loss": 0.782,
      "num_input_tokens_seen": 71971176,
      "step": 123955
    },
    {
      "epoch": 18.46291331546023,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0005366487605154684,
      "loss": 0.8068,
      "num_input_tokens_seen": 71973800,
      "step": 123960
    },
    {
      "epoch": 18.46365802800119,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0005361320473846326,
      "loss": 0.8091,
      "num_input_tokens_seen": 71976616,
      "step": 123965
    },
    {
      "epoch": 18.464402740542152,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.0005356155786061467,
      "loss": 0.7821,
      "num_input_tokens_seen": 71979656,
      "step": 123970
    },
    {
      "epoch": 18.46514745308311,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0005350993541887356,
      "loss": 0.7995,
      "num_input_tokens_seen": 71982600,
      "step": 123975
    },
    {
      "epoch": 18.46589216562407,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0005345833741411155,
      "loss": 0.8089,
      "num_input_tokens_seen": 71985736,
      "step": 123980
    },
    {
      "epoch": 18.46663687816503,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0005340676384720111,
      "loss": 0.7972,
      "num_input_tokens_seen": 71988712,
      "step": 123985
    },
    {
      "epoch": 18.467381590705987,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0005335521471901305,
      "loss": 0.8004,
      "num_input_tokens_seen": 71991720,
      "step": 123990
    },
    {
      "epoch": 18.468126303246947,
      "grad_norm": 0.048828125,
      "learning_rate": 0.0005330369003041835,
      "loss": 0.7798,
      "num_input_tokens_seen": 71994600,
      "step": 123995
    },
    {
      "epoch": 18.468871015787904,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0005325218978228729,
      "loss": 0.792,
      "num_input_tokens_seen": 71997480,
      "step": 124000
    },
    {
      "epoch": 18.469615728328865,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0005320071397549036,
      "loss": 0.7975,
      "num_input_tokens_seen": 72000520,
      "step": 124005
    },
    {
      "epoch": 18.470360440869825,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0005314926261089653,
      "loss": 0.805,
      "num_input_tokens_seen": 72003208,
      "step": 124010
    },
    {
      "epoch": 18.471105153410782,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0005309783568937576,
      "loss": 0.7977,
      "num_input_tokens_seen": 72006088,
      "step": 124015
    },
    {
      "epoch": 18.471849865951743,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.000530464332117962,
      "loss": 0.8024,
      "num_input_tokens_seen": 72009128,
      "step": 124020
    },
    {
      "epoch": 18.472594578492703,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0005299505517902664,
      "loss": 0.7852,
      "num_input_tokens_seen": 72011816,
      "step": 124025
    },
    {
      "epoch": 18.47333929103366,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.000529437015919349,
      "loss": 0.7983,
      "num_input_tokens_seen": 72014536,
      "step": 124030
    },
    {
      "epoch": 18.47408400357462,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0005289237245138877,
      "loss": 0.7949,
      "num_input_tokens_seen": 72017192,
      "step": 124035
    },
    {
      "epoch": 18.474828716115578,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0005284106775825508,
      "loss": 0.8219,
      "num_input_tokens_seen": 72019976,
      "step": 124040
    },
    {
      "epoch": 18.47557342865654,
      "grad_norm": 0.05859375,
      "learning_rate": 0.0005278978751340097,
      "loss": 0.8007,
      "num_input_tokens_seen": 72022600,
      "step": 124045
    },
    {
      "epoch": 18.4763181411975,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0005273853171769238,
      "loss": 0.8033,
      "num_input_tokens_seen": 72025448,
      "step": 124050
    },
    {
      "epoch": 18.477062853738456,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0005268730037199515,
      "loss": 0.8003,
      "num_input_tokens_seen": 72028424,
      "step": 124055
    },
    {
      "epoch": 18.477807566279417,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0005263609347717523,
      "loss": 0.7895,
      "num_input_tokens_seen": 72031400,
      "step": 124060
    },
    {
      "epoch": 18.478552278820374,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0005258491103409729,
      "loss": 0.8165,
      "num_input_tokens_seen": 72034216,
      "step": 124065
    },
    {
      "epoch": 18.479296991361334,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.0005253375304362645,
      "loss": 0.8089,
      "num_input_tokens_seen": 72037224,
      "step": 124070
    },
    {
      "epoch": 18.480041703902295,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0005248261950662635,
      "loss": 0.7962,
      "num_input_tokens_seen": 72040136,
      "step": 124075
    },
    {
      "epoch": 18.48078641644325,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0005243151042396149,
      "loss": 0.7838,
      "num_input_tokens_seen": 72042920,
      "step": 124080
    },
    {
      "epoch": 18.481531128984212,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0005238042579649499,
      "loss": 0.7828,
      "num_input_tokens_seen": 72045864,
      "step": 124085
    },
    {
      "epoch": 18.482275841525173,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.0005232936562508983,
      "loss": 0.8077,
      "num_input_tokens_seen": 72048648,
      "step": 124090
    },
    {
      "epoch": 18.48302055406613,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0005227832991060882,
      "loss": 0.8009,
      "num_input_tokens_seen": 72051624,
      "step": 124095
    },
    {
      "epoch": 18.48376526660709,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0005222731865391377,
      "loss": 0.7993,
      "num_input_tokens_seen": 72054472,
      "step": 124100
    },
    {
      "epoch": 18.484509979148047,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0005217633185586701,
      "loss": 0.7855,
      "num_input_tokens_seen": 72057320,
      "step": 124105
    },
    {
      "epoch": 18.485254691689008,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0005212536951732949,
      "loss": 0.8009,
      "num_input_tokens_seen": 72060200,
      "step": 124110
    },
    {
      "epoch": 18.48599940422997,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0005207443163916237,
      "loss": 0.7897,
      "num_input_tokens_seen": 72063048,
      "step": 124115
    },
    {
      "epoch": 18.486744116770925,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0005202351822222611,
      "loss": 0.7997,
      "num_input_tokens_seen": 72065960,
      "step": 124120
    },
    {
      "epoch": 18.487488829311886,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0005197262926738072,
      "loss": 0.7896,
      "num_input_tokens_seen": 72068680,
      "step": 124125
    },
    {
      "epoch": 18.488233541852846,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0005192176477548616,
      "loss": 0.8036,
      "num_input_tokens_seen": 72071912,
      "step": 124130
    },
    {
      "epoch": 18.488978254393803,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0005187092474740157,
      "loss": 0.7969,
      "num_input_tokens_seen": 72074696,
      "step": 124135
    },
    {
      "epoch": 18.489722966934764,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0005182010918398594,
      "loss": 0.7762,
      "num_input_tokens_seen": 72077704,
      "step": 124140
    },
    {
      "epoch": 18.49046767947572,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0005176931808609758,
      "loss": 0.7935,
      "num_input_tokens_seen": 72080648,
      "step": 124145
    },
    {
      "epoch": 18.49121239201668,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.000517185514545948,
      "loss": 0.7826,
      "num_input_tokens_seen": 72083560,
      "step": 124150
    },
    {
      "epoch": 18.491957104557642,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.0005166780929033493,
      "loss": 0.8007,
      "num_input_tokens_seen": 72086472,
      "step": 124155
    },
    {
      "epoch": 18.4927018170986,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.000516170915941756,
      "loss": 0.8005,
      "num_input_tokens_seen": 72089000,
      "step": 124160
    },
    {
      "epoch": 18.49344652963956,
      "grad_norm": 0.0625,
      "learning_rate": 0.0005156639836697363,
      "loss": 0.7986,
      "num_input_tokens_seen": 72092072,
      "step": 124165
    },
    {
      "epoch": 18.49419124218052,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0005151572960958484,
      "loss": 0.8088,
      "num_input_tokens_seen": 72094696,
      "step": 124170
    },
    {
      "epoch": 18.494935954721477,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0005146508532286587,
      "loss": 0.7941,
      "num_input_tokens_seen": 72097576,
      "step": 124175
    },
    {
      "epoch": 18.495680667262437,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0005141446550767203,
      "loss": 0.7934,
      "num_input_tokens_seen": 72100616,
      "step": 124180
    },
    {
      "epoch": 18.496425379803394,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0005136387016485849,
      "loss": 0.8033,
      "num_input_tokens_seen": 72103656,
      "step": 124185
    },
    {
      "epoch": 18.497170092344355,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0005131329929527989,
      "loss": 0.7881,
      "num_input_tokens_seen": 72106856,
      "step": 124190
    },
    {
      "epoch": 18.497914804885315,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0005126275289979105,
      "loss": 0.7981,
      "num_input_tokens_seen": 72109672,
      "step": 124195
    },
    {
      "epoch": 18.498659517426272,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0005121223097924543,
      "loss": 0.8018,
      "num_input_tokens_seen": 72112392,
      "step": 124200
    },
    {
      "epoch": 18.499404229967233,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0005116173353449637,
      "loss": 0.8199,
      "num_input_tokens_seen": 72115112,
      "step": 124205
    },
    {
      "epoch": 18.500148942508194,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.000511112605663977,
      "loss": 0.7939,
      "num_input_tokens_seen": 72118088,
      "step": 124210
    },
    {
      "epoch": 18.50089365504915,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0005106081207580138,
      "loss": 0.7991,
      "num_input_tokens_seen": 72121704,
      "step": 124215
    },
    {
      "epoch": 18.50163836759011,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0005101038806356022,
      "loss": 0.8297,
      "num_input_tokens_seen": 72124744,
      "step": 124220
    },
    {
      "epoch": 18.502383080131068,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0005095998853052574,
      "loss": 0.8205,
      "num_input_tokens_seen": 72127784,
      "step": 124225
    },
    {
      "epoch": 18.50312779267203,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0005090961347754958,
      "loss": 0.793,
      "num_input_tokens_seen": 72130568,
      "step": 124230
    },
    {
      "epoch": 18.50387250521299,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.000508592629054827,
      "loss": 0.7963,
      "num_input_tokens_seen": 72133448,
      "step": 124235
    },
    {
      "epoch": 18.504617217753946,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0005080893681517579,
      "loss": 0.7866,
      "num_input_tokens_seen": 72136488,
      "step": 124240
    },
    {
      "epoch": 18.505361930294907,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0005075863520747898,
      "loss": 0.7711,
      "num_input_tokens_seen": 72139208,
      "step": 124245
    },
    {
      "epoch": 18.506106642835867,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0005070835808324209,
      "loss": 0.7863,
      "num_input_tokens_seen": 72141960,
      "step": 124250
    },
    {
      "epoch": 18.506851355376824,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0005065810544331461,
      "loss": 0.8207,
      "num_input_tokens_seen": 72144648,
      "step": 124255
    },
    {
      "epoch": 18.507596067917785,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0005060787728854521,
      "loss": 0.7997,
      "num_input_tokens_seen": 72147816,
      "step": 124260
    },
    {
      "epoch": 18.50834078045874,
      "grad_norm": 0.064453125,
      "learning_rate": 0.000505576736197827,
      "loss": 0.8059,
      "num_input_tokens_seen": 72150600,
      "step": 124265
    },
    {
      "epoch": 18.509085492999702,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0005050749443787489,
      "loss": 0.8117,
      "num_input_tokens_seen": 72153352,
      "step": 124270
    },
    {
      "epoch": 18.509830205540663,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0005045733974367011,
      "loss": 0.8207,
      "num_input_tokens_seen": 72156168,
      "step": 124275
    },
    {
      "epoch": 18.51057491808162,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.0005040720953801536,
      "loss": 0.7942,
      "num_input_tokens_seen": 72159016,
      "step": 124280
    },
    {
      "epoch": 18.51131963062258,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0005035710382175712,
      "loss": 0.7925,
      "num_input_tokens_seen": 72161768,
      "step": 124285
    },
    {
      "epoch": 18.512064343163537,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0005030702259574254,
      "loss": 0.8048,
      "num_input_tokens_seen": 72164680,
      "step": 124290
    },
    {
      "epoch": 18.512809055704498,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0005025696586081729,
      "loss": 0.7747,
      "num_input_tokens_seen": 72167272,
      "step": 124295
    },
    {
      "epoch": 18.51355376824546,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0005020693361782735,
      "loss": 0.7906,
      "num_input_tokens_seen": 72170216,
      "step": 124300
    },
    {
      "epoch": 18.514298480786415,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.0005015692586761772,
      "loss": 0.8037,
      "num_input_tokens_seen": 72173416,
      "step": 124305
    },
    {
      "epoch": 18.515043193327376,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0005010694261103304,
      "loss": 0.7943,
      "num_input_tokens_seen": 72176264,
      "step": 124310
    },
    {
      "epoch": 18.515787905868336,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0005005698384891816,
      "loss": 0.8021,
      "num_input_tokens_seen": 72179272,
      "step": 124315
    },
    {
      "epoch": 18.516532618409293,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0005000704958211688,
      "loss": 0.7908,
      "num_input_tokens_seen": 72182248,
      "step": 124320
    },
    {
      "epoch": 18.517277330950254,
      "grad_norm": 0.0546875,
      "learning_rate": 0.000499571398114727,
      "loss": 0.7874,
      "num_input_tokens_seen": 72185320,
      "step": 124325
    },
    {
      "epoch": 18.51802204349121,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0004990725453782895,
      "loss": 0.8019,
      "num_input_tokens_seen": 72188456,
      "step": 124330
    },
    {
      "epoch": 18.51876675603217,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.000498573937620283,
      "loss": 0.8018,
      "num_input_tokens_seen": 72190984,
      "step": 124335
    },
    {
      "epoch": 18.519511468573132,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0004980755748491306,
      "loss": 0.8204,
      "num_input_tokens_seen": 72194024,
      "step": 124340
    },
    {
      "epoch": 18.52025618111409,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0004975774570732522,
      "loss": 0.7865,
      "num_input_tokens_seen": 72196872,
      "step": 124345
    },
    {
      "epoch": 18.52100089365505,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0004970795843010628,
      "loss": 0.8095,
      "num_input_tokens_seen": 72199784,
      "step": 124350
    },
    {
      "epoch": 18.52174560619601,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0004965819565409757,
      "loss": 0.7989,
      "num_input_tokens_seen": 72202536,
      "step": 124355
    },
    {
      "epoch": 18.522490318736967,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0004960845738013958,
      "loss": 0.7937,
      "num_input_tokens_seen": 72205192,
      "step": 124360
    },
    {
      "epoch": 18.523235031277927,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.000495587436090723,
      "loss": 0.7934,
      "num_input_tokens_seen": 72208040,
      "step": 124365
    },
    {
      "epoch": 18.523979743818884,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.0004950905434173624,
      "loss": 0.8013,
      "num_input_tokens_seen": 72210856,
      "step": 124370
    },
    {
      "epoch": 18.524724456359845,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0004945938957897039,
      "loss": 0.7993,
      "num_input_tokens_seen": 72213768,
      "step": 124375
    },
    {
      "epoch": 18.525469168900806,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.000494097493216139,
      "loss": 0.8053,
      "num_input_tokens_seen": 72216808,
      "step": 124380
    },
    {
      "epoch": 18.526213881441763,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0004936013357050512,
      "loss": 0.8138,
      "num_input_tokens_seen": 72219368,
      "step": 124385
    },
    {
      "epoch": 18.526958593982723,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0004931054232648285,
      "loss": 0.8124,
      "num_input_tokens_seen": 72222024,
      "step": 124390
    },
    {
      "epoch": 18.527703306523684,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.0004926097559038427,
      "loss": 0.7863,
      "num_input_tokens_seen": 72224968,
      "step": 124395
    },
    {
      "epoch": 18.52844801906464,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0004921143336304723,
      "loss": 0.8048,
      "num_input_tokens_seen": 72228008,
      "step": 124400
    },
    {
      "epoch": 18.5291927316056,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0004916191564530853,
      "loss": 0.8104,
      "num_input_tokens_seen": 72231112,
      "step": 124405
    },
    {
      "epoch": 18.529937444146558,
      "grad_norm": 0.037109375,
      "learning_rate": 0.000491124224380045,
      "loss": 0.7991,
      "num_input_tokens_seen": 72234056,
      "step": 124410
    },
    {
      "epoch": 18.53068215668752,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0004906295374197167,
      "loss": 0.8033,
      "num_input_tokens_seen": 72236968,
      "step": 124415
    },
    {
      "epoch": 18.53142686922848,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0004901350955804534,
      "loss": 0.7933,
      "num_input_tokens_seen": 72239976,
      "step": 124420
    },
    {
      "epoch": 18.532171581769436,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.0004896408988706135,
      "loss": 0.8174,
      "num_input_tokens_seen": 72242760,
      "step": 124425
    },
    {
      "epoch": 18.532916294310397,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0004891469472985405,
      "loss": 0.8314,
      "num_input_tokens_seen": 72245608,
      "step": 124430
    },
    {
      "epoch": 18.533661006851354,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0004886532408725824,
      "loss": 0.8026,
      "num_input_tokens_seen": 72248456,
      "step": 124435
    },
    {
      "epoch": 18.534405719392314,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0004881597796010811,
      "loss": 0.7945,
      "num_input_tokens_seen": 72251368,
      "step": 124440
    },
    {
      "epoch": 18.535150431933275,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0004876665634923699,
      "loss": 0.8063,
      "num_input_tokens_seen": 72254152,
      "step": 124445
    },
    {
      "epoch": 18.53589514447423,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0004871735925547821,
      "loss": 0.7884,
      "num_input_tokens_seen": 72256840,
      "step": 124450
    },
    {
      "epoch": 18.536639857015192,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.000486680866796646,
      "loss": 0.8181,
      "num_input_tokens_seen": 72259656,
      "step": 124455
    },
    {
      "epoch": 18.537384569556153,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.00048618838622628666,
      "loss": 0.7981,
      "num_input_tokens_seen": 72262600,
      "step": 124460
    },
    {
      "epoch": 18.53812928209711,
      "grad_norm": 0.06640625,
      "learning_rate": 0.00048569615085202074,
      "loss": 0.7816,
      "num_input_tokens_seen": 72265992,
      "step": 124465
    },
    {
      "epoch": 18.53887399463807,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00048520416068217,
      "loss": 0.7769,
      "num_input_tokens_seen": 72268776,
      "step": 124470
    },
    {
      "epoch": 18.539618707179027,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.00048471241572504097,
      "loss": 0.7914,
      "num_input_tokens_seen": 72271688,
      "step": 124475
    },
    {
      "epoch": 18.540363419719988,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.00048422091598894376,
      "loss": 0.8013,
      "num_input_tokens_seen": 72274600,
      "step": 124480
    },
    {
      "epoch": 18.54110813226095,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.00048372966148218176,
      "loss": 0.805,
      "num_input_tokens_seen": 72277288,
      "step": 124485
    },
    {
      "epoch": 18.541852844801905,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0004832386522130516,
      "loss": 0.804,
      "num_input_tokens_seen": 72280072,
      "step": 124490
    },
    {
      "epoch": 18.542597557342866,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.0004827478881898517,
      "loss": 0.8134,
      "num_input_tokens_seen": 72282920,
      "step": 124495
    },
    {
      "epoch": 18.543342269883826,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.00048225736942087035,
      "loss": 0.8075,
      "num_input_tokens_seen": 72285896,
      "step": 124500
    },
    {
      "epoch": 18.544086982424783,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.000481767095914396,
      "loss": 0.7862,
      "num_input_tokens_seen": 72288488,
      "step": 124505
    },
    {
      "epoch": 18.544831694965744,
      "grad_norm": 0.033203125,
      "learning_rate": 0.00048127706767871203,
      "loss": 0.7921,
      "num_input_tokens_seen": 72291272,
      "step": 124510
    },
    {
      "epoch": 18.5455764075067,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00048078728472209507,
      "loss": 0.8028,
      "num_input_tokens_seen": 72293928,
      "step": 124515
    },
    {
      "epoch": 18.54632112004766,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.00048029774705282024,
      "loss": 0.7961,
      "num_input_tokens_seen": 72296808,
      "step": 124520
    },
    {
      "epoch": 18.547065832588622,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.0004798084546791575,
      "loss": 0.7965,
      "num_input_tokens_seen": 72299560,
      "step": 124525
    },
    {
      "epoch": 18.54781054512958,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.00047931940760937374,
      "loss": 0.7973,
      "num_input_tokens_seen": 72302440,
      "step": 124530
    },
    {
      "epoch": 18.54855525767054,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.00047883060585172883,
      "loss": 0.8063,
      "num_input_tokens_seen": 72305256,
      "step": 124535
    },
    {
      "epoch": 18.5492999702115,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.00047834204941448464,
      "loss": 0.7975,
      "num_input_tokens_seen": 72307944,
      "step": 124540
    },
    {
      "epoch": 18.550044682752457,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.0004778537383058895,
      "loss": 0.8011,
      "num_input_tokens_seen": 72310760,
      "step": 124545
    },
    {
      "epoch": 18.550789395293418,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.00047736567253419846,
      "loss": 0.8059,
      "num_input_tokens_seen": 72313704,
      "step": 124550
    },
    {
      "epoch": 18.551534107834375,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.0004768778521076533,
      "loss": 0.7943,
      "num_input_tokens_seen": 72316808,
      "step": 124555
    },
    {
      "epoch": 18.552278820375335,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00047639027703449575,
      "loss": 0.762,
      "num_input_tokens_seen": 72319592,
      "step": 124560
    },
    {
      "epoch": 18.553023532916296,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0004759029473229642,
      "loss": 0.7831,
      "num_input_tokens_seen": 72322536,
      "step": 124565
    },
    {
      "epoch": 18.553768245457253,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00047541586298129046,
      "loss": 0.812,
      "num_input_tokens_seen": 72325448,
      "step": 124570
    },
    {
      "epoch": 18.554512957998213,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.00047492902401770283,
      "loss": 0.7835,
      "num_input_tokens_seen": 72328616,
      "step": 124575
    },
    {
      "epoch": 18.55525767053917,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0004744424304404282,
      "loss": 0.8051,
      "num_input_tokens_seen": 72331528,
      "step": 124580
    },
    {
      "epoch": 18.55600238308013,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0004739560822576849,
      "loss": 0.7972,
      "num_input_tokens_seen": 72334248,
      "step": 124585
    },
    {
      "epoch": 18.55674709562109,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.00047346997947768973,
      "loss": 0.8088,
      "num_input_tokens_seen": 72336936,
      "step": 124590
    },
    {
      "epoch": 18.557491808162048,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0004729841221086561,
      "loss": 0.8083,
      "num_input_tokens_seen": 72339720,
      "step": 124595
    },
    {
      "epoch": 18.55823652070301,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0004724985101587908,
      "loss": 0.7807,
      "num_input_tokens_seen": 72342504,
      "step": 124600
    },
    {
      "epoch": 18.55898123324397,
      "grad_norm": 0.0625,
      "learning_rate": 0.00047201314363629564,
      "loss": 0.7911,
      "num_input_tokens_seen": 72345128,
      "step": 124605
    },
    {
      "epoch": 18.559725945784926,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00047152802254937565,
      "loss": 0.8008,
      "num_input_tokens_seen": 72348392,
      "step": 124610
    },
    {
      "epoch": 18.560470658325887,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00047104314690622095,
      "loss": 0.7732,
      "num_input_tokens_seen": 72351112,
      "step": 124615
    },
    {
      "epoch": 18.561215370866844,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.00047055851671502834,
      "loss": 0.8259,
      "num_input_tokens_seen": 72353960,
      "step": 124620
    },
    {
      "epoch": 18.561960083407804,
      "grad_norm": 0.0546875,
      "learning_rate": 0.00047007413198398127,
      "loss": 0.8084,
      "num_input_tokens_seen": 72356680,
      "step": 124625
    },
    {
      "epoch": 18.562704795948765,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.00046958999272126486,
      "loss": 0.8007,
      "num_input_tokens_seen": 72359400,
      "step": 124630
    },
    {
      "epoch": 18.56344950848972,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0004691060989350576,
      "loss": 0.7886,
      "num_input_tokens_seen": 72362248,
      "step": 124635
    },
    {
      "epoch": 18.564194221030682,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0004686224506335329,
      "loss": 0.79,
      "num_input_tokens_seen": 72365000,
      "step": 124640
    },
    {
      "epoch": 18.564938933571643,
      "grad_norm": 0.05859375,
      "learning_rate": 0.00046813904782486425,
      "loss": 0.7855,
      "num_input_tokens_seen": 72367752,
      "step": 124645
    },
    {
      "epoch": 18.5656836461126,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.00046765589051721675,
      "loss": 0.7989,
      "num_input_tokens_seen": 72370824,
      "step": 124650
    },
    {
      "epoch": 18.56642835865356,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.00046717297871875396,
      "loss": 0.7962,
      "num_input_tokens_seen": 72373608,
      "step": 124655
    },
    {
      "epoch": 18.567173071194517,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0004666903124376309,
      "loss": 0.7895,
      "num_input_tokens_seen": 72376552,
      "step": 124660
    },
    {
      "epoch": 18.567917783735478,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0004662078916820045,
      "loss": 0.7972,
      "num_input_tokens_seen": 72379624,
      "step": 124665
    },
    {
      "epoch": 18.56866249627644,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0004657257164600248,
      "loss": 0.8051,
      "num_input_tokens_seen": 72382472,
      "step": 124670
    },
    {
      "epoch": 18.569407208817395,
      "grad_norm": 0.087890625,
      "learning_rate": 0.000465243786779837,
      "loss": 0.8055,
      "num_input_tokens_seen": 72385704,
      "step": 124675
    },
    {
      "epoch": 18.570151921358356,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0004647621026495813,
      "loss": 0.7999,
      "num_input_tokens_seen": 72388200,
      "step": 124680
    },
    {
      "epoch": 18.570896633899316,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00046428066407739943,
      "loss": 0.8045,
      "num_input_tokens_seen": 72390920,
      "step": 124685
    },
    {
      "epoch": 18.571641346440273,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.0004637994710714233,
      "loss": 0.79,
      "num_input_tokens_seen": 72393768,
      "step": 124690
    },
    {
      "epoch": 18.572386058981234,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00046331852363977797,
      "loss": 0.8179,
      "num_input_tokens_seen": 72396616,
      "step": 124695
    },
    {
      "epoch": 18.57313077152219,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0004628378217905937,
      "loss": 0.7819,
      "num_input_tokens_seen": 72399784,
      "step": 124700
    },
    {
      "epoch": 18.57387548406315,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.000462357365531989,
      "loss": 0.7944,
      "num_input_tokens_seen": 72402920,
      "step": 124705
    },
    {
      "epoch": 18.574620196604112,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.00046187715487208236,
      "loss": 0.8004,
      "num_input_tokens_seen": 72405704,
      "step": 124710
    },
    {
      "epoch": 18.57536490914507,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.00046139718981898226,
      "loss": 0.7995,
      "num_input_tokens_seen": 72408808,
      "step": 124715
    },
    {
      "epoch": 18.57610962168603,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0004609174703808039,
      "loss": 0.8001,
      "num_input_tokens_seen": 72412168,
      "step": 124720
    },
    {
      "epoch": 18.57685433422699,
      "grad_norm": 0.056640625,
      "learning_rate": 0.00046043799656564577,
      "loss": 0.8069,
      "num_input_tokens_seen": 72414760,
      "step": 124725
    },
    {
      "epoch": 18.577599046767947,
      "grad_norm": 0.060546875,
      "learning_rate": 0.00045995876838160975,
      "loss": 0.8315,
      "num_input_tokens_seen": 72417640,
      "step": 124730
    },
    {
      "epoch": 18.578343759308908,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.00045947978583679434,
      "loss": 0.7982,
      "num_input_tokens_seen": 72420456,
      "step": 124735
    },
    {
      "epoch": 18.579088471849865,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0004590010489392882,
      "loss": 0.822,
      "num_input_tokens_seen": 72423464,
      "step": 124740
    },
    {
      "epoch": 18.579833184390825,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0004585225576971813,
      "loss": 0.8093,
      "num_input_tokens_seen": 72426472,
      "step": 124745
    },
    {
      "epoch": 18.580577896931786,
      "grad_norm": 0.056640625,
      "learning_rate": 0.00045804431211855567,
      "loss": 0.7934,
      "num_input_tokens_seen": 72429320,
      "step": 124750
    },
    {
      "epoch": 18.581322609472743,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0004575663122114931,
      "loss": 0.7883,
      "num_input_tokens_seen": 72432360,
      "step": 124755
    },
    {
      "epoch": 18.582067322013703,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.00045708855798406555,
      "loss": 0.8003,
      "num_input_tokens_seen": 72435112,
      "step": 124760
    },
    {
      "epoch": 18.582812034554664,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.00045661104944434813,
      "loss": 0.7968,
      "num_input_tokens_seen": 72438152,
      "step": 124765
    },
    {
      "epoch": 18.58355674709562,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.00045613378660040445,
      "loss": 0.7915,
      "num_input_tokens_seen": 72440968,
      "step": 124770
    },
    {
      "epoch": 18.58430145963658,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.00045565676946029976,
      "loss": 0.8033,
      "num_input_tokens_seen": 72443688,
      "step": 124775
    },
    {
      "epoch": 18.585046172177538,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0004551799980320925,
      "loss": 0.7905,
      "num_input_tokens_seen": 72446600,
      "step": 124780
    },
    {
      "epoch": 18.5857908847185,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00045470347232383297,
      "loss": 0.7906,
      "num_input_tokens_seen": 72449960,
      "step": 124785
    },
    {
      "epoch": 18.58653559725946,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0004542271923435781,
      "loss": 0.7933,
      "num_input_tokens_seen": 72452648,
      "step": 124790
    },
    {
      "epoch": 18.587280309800416,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0004537511580993697,
      "loss": 0.7946,
      "num_input_tokens_seen": 72455624,
      "step": 124795
    },
    {
      "epoch": 18.588025022341377,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.00045327536959925306,
      "loss": 0.7954,
      "num_input_tokens_seen": 72458312,
      "step": 124800
    },
    {
      "epoch": 18.588769734882334,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0004527998268512634,
      "loss": 0.8027,
      "num_input_tokens_seen": 72461000,
      "step": 124805
    },
    {
      "epoch": 18.589514447423294,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.00045232452986343426,
      "loss": 0.7842,
      "num_input_tokens_seen": 72464040,
      "step": 124810
    },
    {
      "epoch": 18.590259159964255,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.00045184947864379766,
      "loss": 0.7982,
      "num_input_tokens_seen": 72466888,
      "step": 124815
    },
    {
      "epoch": 18.591003872505212,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.00045137467320037705,
      "loss": 0.7917,
      "num_input_tokens_seen": 72470088,
      "step": 124820
    },
    {
      "epoch": 18.591748585046172,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00045090011354119604,
      "loss": 0.7981,
      "num_input_tokens_seen": 72472840,
      "step": 124825
    },
    {
      "epoch": 18.592493297587133,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.00045042579967426996,
      "loss": 0.7967,
      "num_input_tokens_seen": 72475848,
      "step": 124830
    },
    {
      "epoch": 18.59323801012809,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00044995173160761404,
      "loss": 0.8255,
      "num_input_tokens_seen": 72478728,
      "step": 124835
    },
    {
      "epoch": 18.59398272266905,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.0004494779093492351,
      "loss": 0.8047,
      "num_input_tokens_seen": 72481800,
      "step": 124840
    },
    {
      "epoch": 18.594727435210007,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0004490043329071369,
      "loss": 0.7979,
      "num_input_tokens_seen": 72484552,
      "step": 124845
    },
    {
      "epoch": 18.595472147750968,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0004485310022893229,
      "loss": 0.7986,
      "num_input_tokens_seen": 72487176,
      "step": 124850
    },
    {
      "epoch": 18.59621686029193,
      "grad_norm": 0.060546875,
      "learning_rate": 0.00044805791750378684,
      "loss": 0.7865,
      "num_input_tokens_seen": 72490120,
      "step": 124855
    },
    {
      "epoch": 18.596961572832885,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0004475850785585239,
      "loss": 0.7879,
      "num_input_tokens_seen": 72493064,
      "step": 124860
    },
    {
      "epoch": 18.597706285373846,
      "grad_norm": 0.041015625,
      "learning_rate": 0.00044711248546151934,
      "loss": 0.809,
      "num_input_tokens_seen": 72496264,
      "step": 124865
    },
    {
      "epoch": 18.598450997914806,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0004466401382207585,
      "loss": 0.8088,
      "num_input_tokens_seen": 72499336,
      "step": 124870
    },
    {
      "epoch": 18.599195710455763,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.00044616803684422167,
      "loss": 0.7811,
      "num_input_tokens_seen": 72502440,
      "step": 124875
    },
    {
      "epoch": 18.599940422996724,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00044569618133988406,
      "loss": 0.802,
      "num_input_tokens_seen": 72505192,
      "step": 124880
    },
    {
      "epoch": 18.60068513553768,
      "grad_norm": 0.0225830078125,
      "learning_rate": 0.0004452245717157177,
      "loss": 0.8044,
      "num_input_tokens_seen": 72508264,
      "step": 124885
    },
    {
      "epoch": 18.60142984807864,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.0004447532079796879,
      "loss": 0.7903,
      "num_input_tokens_seen": 72511176,
      "step": 124890
    },
    {
      "epoch": 18.602174560619602,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.0004442820901397598,
      "loss": 0.7939,
      "num_input_tokens_seen": 72514248,
      "step": 124895
    },
    {
      "epoch": 18.60291927316056,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.00044381121820389225,
      "loss": 0.8042,
      "num_input_tokens_seen": 72517192,
      "step": 124900
    },
    {
      "epoch": 18.60366398570152,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0004433405921800404,
      "loss": 0.8126,
      "num_input_tokens_seen": 72520008,
      "step": 124905
    },
    {
      "epoch": 18.60440869824248,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0004428702120761546,
      "loss": 0.7859,
      "num_input_tokens_seen": 72522760,
      "step": 124910
    },
    {
      "epoch": 18.605153410783437,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.0004424000779001819,
      "loss": 0.7877,
      "num_input_tokens_seen": 72525672,
      "step": 124915
    },
    {
      "epoch": 18.605898123324398,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0004419301896600608,
      "loss": 0.794,
      "num_input_tokens_seen": 72528712,
      "step": 124920
    },
    {
      "epoch": 18.606642835865355,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0004414605473637367,
      "loss": 0.7655,
      "num_input_tokens_seen": 72531432,
      "step": 124925
    },
    {
      "epoch": 18.607387548406315,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.00044099115101913834,
      "loss": 0.7938,
      "num_input_tokens_seen": 72534568,
      "step": 124930
    },
    {
      "epoch": 18.608132260947276,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.00044052200063419755,
      "loss": 0.8105,
      "num_input_tokens_seen": 72538024,
      "step": 124935
    },
    {
      "epoch": 18.608876973488233,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0004400530962168414,
      "loss": 0.8022,
      "num_input_tokens_seen": 72540808,
      "step": 124940
    },
    {
      "epoch": 18.609621686029193,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.00043958443777498686,
      "loss": 0.7991,
      "num_input_tokens_seen": 72543656,
      "step": 124945
    },
    {
      "epoch": 18.61036639857015,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0004391160253165577,
      "loss": 0.8076,
      "num_input_tokens_seen": 72546312,
      "step": 124950
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.00043864785884946243,
      "loss": 0.81,
      "num_input_tokens_seen": 72549032,
      "step": 124955
    },
    {
      "epoch": 18.61185582365207,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00043817993838161316,
      "loss": 0.808,
      "num_input_tokens_seen": 72551848,
      "step": 124960
    },
    {
      "epoch": 18.61260053619303,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.00043771226392091354,
      "loss": 0.7955,
      "num_input_tokens_seen": 72554696,
      "step": 124965
    },
    {
      "epoch": 18.61334524873399,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.00043724483547526394,
      "loss": 0.8032,
      "num_input_tokens_seen": 72557832,
      "step": 124970
    },
    {
      "epoch": 18.61408996127495,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.000436777653052563,
      "loss": 0.7792,
      "num_input_tokens_seen": 72561064,
      "step": 124975
    },
    {
      "epoch": 18.614834673815906,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.00043631071666070276,
      "loss": 0.8141,
      "num_input_tokens_seen": 72563816,
      "step": 124980
    },
    {
      "epoch": 18.615579386356867,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0004358440263075719,
      "loss": 0.7973,
      "num_input_tokens_seen": 72567144,
      "step": 124985
    },
    {
      "epoch": 18.616324098897824,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0004353775820010508,
      "loss": 0.7909,
      "num_input_tokens_seen": 72570024,
      "step": 124990
    },
    {
      "epoch": 18.617068811438784,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.0004349113837490248,
      "loss": 0.8036,
      "num_input_tokens_seen": 72572680,
      "step": 124995
    },
    {
      "epoch": 18.617813523979745,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.000434445431559366,
      "loss": 0.7762,
      "num_input_tokens_seen": 72575432,
      "step": 125000
    },
    {
      "epoch": 18.618558236520702,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0004339797254399497,
      "loss": 0.8049,
      "num_input_tokens_seen": 72578376,
      "step": 125005
    },
    {
      "epoch": 18.619302949061662,
      "grad_norm": 0.064453125,
      "learning_rate": 0.00043351426539864124,
      "loss": 0.7782,
      "num_input_tokens_seen": 72581416,
      "step": 125010
    },
    {
      "epoch": 18.620047661602623,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0004330490514433027,
      "loss": 0.7957,
      "num_input_tokens_seen": 72584424,
      "step": 125015
    },
    {
      "epoch": 18.62079237414358,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.00043258408358179786,
      "loss": 0.8191,
      "num_input_tokens_seen": 72587432,
      "step": 125020
    },
    {
      "epoch": 18.62153708668454,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00043211936182197707,
      "loss": 0.7887,
      "num_input_tokens_seen": 72590184,
      "step": 125025
    },
    {
      "epoch": 18.622281799225497,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00043165488617169566,
      "loss": 0.8088,
      "num_input_tokens_seen": 72593096,
      "step": 125030
    },
    {
      "epoch": 18.623026511766458,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0004311906566387974,
      "loss": 0.8068,
      "num_input_tokens_seen": 72595912,
      "step": 125035
    },
    {
      "epoch": 18.62377122430742,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.000430726673231126,
      "loss": 0.8114,
      "num_input_tokens_seen": 72598760,
      "step": 125040
    },
    {
      "epoch": 18.624515936848375,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0004302629359565202,
      "loss": 0.8021,
      "num_input_tokens_seen": 72601448,
      "step": 125045
    },
    {
      "epoch": 18.625260649389336,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0004297994448228137,
      "loss": 0.7975,
      "num_input_tokens_seen": 72604200,
      "step": 125050
    },
    {
      "epoch": 18.626005361930297,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.000429336199837837,
      "loss": 0.8105,
      "num_input_tokens_seen": 72608168,
      "step": 125055
    },
    {
      "epoch": 18.626750074471254,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.00042887320100941536,
      "loss": 0.809,
      "num_input_tokens_seen": 72610984,
      "step": 125060
    },
    {
      "epoch": 18.627494787012214,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00042841044834537265,
      "loss": 0.7907,
      "num_input_tokens_seen": 72613832,
      "step": 125065
    },
    {
      "epoch": 18.62823949955317,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.00042794794185352424,
      "loss": 0.8039,
      "num_input_tokens_seen": 72617000,
      "step": 125070
    },
    {
      "epoch": 18.62898421209413,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0004274856815416872,
      "loss": 0.812,
      "num_input_tokens_seen": 72619560,
      "step": 125075
    },
    {
      "epoch": 18.629728924635092,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.00042702366741766693,
      "loss": 0.7873,
      "num_input_tokens_seen": 72622536,
      "step": 125080
    },
    {
      "epoch": 18.63047363717605,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.00042656189948927223,
      "loss": 0.8085,
      "num_input_tokens_seen": 72625416,
      "step": 125085
    },
    {
      "epoch": 18.63121834971701,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0004261003777643019,
      "loss": 0.782,
      "num_input_tokens_seen": 72628168,
      "step": 125090
    },
    {
      "epoch": 18.631963062257967,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.00042563910225055124,
      "loss": 0.7841,
      "num_input_tokens_seen": 72631176,
      "step": 125095
    },
    {
      "epoch": 18.632707774798927,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.00042517807295581745,
      "loss": 0.781,
      "num_input_tokens_seen": 72634024,
      "step": 125100
    },
    {
      "epoch": 18.633452487339888,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0004247172898878876,
      "loss": 0.797,
      "num_input_tokens_seen": 72636904,
      "step": 125105
    },
    {
      "epoch": 18.634197199880845,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.00042425675305454377,
      "loss": 0.7889,
      "num_input_tokens_seen": 72639976,
      "step": 125110
    },
    {
      "epoch": 18.634941912421805,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.00042379646246356814,
      "loss": 0.8,
      "num_input_tokens_seen": 72642888,
      "step": 125115
    },
    {
      "epoch": 18.635686624962766,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.0004233364181227378,
      "loss": 0.7978,
      "num_input_tokens_seen": 72645800,
      "step": 125120
    },
    {
      "epoch": 18.636431337503723,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0004228766200398232,
      "loss": 0.7919,
      "num_input_tokens_seen": 72648584,
      "step": 125125
    },
    {
      "epoch": 18.637176050044683,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0004224170682225914,
      "loss": 0.7979,
      "num_input_tokens_seen": 72651496,
      "step": 125130
    },
    {
      "epoch": 18.63792076258564,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0004219577626788079,
      "loss": 0.8248,
      "num_input_tokens_seen": 72654472,
      "step": 125135
    },
    {
      "epoch": 18.6386654751266,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.00042149870341623153,
      "loss": 0.7974,
      "num_input_tokens_seen": 72657416,
      "step": 125140
    },
    {
      "epoch": 18.63941018766756,
      "grad_norm": 0.0262451171875,
      "learning_rate": 0.0004210398904426177,
      "loss": 0.8017,
      "num_input_tokens_seen": 72660424,
      "step": 125145
    },
    {
      "epoch": 18.64015490020852,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0004205813237657169,
      "loss": 0.7956,
      "num_input_tokens_seen": 72663080,
      "step": 125150
    },
    {
      "epoch": 18.64089961274948,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0004201230033932779,
      "loss": 0.8098,
      "num_input_tokens_seen": 72666216,
      "step": 125155
    },
    {
      "epoch": 18.64164432529044,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.0004196649293330412,
      "loss": 0.8113,
      "num_input_tokens_seen": 72669032,
      "step": 125160
    },
    {
      "epoch": 18.642389037831396,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0004192071015927473,
      "loss": 0.8126,
      "num_input_tokens_seen": 72671592,
      "step": 125165
    },
    {
      "epoch": 18.643133750372357,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.00041874952018013166,
      "loss": 0.7891,
      "num_input_tokens_seen": 72674472,
      "step": 125170
    },
    {
      "epoch": 18.643878462913314,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0004182921851029231,
      "loss": 0.8008,
      "num_input_tokens_seen": 72677448,
      "step": 125175
    },
    {
      "epoch": 18.644623175454274,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00041783509636884707,
      "loss": 0.7826,
      "num_input_tokens_seen": 72680136,
      "step": 125180
    },
    {
      "epoch": 18.645367887995235,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0004173782539856241,
      "loss": 0.8064,
      "num_input_tokens_seen": 72683272,
      "step": 125185
    },
    {
      "epoch": 18.646112600536192,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.000416921657960978,
      "loss": 0.7937,
      "num_input_tokens_seen": 72686440,
      "step": 125190
    },
    {
      "epoch": 18.646857313077152,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00041646530830261594,
      "loss": 0.8001,
      "num_input_tokens_seen": 72689000,
      "step": 125195
    },
    {
      "epoch": 18.647602025618113,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.0004160092050182518,
      "loss": 0.8128,
      "num_input_tokens_seen": 72691816,
      "step": 125200
    },
    {
      "epoch": 18.64834673815907,
      "grad_norm": 0.076171875,
      "learning_rate": 0.00041555334811559097,
      "loss": 0.8139,
      "num_input_tokens_seen": 72694888,
      "step": 125205
    },
    {
      "epoch": 18.64909145070003,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.0004150977376023307,
      "loss": 0.8027,
      "num_input_tokens_seen": 72697896,
      "step": 125210
    },
    {
      "epoch": 18.649836163240987,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.00041464237348617315,
      "loss": 0.8028,
      "num_input_tokens_seen": 72700872,
      "step": 125215
    },
    {
      "epoch": 18.650580875781948,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.0004141872557748055,
      "loss": 0.8069,
      "num_input_tokens_seen": 72703912,
      "step": 125220
    },
    {
      "epoch": 18.65132558832291,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0004137323844759233,
      "loss": 0.7964,
      "num_input_tokens_seen": 72706824,
      "step": 125225
    },
    {
      "epoch": 18.652070300863866,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0004132777595972054,
      "loss": 0.8008,
      "num_input_tokens_seen": 72709576,
      "step": 125230
    },
    {
      "epoch": 18.652815013404826,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0004128233811463355,
      "loss": 0.7997,
      "num_input_tokens_seen": 72712456,
      "step": 125235
    },
    {
      "epoch": 18.653559725945787,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0004123692491309894,
      "loss": 0.7858,
      "num_input_tokens_seen": 72715304,
      "step": 125240
    },
    {
      "epoch": 18.654304438486744,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00041191536355883907,
      "loss": 0.7924,
      "num_input_tokens_seen": 72718120,
      "step": 125245
    },
    {
      "epoch": 18.655049151027704,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.0004114617244375501,
      "loss": 0.7993,
      "num_input_tokens_seen": 72720872,
      "step": 125250
    },
    {
      "epoch": 18.65579386356866,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0004110083317747881,
      "loss": 0.7899,
      "num_input_tokens_seen": 72723656,
      "step": 125255
    },
    {
      "epoch": 18.65653857610962,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.00041055518557821523,
      "loss": 0.7892,
      "num_input_tokens_seen": 72726504,
      "step": 125260
    },
    {
      "epoch": 18.657283288650582,
      "grad_norm": 0.0546875,
      "learning_rate": 0.000410102285855482,
      "loss": 0.7889,
      "num_input_tokens_seen": 72729832,
      "step": 125265
    },
    {
      "epoch": 18.65802800119154,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.00040964963261424234,
      "loss": 0.7846,
      "num_input_tokens_seen": 72732584,
      "step": 125270
    },
    {
      "epoch": 18.6587727137325,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0004091972258621435,
      "loss": 0.8111,
      "num_input_tokens_seen": 72735208,
      "step": 125275
    },
    {
      "epoch": 18.65951742627346,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0004087450656068292,
      "loss": 0.7929,
      "num_input_tokens_seen": 72738152,
      "step": 125280
    },
    {
      "epoch": 18.660262138814417,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.00040829315185593516,
      "loss": 0.8054,
      "num_input_tokens_seen": 72740936,
      "step": 125285
    },
    {
      "epoch": 18.661006851355378,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.00040784148461710023,
      "loss": 0.7982,
      "num_input_tokens_seen": 72743752,
      "step": 125290
    },
    {
      "epoch": 18.661751563896335,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.0004073900638979516,
      "loss": 0.8152,
      "num_input_tokens_seen": 72746472,
      "step": 125295
    },
    {
      "epoch": 18.662496276437295,
      "grad_norm": 0.078125,
      "learning_rate": 0.0004069388897061149,
      "loss": 0.7971,
      "num_input_tokens_seen": 72749320,
      "step": 125300
    },
    {
      "epoch": 18.663240988978256,
      "grad_norm": 0.064453125,
      "learning_rate": 0.00040648796204921734,
      "loss": 0.8055,
      "num_input_tokens_seen": 72752488,
      "step": 125305
    },
    {
      "epoch": 18.663985701519213,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0004060372809348711,
      "loss": 0.8142,
      "num_input_tokens_seen": 72755240,
      "step": 125310
    },
    {
      "epoch": 18.664730414060173,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0004055868463706935,
      "loss": 0.7843,
      "num_input_tokens_seen": 72758152,
      "step": 125315
    },
    {
      "epoch": 18.66547512660113,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0004051366583642918,
      "loss": 0.8039,
      "num_input_tokens_seen": 72761128,
      "step": 125320
    },
    {
      "epoch": 18.66621983914209,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.00040468671692327484,
      "loss": 0.8076,
      "num_input_tokens_seen": 72763816,
      "step": 125325
    },
    {
      "epoch": 18.66696455168305,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0004042370220552399,
      "loss": 0.8137,
      "num_input_tokens_seen": 72766376,
      "step": 125330
    },
    {
      "epoch": 18.66770926422401,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00040378757376778596,
      "loss": 0.8127,
      "num_input_tokens_seen": 72769160,
      "step": 125335
    },
    {
      "epoch": 18.66845397676497,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00040333837206850686,
      "loss": 0.8063,
      "num_input_tokens_seen": 72771944,
      "step": 125340
    },
    {
      "epoch": 18.66919868930593,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0004028894169649899,
      "loss": 0.8106,
      "num_input_tokens_seen": 72774760,
      "step": 125345
    },
    {
      "epoch": 18.669943401846886,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0004024407084648207,
      "loss": 0.7911,
      "num_input_tokens_seen": 72777480,
      "step": 125350
    },
    {
      "epoch": 18.670688114387847,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0004019922465755782,
      "loss": 0.7826,
      "num_input_tokens_seen": 72780328,
      "step": 125355
    },
    {
      "epoch": 18.671432826928804,
      "grad_norm": 0.0272216796875,
      "learning_rate": 0.00040154403130484125,
      "loss": 0.7843,
      "num_input_tokens_seen": 72783048,
      "step": 125360
    },
    {
      "epoch": 18.672177539469764,
      "grad_norm": 0.06640625,
      "learning_rate": 0.00040109606266018226,
      "loss": 0.8011,
      "num_input_tokens_seen": 72786056,
      "step": 125365
    },
    {
      "epoch": 18.672922252010725,
      "grad_norm": 0.0859375,
      "learning_rate": 0.00040064834064916673,
      "loss": 0.7967,
      "num_input_tokens_seen": 72788744,
      "step": 125370
    },
    {
      "epoch": 18.673666964551682,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.00040020086527935703,
      "loss": 0.8092,
      "num_input_tokens_seen": 72791560,
      "step": 125375
    },
    {
      "epoch": 18.674411677092642,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.00039975363655831705,
      "loss": 0.7938,
      "num_input_tokens_seen": 72794600,
      "step": 125380
    },
    {
      "epoch": 18.675156389633603,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0003993066544936008,
      "loss": 0.8152,
      "num_input_tokens_seen": 72797416,
      "step": 125385
    },
    {
      "epoch": 18.67590110217456,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.0003988599190927572,
      "loss": 0.793,
      "num_input_tokens_seen": 72800296,
      "step": 125390
    },
    {
      "epoch": 18.67664581471552,
      "grad_norm": 0.123046875,
      "learning_rate": 0.00039841343036333684,
      "loss": 0.7909,
      "num_input_tokens_seen": 72803144,
      "step": 125395
    },
    {
      "epoch": 18.677390527256478,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0003979671883128805,
      "loss": 0.8118,
      "num_input_tokens_seen": 72806184,
      "step": 125400
    },
    {
      "epoch": 18.678135239797438,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0003975211929489286,
      "loss": 0.7884,
      "num_input_tokens_seen": 72809032,
      "step": 125405
    },
    {
      "epoch": 18.6788799523384,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.00039707544427901363,
      "loss": 0.783,
      "num_input_tokens_seen": 72811816,
      "step": 125410
    },
    {
      "epoch": 18.679624664879356,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0003966299423106678,
      "loss": 0.7922,
      "num_input_tokens_seen": 72814568,
      "step": 125415
    },
    {
      "epoch": 18.680369377420316,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00039618468705141684,
      "loss": 0.7985,
      "num_input_tokens_seen": 72817192,
      "step": 125420
    },
    {
      "epoch": 18.681114089961277,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.000395739678508783,
      "loss": 0.7981,
      "num_input_tokens_seen": 72819880,
      "step": 125425
    },
    {
      "epoch": 18.681858802502234,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.00039529491669028523,
      "loss": 0.8099,
      "num_input_tokens_seen": 72822760,
      "step": 125430
    },
    {
      "epoch": 18.682603515043194,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.000394850401603436,
      "loss": 0.7905,
      "num_input_tokens_seen": 72825704,
      "step": 125435
    },
    {
      "epoch": 18.68334822758415,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.00039440613325574414,
      "loss": 0.7953,
      "num_input_tokens_seen": 72828520,
      "step": 125440
    },
    {
      "epoch": 18.68409294012511,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0003939621116547154,
      "loss": 0.8125,
      "num_input_tokens_seen": 72831208,
      "step": 125445
    },
    {
      "epoch": 18.684837652666072,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0003935183368078537,
      "loss": 0.8121,
      "num_input_tokens_seen": 72834152,
      "step": 125450
    },
    {
      "epoch": 18.68558236520703,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00039307480872265485,
      "loss": 0.8111,
      "num_input_tokens_seen": 72837096,
      "step": 125455
    },
    {
      "epoch": 18.68632707774799,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.00039263152740660944,
      "loss": 0.7924,
      "num_input_tokens_seen": 72839976,
      "step": 125460
    },
    {
      "epoch": 18.687071790288947,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.0003921884928672098,
      "loss": 0.8016,
      "num_input_tokens_seen": 72842664,
      "step": 125465
    },
    {
      "epoch": 18.687816502829907,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.0003917457051119366,
      "loss": 0.8061,
      "num_input_tokens_seen": 72845832,
      "step": 125470
    },
    {
      "epoch": 18.688561215370868,
      "grad_norm": 0.026611328125,
      "learning_rate": 0.00039130316414827393,
      "loss": 0.7858,
      "num_input_tokens_seen": 72848648,
      "step": 125475
    },
    {
      "epoch": 18.689305927911825,
      "grad_norm": 0.060546875,
      "learning_rate": 0.00039086086998369574,
      "loss": 0.8059,
      "num_input_tokens_seen": 72851624,
      "step": 125480
    },
    {
      "epoch": 18.690050640452785,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.00039041882262567603,
      "loss": 0.783,
      "num_input_tokens_seen": 72854504,
      "step": 125485
    },
    {
      "epoch": 18.690795352993746,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00038997702208168225,
      "loss": 0.7949,
      "num_input_tokens_seen": 72857544,
      "step": 125490
    },
    {
      "epoch": 18.691540065534703,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.00038953546835917674,
      "loss": 0.8029,
      "num_input_tokens_seen": 72860840,
      "step": 125495
    },
    {
      "epoch": 18.692284778075663,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.00038909416146562016,
      "loss": 0.7884,
      "num_input_tokens_seen": 72863944,
      "step": 125500
    },
    {
      "epoch": 18.69302949061662,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.00038865310140846987,
      "loss": 0.7909,
      "num_input_tokens_seen": 72866600,
      "step": 125505
    },
    {
      "epoch": 18.69377420315758,
      "grad_norm": 0.03515625,
      "learning_rate": 0.00038821228819517327,
      "loss": 0.8217,
      "num_input_tokens_seen": 72869352,
      "step": 125510
    },
    {
      "epoch": 18.69451891569854,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0003877717218331794,
      "loss": 0.7902,
      "num_input_tokens_seen": 72872168,
      "step": 125515
    },
    {
      "epoch": 18.6952636282395,
      "grad_norm": 0.0595703125,
      "learning_rate": 0.0003873314023299307,
      "loss": 0.8026,
      "num_input_tokens_seen": 72874952,
      "step": 125520
    },
    {
      "epoch": 18.69600834078046,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0003868913296928661,
      "loss": 0.8057,
      "num_input_tokens_seen": 72878280,
      "step": 125525
    },
    {
      "epoch": 18.69675305332142,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.00038645150392942136,
      "loss": 0.8001,
      "num_input_tokens_seen": 72881480,
      "step": 125530
    },
    {
      "epoch": 18.697497765862376,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0003860119250470273,
      "loss": 0.7808,
      "num_input_tokens_seen": 72884456,
      "step": 125535
    },
    {
      "epoch": 18.698242478403337,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0003855725930531062,
      "loss": 0.7831,
      "num_input_tokens_seen": 72887560,
      "step": 125540
    },
    {
      "epoch": 18.698987190944294,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0003851335079550838,
      "loss": 0.8356,
      "num_input_tokens_seen": 72890376,
      "step": 125545
    },
    {
      "epoch": 18.699731903485254,
      "grad_norm": 0.030517578125,
      "learning_rate": 0.0003846946697603759,
      "loss": 0.7995,
      "num_input_tokens_seen": 72893288,
      "step": 125550
    },
    {
      "epoch": 18.700476616026215,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.0003842560784763982,
      "loss": 0.7828,
      "num_input_tokens_seen": 72896168,
      "step": 125555
    },
    {
      "epoch": 18.701221328567172,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.00038381773411055804,
      "loss": 0.791,
      "num_input_tokens_seen": 72899432,
      "step": 125560
    },
    {
      "epoch": 18.701966041108133,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.00038337963667026294,
      "loss": 0.7918,
      "num_input_tokens_seen": 72902344,
      "step": 125565
    },
    {
      "epoch": 18.702710753649093,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0003829417861629153,
      "loss": 0.808,
      "num_input_tokens_seen": 72905128,
      "step": 125570
    },
    {
      "epoch": 18.70345546619005,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.00038250418259590745,
      "loss": 0.799,
      "num_input_tokens_seen": 72907784,
      "step": 125575
    },
    {
      "epoch": 18.70420017873101,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0003820668259766369,
      "loss": 0.8196,
      "num_input_tokens_seen": 72910536,
      "step": 125580
    },
    {
      "epoch": 18.704944891271968,
      "grad_norm": 0.0390625,
      "learning_rate": 0.00038162971631248765,
      "loss": 0.8039,
      "num_input_tokens_seen": 72913256,
      "step": 125585
    },
    {
      "epoch": 18.705689603812928,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00038119285361084887,
      "loss": 0.8045,
      "num_input_tokens_seen": 72916328,
      "step": 125590
    },
    {
      "epoch": 18.70643431635389,
      "grad_norm": 0.078125,
      "learning_rate": 0.00038075623787909793,
      "loss": 0.8043,
      "num_input_tokens_seen": 72919432,
      "step": 125595
    },
    {
      "epoch": 18.707179028894846,
      "grad_norm": 0.10546875,
      "learning_rate": 0.00038031986912461234,
      "loss": 0.7919,
      "num_input_tokens_seen": 72922344,
      "step": 125600
    },
    {
      "epoch": 18.707923741435806,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.0003798837473547628,
      "loss": 0.8119,
      "num_input_tokens_seen": 72925320,
      "step": 125605
    },
    {
      "epoch": 18.708668453976763,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.0003794478725769201,
      "loss": 0.8152,
      "num_input_tokens_seen": 72928264,
      "step": 125610
    },
    {
      "epoch": 18.709413166517724,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.00037901224479844505,
      "loss": 0.8,
      "num_input_tokens_seen": 72930728,
      "step": 125615
    },
    {
      "epoch": 18.710157879058684,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00037857686402669675,
      "loss": 0.8012,
      "num_input_tokens_seen": 72933672,
      "step": 125620
    },
    {
      "epoch": 18.71090259159964,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0003781417302690343,
      "loss": 0.7856,
      "num_input_tokens_seen": 72936552,
      "step": 125625
    },
    {
      "epoch": 18.7116473041406,
      "grad_norm": 0.0284423828125,
      "learning_rate": 0.00037770684353280345,
      "loss": 0.7938,
      "num_input_tokens_seen": 72939944,
      "step": 125630
    },
    {
      "epoch": 18.712392016681562,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.00037727220382535677,
      "loss": 0.7841,
      "num_input_tokens_seen": 72942632,
      "step": 125635
    },
    {
      "epoch": 18.71313672922252,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.00037683781115403325,
      "loss": 0.7985,
      "num_input_tokens_seen": 72945672,
      "step": 125640
    },
    {
      "epoch": 18.71388144176348,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0003764036655261721,
      "loss": 0.7865,
      "num_input_tokens_seen": 72948520,
      "step": 125645
    },
    {
      "epoch": 18.714626154304437,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.00037596976694911075,
      "loss": 0.8211,
      "num_input_tokens_seen": 72951400,
      "step": 125650
    },
    {
      "epoch": 18.715370866845397,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0003755361154301734,
      "loss": 0.816,
      "num_input_tokens_seen": 72954408,
      "step": 125655
    },
    {
      "epoch": 18.716115579386358,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.00037510271097669076,
      "loss": 0.8048,
      "num_input_tokens_seen": 72957608,
      "step": 125660
    },
    {
      "epoch": 18.716860291927315,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0003746695535959837,
      "loss": 0.7909,
      "num_input_tokens_seen": 72960328,
      "step": 125665
    },
    {
      "epoch": 18.717605004468275,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0003742366432953714,
      "loss": 0.8073,
      "num_input_tokens_seen": 72963304,
      "step": 125670
    },
    {
      "epoch": 18.718349717009236,
      "grad_norm": 0.0390625,
      "learning_rate": 0.00037380398008216465,
      "loss": 0.7938,
      "num_input_tokens_seen": 72966056,
      "step": 125675
    },
    {
      "epoch": 18.719094429550193,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0003733715639636742,
      "loss": 0.8041,
      "num_input_tokens_seen": 72968744,
      "step": 125680
    },
    {
      "epoch": 18.719839142091153,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.00037293939494720594,
      "loss": 0.8112,
      "num_input_tokens_seen": 72971720,
      "step": 125685
    },
    {
      "epoch": 18.72058385463211,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.0003725074730400607,
      "loss": 0.7993,
      "num_input_tokens_seen": 72974600,
      "step": 125690
    },
    {
      "epoch": 18.72132856717307,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.00037207579824953427,
      "loss": 0.7957,
      "num_input_tokens_seen": 72977320,
      "step": 125695
    },
    {
      "epoch": 18.72207327971403,
      "grad_norm": 0.052734375,
      "learning_rate": 0.00037164437058292087,
      "loss": 0.7953,
      "num_input_tokens_seen": 72980200,
      "step": 125700
    },
    {
      "epoch": 18.72281799225499,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.00037121319004750796,
      "loss": 0.7831,
      "num_input_tokens_seen": 72983208,
      "step": 125705
    },
    {
      "epoch": 18.72356270479595,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0003707822566505797,
      "loss": 0.829,
      "num_input_tokens_seen": 72986056,
      "step": 125710
    },
    {
      "epoch": 18.72430741733691,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.000370351570399417,
      "loss": 0.8143,
      "num_input_tokens_seen": 72989000,
      "step": 125715
    },
    {
      "epoch": 18.725052129877866,
      "grad_norm": 0.201171875,
      "learning_rate": 0.0003699211313012957,
      "loss": 0.8189,
      "num_input_tokens_seen": 72991976,
      "step": 125720
    },
    {
      "epoch": 18.725796842418827,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0003694909393634882,
      "loss": 0.8049,
      "num_input_tokens_seen": 72995080,
      "step": 125725
    },
    {
      "epoch": 18.726541554959784,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.00036906099459326045,
      "loss": 0.8166,
      "num_input_tokens_seen": 72998088,
      "step": 125730
    },
    {
      "epoch": 18.727286267500745,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00036863129699787667,
      "loss": 0.793,
      "num_input_tokens_seen": 73000744,
      "step": 125735
    },
    {
      "epoch": 18.728030980041705,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00036820184658459764,
      "loss": 0.8029,
      "num_input_tokens_seen": 73003368,
      "step": 125740
    },
    {
      "epoch": 18.728775692582662,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0003677726433606759,
      "loss": 0.7941,
      "num_input_tokens_seen": 73006248,
      "step": 125745
    },
    {
      "epoch": 18.729520405123623,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.00036734368733336406,
      "loss": 0.774,
      "num_input_tokens_seen": 73009032,
      "step": 125750
    },
    {
      "epoch": 18.730265117664583,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0003669149785099096,
      "loss": 0.8093,
      "num_input_tokens_seen": 73011624,
      "step": 125755
    },
    {
      "epoch": 18.73100983020554,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.00036648651689755337,
      "loss": 0.8079,
      "num_input_tokens_seen": 73014440,
      "step": 125760
    },
    {
      "epoch": 18.7317545427465,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0003660583025035363,
      "loss": 0.786,
      "num_input_tokens_seen": 73017704,
      "step": 125765
    },
    {
      "epoch": 18.732499255287458,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.00036563033533508927,
      "loss": 0.7756,
      "num_input_tokens_seen": 73020296,
      "step": 125770
    },
    {
      "epoch": 18.733243967828418,
      "grad_norm": 0.0267333984375,
      "learning_rate": 0.00036520261539944475,
      "loss": 0.7911,
      "num_input_tokens_seen": 73023176,
      "step": 125775
    },
    {
      "epoch": 18.73398868036938,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.000364775142703827,
      "loss": 0.8041,
      "num_input_tokens_seen": 73025992,
      "step": 125780
    },
    {
      "epoch": 18.734733392910336,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.00036434791725546033,
      "loss": 0.8135,
      "num_input_tokens_seen": 73028968,
      "step": 125785
    },
    {
      "epoch": 18.735478105451296,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.00036392093906155887,
      "loss": 0.7849,
      "num_input_tokens_seen": 73031784,
      "step": 125790
    },
    {
      "epoch": 18.736222817992257,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.00036349420812933854,
      "loss": 0.7934,
      "num_input_tokens_seen": 73034728,
      "step": 125795
    },
    {
      "epoch": 18.736967530533214,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0003630677244660069,
      "loss": 0.8326,
      "num_input_tokens_seen": 73037928,
      "step": 125800
    },
    {
      "epoch": 18.737712243074174,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0003626414880787698,
      "loss": 0.7819,
      "num_input_tokens_seen": 73040744,
      "step": 125805
    },
    {
      "epoch": 18.73845695561513,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0003622154989748283,
      "loss": 0.7915,
      "num_input_tokens_seen": 73043848,
      "step": 125810
    },
    {
      "epoch": 18.739201668156092,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0003617897571613798,
      "loss": 0.8131,
      "num_input_tokens_seen": 73046472,
      "step": 125815
    },
    {
      "epoch": 18.739946380697052,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.00036136426264561536,
      "loss": 0.7931,
      "num_input_tokens_seen": 73049512,
      "step": 125820
    },
    {
      "epoch": 18.74069109323801,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00036093901543472084,
      "loss": 0.8013,
      "num_input_tokens_seen": 73052520,
      "step": 125825
    },
    {
      "epoch": 18.74143580577897,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0003605140155358871,
      "loss": 0.7876,
      "num_input_tokens_seen": 73055240,
      "step": 125830
    },
    {
      "epoch": 18.742180518319927,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0003600892629562885,
      "loss": 0.7843,
      "num_input_tokens_seen": 73058152,
      "step": 125835
    },
    {
      "epoch": 18.742925230860887,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00035966475770310256,
      "loss": 0.788,
      "num_input_tokens_seen": 73061096,
      "step": 125840
    },
    {
      "epoch": 18.743669943401848,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.00035924049978350027,
      "loss": 0.8022,
      "num_input_tokens_seen": 73063720,
      "step": 125845
    },
    {
      "epoch": 18.744414655942805,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0003588164892046508,
      "loss": 0.7942,
      "num_input_tokens_seen": 73066728,
      "step": 125850
    },
    {
      "epoch": 18.745159368483765,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0003583927259737152,
      "loss": 0.7956,
      "num_input_tokens_seen": 73069512,
      "step": 125855
    },
    {
      "epoch": 18.745904081024726,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.00035796921009785266,
      "loss": 0.7796,
      "num_input_tokens_seen": 73072328,
      "step": 125860
    },
    {
      "epoch": 18.746648793565683,
      "grad_norm": 0.0615234375,
      "learning_rate": 0.0003575459415842208,
      "loss": 0.8236,
      "num_input_tokens_seen": 73075176,
      "step": 125865
    },
    {
      "epoch": 18.747393506106643,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0003571229204399656,
      "loss": 0.817,
      "num_input_tokens_seen": 73077928,
      "step": 125870
    },
    {
      "epoch": 18.7481382186476,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00035670014667223967,
      "loss": 0.7864,
      "num_input_tokens_seen": 73081032,
      "step": 125875
    },
    {
      "epoch": 18.74888293118856,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.000356277620288179,
      "loss": 0.7841,
      "num_input_tokens_seen": 73084136,
      "step": 125880
    },
    {
      "epoch": 18.74962764372952,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0003558553412949261,
      "loss": 0.8112,
      "num_input_tokens_seen": 73087112,
      "step": 125885
    },
    {
      "epoch": 18.75037235627048,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.0003554333096996137,
      "loss": 0.8065,
      "num_input_tokens_seen": 73089992,
      "step": 125890
    },
    {
      "epoch": 18.75111706881144,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00035501152550937106,
      "loss": 0.8076,
      "num_input_tokens_seen": 73092808,
      "step": 125895
    },
    {
      "epoch": 18.7518617813524,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0003545899887313242,
      "loss": 0.7927,
      "num_input_tokens_seen": 73095784,
      "step": 125900
    },
    {
      "epoch": 18.752606493893357,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0003541686993725956,
      "loss": 0.8334,
      "num_input_tokens_seen": 73098824,
      "step": 125905
    },
    {
      "epoch": 18.753351206434317,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.00035374765744030143,
      "loss": 0.7941,
      "num_input_tokens_seen": 73101320,
      "step": 125910
    },
    {
      "epoch": 18.754095918975274,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.0003533268629415526,
      "loss": 0.8087,
      "num_input_tokens_seen": 73104104,
      "step": 125915
    },
    {
      "epoch": 18.754840631516235,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.00035290631588346164,
      "loss": 0.8104,
      "num_input_tokens_seen": 73106984,
      "step": 125920
    },
    {
      "epoch": 18.755585344057195,
      "grad_norm": 0.0546875,
      "learning_rate": 0.00035248601627313133,
      "loss": 0.7902,
      "num_input_tokens_seen": 73110696,
      "step": 125925
    },
    {
      "epoch": 18.756330056598152,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0003520659641176643,
      "loss": 0.7961,
      "num_input_tokens_seen": 73113704,
      "step": 125930
    },
    {
      "epoch": 18.757074769139113,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0003516461594241532,
      "loss": 0.7841,
      "num_input_tokens_seen": 73116584,
      "step": 125935
    },
    {
      "epoch": 18.757819481680073,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.00035122660219969235,
      "loss": 0.7685,
      "num_input_tokens_seen": 73119240,
      "step": 125940
    },
    {
      "epoch": 18.75856419422103,
      "grad_norm": 0.37109375,
      "learning_rate": 0.0003508072924513711,
      "loss": 0.8534,
      "num_input_tokens_seen": 73122280,
      "step": 125945
    },
    {
      "epoch": 18.75930890676199,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0003503882301862704,
      "loss": 0.8402,
      "num_input_tokens_seen": 73125128,
      "step": 125950
    },
    {
      "epoch": 18.760053619302948,
      "grad_norm": 0.099609375,
      "learning_rate": 0.00034996941541147296,
      "loss": 0.7921,
      "num_input_tokens_seen": 73128072,
      "step": 125955
    },
    {
      "epoch": 18.760798331843908,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00034955084813405145,
      "loss": 0.7886,
      "num_input_tokens_seen": 73131048,
      "step": 125960
    },
    {
      "epoch": 18.76154304438487,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.0003491325283610802,
      "loss": 0.7904,
      "num_input_tokens_seen": 73133768,
      "step": 125965
    },
    {
      "epoch": 18.762287756925826,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0003487144560996236,
      "loss": 0.7921,
      "num_input_tokens_seen": 73136904,
      "step": 125970
    },
    {
      "epoch": 18.763032469466786,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00034829663135674423,
      "loss": 0.8096,
      "num_input_tokens_seen": 73139944,
      "step": 125975
    },
    {
      "epoch": 18.763777182007743,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.00034787905413950326,
      "loss": 0.7837,
      "num_input_tokens_seen": 73142984,
      "step": 125980
    },
    {
      "epoch": 18.764521894548704,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00034746172445495325,
      "loss": 0.8101,
      "num_input_tokens_seen": 73145864,
      "step": 125985
    },
    {
      "epoch": 18.765266607089664,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0003470446423101453,
      "loss": 0.8167,
      "num_input_tokens_seen": 73148584,
      "step": 125990
    },
    {
      "epoch": 18.76601131963062,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00034662780771212376,
      "loss": 0.7933,
      "num_input_tokens_seen": 73151400,
      "step": 125995
    },
    {
      "epoch": 18.766756032171582,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.00034621122066793463,
      "loss": 0.8055,
      "num_input_tokens_seen": 73154152,
      "step": 126000
    },
    {
      "epoch": 18.767500744712542,
      "grad_norm": 0.060546875,
      "learning_rate": 0.00034579488118461233,
      "loss": 0.7922,
      "num_input_tokens_seen": 73156808,
      "step": 126005
    },
    {
      "epoch": 18.7682454572535,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.00034537878926919283,
      "loss": 0.7931,
      "num_input_tokens_seen": 73159688,
      "step": 126010
    },
    {
      "epoch": 18.76899016979446,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00034496294492870395,
      "loss": 0.8102,
      "num_input_tokens_seen": 73162376,
      "step": 126015
    },
    {
      "epoch": 18.769734882335417,
      "grad_norm": 0.035400390625,
      "learning_rate": 0.00034454734817016995,
      "loss": 0.8009,
      "num_input_tokens_seen": 73165032,
      "step": 126020
    },
    {
      "epoch": 18.770479594876377,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0003441319990006153,
      "loss": 0.804,
      "num_input_tokens_seen": 73167816,
      "step": 126025
    },
    {
      "epoch": 18.771224307417338,
      "grad_norm": 0.060546875,
      "learning_rate": 0.00034371689742705437,
      "loss": 0.7806,
      "num_input_tokens_seen": 73170504,
      "step": 126030
    },
    {
      "epoch": 18.771969019958295,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0003433020434564998,
      "loss": 0.8051,
      "num_input_tokens_seen": 73173128,
      "step": 126035
    },
    {
      "epoch": 18.772713732499255,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.00034288743709596115,
      "loss": 0.7964,
      "num_input_tokens_seen": 73175880,
      "step": 126040
    },
    {
      "epoch": 18.773458445040216,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0003424730783524443,
      "loss": 0.8139,
      "num_input_tokens_seen": 73178760,
      "step": 126045
    },
    {
      "epoch": 18.774203157581173,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0003420589672329455,
      "loss": 0.7863,
      "num_input_tokens_seen": 73181512,
      "step": 126050
    },
    {
      "epoch": 18.774947870122134,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0003416451037444623,
      "loss": 0.7957,
      "num_input_tokens_seen": 73184648,
      "step": 126055
    },
    {
      "epoch": 18.77569258266309,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.00034123148789398926,
      "loss": 0.8144,
      "num_input_tokens_seen": 73187784,
      "step": 126060
    },
    {
      "epoch": 18.77643729520405,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.00034081811968850905,
      "loss": 0.8113,
      "num_input_tokens_seen": 73190760,
      "step": 126065
    },
    {
      "epoch": 18.77718200774501,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0003404049991350094,
      "loss": 0.8165,
      "num_input_tokens_seen": 73193416,
      "step": 126070
    },
    {
      "epoch": 18.77792672028597,
      "grad_norm": 0.025390625,
      "learning_rate": 0.0003399921262404665,
      "loss": 0.7906,
      "num_input_tokens_seen": 73196424,
      "step": 126075
    },
    {
      "epoch": 18.77867143282693,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0003395795010118596,
      "loss": 0.7838,
      "num_input_tokens_seen": 73199240,
      "step": 126080
    },
    {
      "epoch": 18.77941614536789,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00033916712345615326,
      "loss": 0.7939,
      "num_input_tokens_seen": 73202056,
      "step": 126085
    },
    {
      "epoch": 18.780160857908847,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.0003387549935803202,
      "loss": 0.8165,
      "num_input_tokens_seen": 73205096,
      "step": 126090
    },
    {
      "epoch": 18.780905570449807,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0003383431113913215,
      "loss": 0.8002,
      "num_input_tokens_seen": 73207624,
      "step": 126095
    },
    {
      "epoch": 18.781650282990764,
      "grad_norm": 0.041015625,
      "learning_rate": 0.00033793147689611334,
      "loss": 0.8041,
      "num_input_tokens_seen": 73210504,
      "step": 126100
    },
    {
      "epoch": 18.782394995531725,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.00033752009010165005,
      "loss": 0.8103,
      "num_input_tokens_seen": 73213256,
      "step": 126105
    },
    {
      "epoch": 18.783139708072685,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.0003371089510148828,
      "loss": 0.8063,
      "num_input_tokens_seen": 73216680,
      "step": 126110
    },
    {
      "epoch": 18.783884420613642,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.00033669805964275766,
      "loss": 0.7993,
      "num_input_tokens_seen": 73219848,
      "step": 126115
    },
    {
      "epoch": 18.784629133154603,
      "grad_norm": 0.0673828125,
      "learning_rate": 0.00033628741599221577,
      "loss": 0.7887,
      "num_input_tokens_seen": 73222856,
      "step": 126120
    },
    {
      "epoch": 18.785373845695563,
      "grad_norm": 0.0303955078125,
      "learning_rate": 0.0003358770200701949,
      "loss": 0.7839,
      "num_input_tokens_seen": 73225608,
      "step": 126125
    },
    {
      "epoch": 18.78611855823652,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.0003354668718836262,
      "loss": 0.7878,
      "num_input_tokens_seen": 73228712,
      "step": 126130
    },
    {
      "epoch": 18.78686327077748,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0003350569714394408,
      "loss": 0.7999,
      "num_input_tokens_seen": 73231560,
      "step": 126135
    },
    {
      "epoch": 18.787607983318438,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00033464731874456474,
      "loss": 0.8043,
      "num_input_tokens_seen": 73234760,
      "step": 126140
    },
    {
      "epoch": 18.7883526958594,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.00033423791380591427,
      "loss": 0.7968,
      "num_input_tokens_seen": 73237544,
      "step": 126145
    },
    {
      "epoch": 18.78909740840036,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00033382875663041044,
      "loss": 0.7882,
      "num_input_tokens_seen": 73240520,
      "step": 126150
    },
    {
      "epoch": 18.789842120941316,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.00033341984722496277,
      "loss": 0.771,
      "num_input_tokens_seen": 73243560,
      "step": 126155
    },
    {
      "epoch": 18.790586833482276,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.00033301118559648233,
      "loss": 0.8025,
      "num_input_tokens_seen": 73246536,
      "step": 126160
    },
    {
      "epoch": 18.791331546023237,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.00033260277175187035,
      "loss": 0.7868,
      "num_input_tokens_seen": 73249640,
      "step": 126165
    },
    {
      "epoch": 18.792076258564194,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.00033219460569802625,
      "loss": 0.7942,
      "num_input_tokens_seen": 73252712,
      "step": 126170
    },
    {
      "epoch": 18.792820971105154,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0003317866874418479,
      "loss": 0.7921,
      "num_input_tokens_seen": 73255656,
      "step": 126175
    },
    {
      "epoch": 18.79356568364611,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.00033137901699022474,
      "loss": 0.8102,
      "num_input_tokens_seen": 73258440,
      "step": 126180
    },
    {
      "epoch": 18.794310396187072,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0003309715943500446,
      "loss": 0.8412,
      "num_input_tokens_seen": 73261448,
      "step": 126185
    },
    {
      "epoch": 18.795055108728032,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0003305644195281887,
      "loss": 0.7964,
      "num_input_tokens_seen": 73264232,
      "step": 126190
    },
    {
      "epoch": 18.79579982126899,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0003301574925315398,
      "loss": 0.7938,
      "num_input_tokens_seen": 73266856,
      "step": 126195
    },
    {
      "epoch": 18.79654453380995,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.00032975081336696907,
      "loss": 0.8376,
      "num_input_tokens_seen": 73269544,
      "step": 126200
    },
    {
      "epoch": 18.797289246350907,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0003293443820413494,
      "loss": 0.787,
      "num_input_tokens_seen": 73272488,
      "step": 126205
    },
    {
      "epoch": 18.798033958891867,
      "grad_norm": 0.09375,
      "learning_rate": 0.0003289381985615436,
      "loss": 0.7808,
      "num_input_tokens_seen": 73275016,
      "step": 126210
    },
    {
      "epoch": 18.798778671432828,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.0003285322629344178,
      "loss": 0.8068,
      "num_input_tokens_seen": 73278120,
      "step": 126215
    },
    {
      "epoch": 18.799523383973785,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0003281265751668283,
      "loss": 0.7865,
      "num_input_tokens_seen": 73281224,
      "step": 126220
    },
    {
      "epoch": 18.800268096514746,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.0003277211352656262,
      "loss": 0.7835,
      "num_input_tokens_seen": 73284232,
      "step": 126225
    },
    {
      "epoch": 18.801012809055706,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.000327315943237666,
      "loss": 0.7954,
      "num_input_tokens_seen": 73286984,
      "step": 126230
    },
    {
      "epoch": 18.801757521596663,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0003269109990897889,
      "loss": 0.7927,
      "num_input_tokens_seen": 73289768,
      "step": 126235
    },
    {
      "epoch": 18.802502234137624,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0003265063028288395,
      "loss": 0.8101,
      "num_input_tokens_seen": 73292712,
      "step": 126240
    },
    {
      "epoch": 18.80324694667858,
      "grad_norm": 0.3359375,
      "learning_rate": 0.000326101854461649,
      "loss": 0.8098,
      "num_input_tokens_seen": 73295720,
      "step": 126245
    },
    {
      "epoch": 18.80399165921954,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0003256976539950568,
      "loss": 0.782,
      "num_input_tokens_seen": 73298568,
      "step": 126250
    },
    {
      "epoch": 18.8047363717605,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0003252937014358875,
      "loss": 0.7952,
      "num_input_tokens_seen": 73301320,
      "step": 126255
    },
    {
      "epoch": 18.80548108430146,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0003248899967909657,
      "loss": 0.812,
      "num_input_tokens_seen": 73304104,
      "step": 126260
    },
    {
      "epoch": 18.80622579684242,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00032448654006711264,
      "loss": 0.7811,
      "num_input_tokens_seen": 73306952,
      "step": 126265
    },
    {
      "epoch": 18.80697050938338,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.00032408333127114275,
      "loss": 0.7993,
      "num_input_tokens_seen": 73309800,
      "step": 126270
    },
    {
      "epoch": 18.807715221924337,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0003236803704098706,
      "loss": 0.7913,
      "num_input_tokens_seen": 73312808,
      "step": 126275
    },
    {
      "epoch": 18.808459934465297,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.00032327765749010247,
      "loss": 0.804,
      "num_input_tokens_seen": 73315752,
      "step": 126280
    },
    {
      "epoch": 18.809204647006254,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.00032287519251864125,
      "loss": 0.8141,
      "num_input_tokens_seen": 73318760,
      "step": 126285
    },
    {
      "epoch": 18.809949359547215,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0003224729755022848,
      "loss": 0.7938,
      "num_input_tokens_seen": 73321576,
      "step": 126290
    },
    {
      "epoch": 18.810694072088175,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0003220710064478327,
      "loss": 0.7945,
      "num_input_tokens_seen": 73324744,
      "step": 126295
    },
    {
      "epoch": 18.811438784629132,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.00032166928536207117,
      "loss": 0.7864,
      "num_input_tokens_seen": 73327560,
      "step": 126300
    },
    {
      "epoch": 18.812183497170093,
      "grad_norm": 0.060546875,
      "learning_rate": 0.0003212678122517898,
      "loss": 0.8083,
      "num_input_tokens_seen": 73330600,
      "step": 126305
    },
    {
      "epoch": 18.812928209711053,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0003208665871237698,
      "loss": 0.8085,
      "num_input_tokens_seen": 73333320,
      "step": 126310
    },
    {
      "epoch": 18.81367292225201,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0003204656099847891,
      "loss": 0.8148,
      "num_input_tokens_seen": 73336616,
      "step": 126315
    },
    {
      "epoch": 18.81441763479297,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0003200648808416223,
      "loss": 0.7998,
      "num_input_tokens_seen": 73339496,
      "step": 126320
    },
    {
      "epoch": 18.815162347333928,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.00031966439970103897,
      "loss": 0.8138,
      "num_input_tokens_seen": 73342248,
      "step": 126325
    },
    {
      "epoch": 18.81590705987489,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00031926416656980535,
      "loss": 0.8023,
      "num_input_tokens_seen": 73344872,
      "step": 126330
    },
    {
      "epoch": 18.81665177241585,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0003188641814546844,
      "loss": 0.7884,
      "num_input_tokens_seen": 73347976,
      "step": 126335
    },
    {
      "epoch": 18.817396484956806,
      "grad_norm": 0.033203125,
      "learning_rate": 0.00031846444436242903,
      "loss": 0.7876,
      "num_input_tokens_seen": 73351016,
      "step": 126340
    },
    {
      "epoch": 18.818141197497766,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0003180649552997972,
      "loss": 0.8094,
      "num_input_tokens_seen": 73354248,
      "step": 126345
    },
    {
      "epoch": 18.818885910038723,
      "grad_norm": 0.044921875,
      "learning_rate": 0.0003176657142735351,
      "loss": 0.8164,
      "num_input_tokens_seen": 73357224,
      "step": 126350
    },
    {
      "epoch": 18.819630622579684,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.000317266721290389,
      "loss": 0.8048,
      "num_input_tokens_seen": 73360104,
      "step": 126355
    },
    {
      "epoch": 18.820375335120644,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0003168679763570986,
      "loss": 0.7833,
      "num_input_tokens_seen": 73362984,
      "step": 126360
    },
    {
      "epoch": 18.8211200476616,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.00031646947948039847,
      "loss": 0.808,
      "num_input_tokens_seen": 73365992,
      "step": 126365
    },
    {
      "epoch": 18.821864760202562,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.0003160712306670249,
      "loss": 0.7965,
      "num_input_tokens_seen": 73368744,
      "step": 126370
    },
    {
      "epoch": 18.822609472743522,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00031567322992370414,
      "loss": 0.7815,
      "num_input_tokens_seen": 73371656,
      "step": 126375
    },
    {
      "epoch": 18.82335418528448,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0003152754772571609,
      "loss": 0.7996,
      "num_input_tokens_seen": 73374792,
      "step": 126380
    },
    {
      "epoch": 18.82409889782544,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.00031487797267410963,
      "loss": 0.7973,
      "num_input_tokens_seen": 73378024,
      "step": 126385
    },
    {
      "epoch": 18.824843610366397,
      "grad_norm": 0.056640625,
      "learning_rate": 0.0003144807161812718,
      "loss": 0.8133,
      "num_input_tokens_seen": 73381192,
      "step": 126390
    },
    {
      "epoch": 18.825588322907358,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00031408370778535696,
      "loss": 0.8094,
      "num_input_tokens_seen": 73384104,
      "step": 126395
    },
    {
      "epoch": 18.826333035448318,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.00031368694749307145,
      "loss": 0.7885,
      "num_input_tokens_seen": 73387080,
      "step": 126400
    },
    {
      "epoch": 18.827077747989275,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0003132904353111182,
      "loss": 0.8125,
      "num_input_tokens_seen": 73389928,
      "step": 126405
    },
    {
      "epoch": 18.827822460530236,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0003128941712461969,
      "loss": 0.8127,
      "num_input_tokens_seen": 73392488,
      "step": 126410
    },
    {
      "epoch": 18.828567173071196,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0003124981553049988,
      "loss": 0.8166,
      "num_input_tokens_seen": 73395240,
      "step": 126415
    },
    {
      "epoch": 18.829311885612153,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0003121023874942186,
      "loss": 0.8015,
      "num_input_tokens_seen": 73398216,
      "step": 126420
    },
    {
      "epoch": 18.830056598153114,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0003117068678205409,
      "loss": 0.7886,
      "num_input_tokens_seen": 73401128,
      "step": 126425
    },
    {
      "epoch": 18.83080131069407,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.00031131159629064706,
      "loss": 0.7831,
      "num_input_tokens_seen": 73404328,
      "step": 126430
    },
    {
      "epoch": 18.83154602323503,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.0003109165729112151,
      "loss": 0.7947,
      "num_input_tokens_seen": 73406952,
      "step": 126435
    },
    {
      "epoch": 18.83229073577599,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.000310521797688918,
      "loss": 0.7897,
      "num_input_tokens_seen": 73410152,
      "step": 126440
    },
    {
      "epoch": 18.83303544831695,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.00031012727063042534,
      "loss": 0.7973,
      "num_input_tokens_seen": 73412744,
      "step": 126445
    },
    {
      "epoch": 18.83378016085791,
      "grad_norm": 0.049560546875,
      "learning_rate": 0.0003097329917424002,
      "loss": 0.8055,
      "num_input_tokens_seen": 73415688,
      "step": 126450
    },
    {
      "epoch": 18.83452487339887,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.0003093389610315089,
      "loss": 0.7817,
      "num_input_tokens_seen": 73418440,
      "step": 126455
    },
    {
      "epoch": 18.835269585939827,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0003089451785044045,
      "loss": 0.8131,
      "num_input_tokens_seen": 73421288,
      "step": 126460
    },
    {
      "epoch": 18.836014298480787,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.00030855164416773826,
      "loss": 0.7651,
      "num_input_tokens_seen": 73424264,
      "step": 126465
    },
    {
      "epoch": 18.836759011021744,
      "grad_norm": 0.068359375,
      "learning_rate": 0.0003081583580281616,
      "loss": 0.7972,
      "num_input_tokens_seen": 73427272,
      "step": 126470
    },
    {
      "epoch": 18.837503723562705,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.0003077653200923158,
      "loss": 0.7985,
      "num_input_tokens_seen": 73430472,
      "step": 126475
    },
    {
      "epoch": 18.838248436103665,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.00030737253036684396,
      "loss": 0.784,
      "num_input_tokens_seen": 73433672,
      "step": 126480
    },
    {
      "epoch": 18.838993148644622,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.00030697998885837905,
      "loss": 0.803,
      "num_input_tokens_seen": 73436552,
      "step": 126485
    },
    {
      "epoch": 18.839737861185583,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.00030658769557355414,
      "loss": 0.8116,
      "num_input_tokens_seen": 73439272,
      "step": 126490
    },
    {
      "epoch": 18.84048257372654,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0003061956505189972,
      "loss": 0.8034,
      "num_input_tokens_seen": 73442216,
      "step": 126495
    },
    {
      "epoch": 18.8412272862675,
      "grad_norm": 0.05078125,
      "learning_rate": 0.0003058038537013313,
      "loss": 0.799,
      "num_input_tokens_seen": 73445160,
      "step": 126500
    },
    {
      "epoch": 18.84197199880846,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0003054123051271745,
      "loss": 0.7963,
      "num_input_tokens_seen": 73448072,
      "step": 126505
    },
    {
      "epoch": 18.842716711349418,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.00030502100480313987,
      "loss": 0.8006,
      "num_input_tokens_seen": 73451144,
      "step": 126510
    },
    {
      "epoch": 18.84346142389038,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00030462995273584204,
      "loss": 0.8123,
      "num_input_tokens_seen": 73454056,
      "step": 126515
    },
    {
      "epoch": 18.84420613643134,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.00030423914893188406,
      "loss": 0.8018,
      "num_input_tokens_seen": 73457064,
      "step": 126520
    },
    {
      "epoch": 18.844950848972296,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.0003038485933978707,
      "loss": 0.8012,
      "num_input_tokens_seen": 73459816,
      "step": 126525
    },
    {
      "epoch": 18.845695561513256,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.0003034582861403967,
      "loss": 0.7982,
      "num_input_tokens_seen": 73462248,
      "step": 126530
    },
    {
      "epoch": 18.846440274054213,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0003030682271660617,
      "loss": 0.797,
      "num_input_tokens_seen": 73465160,
      "step": 126535
    },
    {
      "epoch": 18.847184986595174,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0003026784164814489,
      "loss": 0.8096,
      "num_input_tokens_seen": 73468072,
      "step": 126540
    },
    {
      "epoch": 18.847929699136134,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0003022888540931479,
      "loss": 0.8215,
      "num_input_tokens_seen": 73471176,
      "step": 126545
    },
    {
      "epoch": 18.84867441167709,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.00030189954000773845,
      "loss": 0.8109,
      "num_input_tokens_seen": 73473896,
      "step": 126550
    },
    {
      "epoch": 18.849419124218052,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.000301510474231797,
      "loss": 0.7916,
      "num_input_tokens_seen": 73476968,
      "step": 126555
    },
    {
      "epoch": 18.850163836759013,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00030112165677189827,
      "loss": 0.7956,
      "num_input_tokens_seen": 73480168,
      "step": 126560
    },
    {
      "epoch": 18.85090854929997,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0003007330876346087,
      "loss": 0.8065,
      "num_input_tokens_seen": 73483016,
      "step": 126565
    },
    {
      "epoch": 18.85165326184093,
      "grad_norm": 0.0390625,
      "learning_rate": 0.0003003447668264947,
      "loss": 0.7876,
      "num_input_tokens_seen": 73485800,
      "step": 126570
    },
    {
      "epoch": 18.852397974381887,
      "grad_norm": 0.052734375,
      "learning_rate": 0.00029995669435411597,
      "loss": 0.7868,
      "num_input_tokens_seen": 73488616,
      "step": 126575
    },
    {
      "epoch": 18.853142686922848,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.000299568870224029,
      "loss": 0.7863,
      "num_input_tokens_seen": 73491560,
      "step": 126580
    },
    {
      "epoch": 18.853887399463808,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0002991812944427835,
      "loss": 0.8256,
      "num_input_tokens_seen": 73494504,
      "step": 126585
    },
    {
      "epoch": 18.854632112004765,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.00029879396701692926,
      "loss": 0.7956,
      "num_input_tokens_seen": 73497672,
      "step": 126590
    },
    {
      "epoch": 18.855376824545726,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.0002984068879530094,
      "loss": 0.7871,
      "num_input_tokens_seen": 73500552,
      "step": 126595
    },
    {
      "epoch": 18.856121537086686,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0002980200572575636,
      "loss": 0.7844,
      "num_input_tokens_seen": 73503592,
      "step": 126600
    },
    {
      "epoch": 18.856866249627643,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.00029763347493712674,
      "loss": 0.7885,
      "num_input_tokens_seen": 73506664,
      "step": 126605
    },
    {
      "epoch": 18.857610962168604,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0002972471409982269,
      "loss": 0.8023,
      "num_input_tokens_seen": 73509480,
      "step": 126610
    },
    {
      "epoch": 18.85835567470956,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00029686105544739715,
      "loss": 0.7949,
      "num_input_tokens_seen": 73512392,
      "step": 126615
    },
    {
      "epoch": 18.85910038725052,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.00029647521829115396,
      "loss": 0.8022,
      "num_input_tokens_seen": 73515208,
      "step": 126620
    },
    {
      "epoch": 18.85984509979148,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0002960896295360188,
      "loss": 0.7982,
      "num_input_tokens_seen": 73518376,
      "step": 126625
    },
    {
      "epoch": 18.86058981233244,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0002957042891885048,
      "loss": 0.7984,
      "num_input_tokens_seen": 73521096,
      "step": 126630
    },
    {
      "epoch": 18.8613345248734,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0002953191972551233,
      "loss": 0.8115,
      "num_input_tokens_seen": 73524008,
      "step": 126635
    },
    {
      "epoch": 18.86207923741436,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0002949343537423776,
      "loss": 0.8071,
      "num_input_tokens_seen": 73526792,
      "step": 126640
    },
    {
      "epoch": 18.862823949955317,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.00029454975865676915,
      "loss": 0.8048,
      "num_input_tokens_seen": 73529352,
      "step": 126645
    },
    {
      "epoch": 18.863568662496277,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0002941654120047976,
      "loss": 0.8164,
      "num_input_tokens_seen": 73532392,
      "step": 126650
    },
    {
      "epoch": 18.864313375037234,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0002937813137929546,
      "loss": 0.8093,
      "num_input_tokens_seen": 73535400,
      "step": 126655
    },
    {
      "epoch": 18.865058087578195,
      "grad_norm": 0.052734375,
      "learning_rate": 0.0002933974640277315,
      "loss": 0.792,
      "num_input_tokens_seen": 73538536,
      "step": 126660
    },
    {
      "epoch": 18.865802800119155,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.0002930138627156098,
      "loss": 0.7997,
      "num_input_tokens_seen": 73541288,
      "step": 126665
    },
    {
      "epoch": 18.866547512660112,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.00029263050986307103,
      "loss": 0.8018,
      "num_input_tokens_seen": 73544232,
      "step": 126670
    },
    {
      "epoch": 18.867292225201073,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0002922474054765933,
      "loss": 0.7913,
      "num_input_tokens_seen": 73547176,
      "step": 126675
    },
    {
      "epoch": 18.868036937742033,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0002918645495626448,
      "loss": 0.802,
      "num_input_tokens_seen": 73550056,
      "step": 126680
    },
    {
      "epoch": 18.86878165028299,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.00029148194212769873,
      "loss": 0.7722,
      "num_input_tokens_seen": 73553096,
      "step": 126685
    },
    {
      "epoch": 18.86952636282395,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0002910995831782148,
      "loss": 0.8082,
      "num_input_tokens_seen": 73556200,
      "step": 126690
    },
    {
      "epoch": 18.870271075364908,
      "grad_norm": 0.044921875,
      "learning_rate": 0.00029071747272065295,
      "loss": 0.7852,
      "num_input_tokens_seen": 73558824,
      "step": 126695
    },
    {
      "epoch": 18.87101578790587,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0002903356107614713,
      "loss": 0.8081,
      "num_input_tokens_seen": 73561512,
      "step": 126700
    },
    {
      "epoch": 18.87176050044683,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.00028995399730711967,
      "loss": 0.8195,
      "num_input_tokens_seen": 73564648,
      "step": 126705
    },
    {
      "epoch": 18.872505212987786,
      "grad_norm": 0.0277099609375,
      "learning_rate": 0.00028957263236404463,
      "loss": 0.7909,
      "num_input_tokens_seen": 73567368,
      "step": 126710
    },
    {
      "epoch": 18.873249925528746,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0002891915159386876,
      "loss": 0.7882,
      "num_input_tokens_seen": 73570216,
      "step": 126715
    },
    {
      "epoch": 18.873994638069703,
      "grad_norm": 0.0294189453125,
      "learning_rate": 0.00028881064803748855,
      "loss": 0.8037,
      "num_input_tokens_seen": 73572968,
      "step": 126720
    },
    {
      "epoch": 18.874739350610664,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0002884300286668823,
      "loss": 0.8001,
      "num_input_tokens_seen": 73576008,
      "step": 126725
    },
    {
      "epoch": 18.875484063151625,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.0002880496578332986,
      "loss": 0.7974,
      "num_input_tokens_seen": 73578856,
      "step": 126730
    },
    {
      "epoch": 18.87622877569258,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.00028766953554316244,
      "loss": 0.7833,
      "num_input_tokens_seen": 73581704,
      "step": 126735
    },
    {
      "epoch": 18.876973488233542,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0002872896618028969,
      "loss": 0.7935,
      "num_input_tokens_seen": 73584776,
      "step": 126740
    },
    {
      "epoch": 18.877718200774503,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.00028691003661892033,
      "loss": 0.8231,
      "num_input_tokens_seen": 73588008,
      "step": 126745
    },
    {
      "epoch": 18.87846291331546,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.00028653065999764417,
      "loss": 0.8121,
      "num_input_tokens_seen": 73591048,
      "step": 126750
    },
    {
      "epoch": 18.87920762585642,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.00028615153194547823,
      "loss": 0.7924,
      "num_input_tokens_seen": 73594248,
      "step": 126755
    },
    {
      "epoch": 18.879952338397377,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0002857726524688292,
      "loss": 0.7688,
      "num_input_tokens_seen": 73597096,
      "step": 126760
    },
    {
      "epoch": 18.880697050938338,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.0002853940215740952,
      "loss": 0.7901,
      "num_input_tokens_seen": 73600168,
      "step": 126765
    },
    {
      "epoch": 18.881441763479298,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.0002850156392676728,
      "loss": 0.8059,
      "num_input_tokens_seen": 73603144,
      "step": 126770
    },
    {
      "epoch": 18.882186476020255,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.00028463750555595687,
      "loss": 0.8016,
      "num_input_tokens_seen": 73606248,
      "step": 126775
    },
    {
      "epoch": 18.882931188561216,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.0002842596204453357,
      "loss": 0.7935,
      "num_input_tokens_seen": 73609256,
      "step": 126780
    },
    {
      "epoch": 18.883675901102176,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0002838819839421891,
      "loss": 0.8033,
      "num_input_tokens_seen": 73611816,
      "step": 126785
    },
    {
      "epoch": 18.884420613643133,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.00028350459605290036,
      "loss": 0.7873,
      "num_input_tokens_seen": 73614696,
      "step": 126790
    },
    {
      "epoch": 18.885165326184094,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.0002831274567838443,
      "loss": 0.8102,
      "num_input_tokens_seen": 73617736,
      "step": 126795
    },
    {
      "epoch": 18.88591003872505,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00028275056614139257,
      "loss": 0.8079,
      "num_input_tokens_seen": 73620616,
      "step": 126800
    },
    {
      "epoch": 18.88665475126601,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.00028237392413191177,
      "loss": 0.8248,
      "num_input_tokens_seen": 73624040,
      "step": 126805
    },
    {
      "epoch": 18.88739946380697,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.00028199753076176503,
      "loss": 0.783,
      "num_input_tokens_seen": 73627176,
      "step": 126810
    },
    {
      "epoch": 18.88814417634793,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0002816213860373107,
      "loss": 0.7986,
      "num_input_tokens_seen": 73630280,
      "step": 126815
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0002812454899649069,
      "loss": 0.7985,
      "num_input_tokens_seen": 73632968,
      "step": 126820
    },
    {
      "epoch": 18.88963360142985,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00028086984255089873,
      "loss": 0.8088,
      "num_input_tokens_seen": 73636008,
      "step": 126825
    },
    {
      "epoch": 18.890378313970807,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.000280494443801636,
      "loss": 0.8031,
      "num_input_tokens_seen": 73638664,
      "step": 126830
    },
    {
      "epoch": 18.891123026511767,
      "grad_norm": 0.0308837890625,
      "learning_rate": 0.00028011929372345865,
      "loss": 0.7865,
      "num_input_tokens_seen": 73641608,
      "step": 126835
    },
    {
      "epoch": 18.891867739052724,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.000279744392322705,
      "loss": 0.8057,
      "num_input_tokens_seen": 73644392,
      "step": 126840
    },
    {
      "epoch": 18.892612451593685,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0002793697396057082,
      "loss": 0.7952,
      "num_input_tokens_seen": 73647400,
      "step": 126845
    },
    {
      "epoch": 18.893357164134645,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.0002789953355787983,
      "loss": 0.7853,
      "num_input_tokens_seen": 73650152,
      "step": 126850
    },
    {
      "epoch": 18.894101876675602,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.00027862118024830185,
      "loss": 0.7913,
      "num_input_tokens_seen": 73653032,
      "step": 126855
    },
    {
      "epoch": 18.894846589216563,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.0002782472736205388,
      "loss": 0.7906,
      "num_input_tokens_seen": 73656264,
      "step": 126860
    },
    {
      "epoch": 18.89559130175752,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0002778736157018241,
      "loss": 0.7777,
      "num_input_tokens_seen": 73658952,
      "step": 126865
    },
    {
      "epoch": 18.89633601429848,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.00027750020649847107,
      "loss": 0.7912,
      "num_input_tokens_seen": 73661896,
      "step": 126870
    },
    {
      "epoch": 18.89708072683944,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.0002771270460167896,
      "loss": 0.8087,
      "num_input_tokens_seen": 73664648,
      "step": 126875
    },
    {
      "epoch": 18.897825439380398,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.0002767541342630847,
      "loss": 0.7907,
      "num_input_tokens_seen": 73667368,
      "step": 126880
    },
    {
      "epoch": 18.89857015192136,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0002763814712436513,
      "loss": 0.8094,
      "num_input_tokens_seen": 73670056,
      "step": 126885
    },
    {
      "epoch": 18.89931486446232,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00027600905696479103,
      "loss": 0.8097,
      "num_input_tokens_seen": 73672744,
      "step": 126890
    },
    {
      "epoch": 18.900059577003276,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.0002756368914327939,
      "loss": 0.8134,
      "num_input_tokens_seen": 73675496,
      "step": 126895
    },
    {
      "epoch": 18.900804289544237,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.00027526497465394484,
      "loss": 0.8052,
      "num_input_tokens_seen": 73678600,
      "step": 126900
    },
    {
      "epoch": 18.901549002085194,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.0002748933066345288,
      "loss": 0.8029,
      "num_input_tokens_seen": 73681224,
      "step": 126905
    },
    {
      "epoch": 18.902293714626154,
      "grad_norm": 0.030029296875,
      "learning_rate": 0.0002745218873808225,
      "loss": 0.8018,
      "num_input_tokens_seen": 73684328,
      "step": 126910
    },
    {
      "epoch": 18.903038427167115,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.00027415071689910586,
      "loss": 0.7701,
      "num_input_tokens_seen": 73686920,
      "step": 126915
    },
    {
      "epoch": 18.90378313970807,
      "grad_norm": 0.5,
      "learning_rate": 0.0002737797951956422,
      "loss": 0.8296,
      "num_input_tokens_seen": 73689704,
      "step": 126920
    },
    {
      "epoch": 18.904527852249032,
      "grad_norm": 0.0281982421875,
      "learning_rate": 0.0002734091222767049,
      "loss": 0.7883,
      "num_input_tokens_seen": 73692552,
      "step": 126925
    },
    {
      "epoch": 18.905272564789993,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0002730386981485522,
      "loss": 0.7862,
      "num_input_tokens_seen": 73695624,
      "step": 126930
    },
    {
      "epoch": 18.90601727733095,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.00027266852281744244,
      "loss": 0.8164,
      "num_input_tokens_seen": 73698280,
      "step": 126935
    },
    {
      "epoch": 18.90676198987191,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.0002722985962896307,
      "loss": 0.782,
      "num_input_tokens_seen": 73701064,
      "step": 126940
    },
    {
      "epoch": 18.907506702412867,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.00027192891857136524,
      "loss": 0.7874,
      "num_input_tokens_seen": 73703880,
      "step": 126945
    },
    {
      "epoch": 18.908251414953828,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0002715594896688911,
      "loss": 0.7978,
      "num_input_tokens_seen": 73706792,
      "step": 126950
    },
    {
      "epoch": 18.908996127494788,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0002711903095884499,
      "loss": 0.8044,
      "num_input_tokens_seen": 73709672,
      "step": 126955
    },
    {
      "epoch": 18.909740840035745,
      "grad_norm": 0.07275390625,
      "learning_rate": 0.00027082137833628003,
      "loss": 0.8057,
      "num_input_tokens_seen": 73712616,
      "step": 126960
    },
    {
      "epoch": 18.910485552576706,
      "grad_norm": 0.3515625,
      "learning_rate": 0.0002704526959186132,
      "loss": 0.8275,
      "num_input_tokens_seen": 73715496,
      "step": 126965
    },
    {
      "epoch": 18.911230265117666,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.0002700842623416777,
      "loss": 0.7972,
      "num_input_tokens_seen": 73718248,
      "step": 126970
    },
    {
      "epoch": 18.911974977658623,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.00026971607761169526,
      "loss": 0.7864,
      "num_input_tokens_seen": 73721320,
      "step": 126975
    },
    {
      "epoch": 18.912719690199584,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00026934814173489084,
      "loss": 0.7931,
      "num_input_tokens_seen": 73724136,
      "step": 126980
    },
    {
      "epoch": 18.91346440274054,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00026898045471747785,
      "loss": 0.7924,
      "num_input_tokens_seen": 73726952,
      "step": 126985
    },
    {
      "epoch": 18.9142091152815,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.0002686130165656664,
      "loss": 0.8341,
      "num_input_tokens_seen": 73730184,
      "step": 126990
    },
    {
      "epoch": 18.914953827822462,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00026824582728566805,
      "loss": 0.8024,
      "num_input_tokens_seen": 73733224,
      "step": 126995
    },
    {
      "epoch": 18.91569854036342,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.0002678788868836812,
      "loss": 0.8108,
      "num_input_tokens_seen": 73736040,
      "step": 127000
    },
    {
      "epoch": 18.91644325290438,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.0002675121953659093,
      "loss": 0.7929,
      "num_input_tokens_seen": 73739016,
      "step": 127005
    },
    {
      "epoch": 18.917187965445336,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.000267145752738544,
      "loss": 0.8181,
      "num_input_tokens_seen": 73741672,
      "step": 127010
    },
    {
      "epoch": 18.917932677986297,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00026677955900778036,
      "loss": 0.7925,
      "num_input_tokens_seen": 73744904,
      "step": 127015
    },
    {
      "epoch": 18.918677390527257,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00026641361417980013,
      "loss": 0.7952,
      "num_input_tokens_seen": 73747560,
      "step": 127020
    },
    {
      "epoch": 18.919422103068214,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.00026604791826078663,
      "loss": 0.7722,
      "num_input_tokens_seen": 73750472,
      "step": 127025
    },
    {
      "epoch": 18.920166815609175,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00026568247125691836,
      "loss": 0.7557,
      "num_input_tokens_seen": 73753128,
      "step": 127030
    },
    {
      "epoch": 18.920911528150135,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0002653172731743719,
      "loss": 0.7947,
      "num_input_tokens_seen": 73756296,
      "step": 127035
    },
    {
      "epoch": 18.921656240691092,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00026495232401931246,
      "loss": 0.8081,
      "num_input_tokens_seen": 73760552,
      "step": 127040
    },
    {
      "epoch": 18.922400953232053,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.00026458762379790666,
      "loss": 0.7978,
      "num_input_tokens_seen": 73763528,
      "step": 127045
    },
    {
      "epoch": 18.92314566577301,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.00026422317251631797,
      "loss": 0.8036,
      "num_input_tokens_seen": 73766600,
      "step": 127050
    },
    {
      "epoch": 18.92389037831397,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00026385897018070146,
      "loss": 0.783,
      "num_input_tokens_seen": 73769416,
      "step": 127055
    },
    {
      "epoch": 18.92463509085493,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.0002634950167972105,
      "loss": 0.8043,
      "num_input_tokens_seen": 73772232,
      "step": 127060
    },
    {
      "epoch": 18.925379803395888,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00026313131237199524,
      "loss": 0.7848,
      "num_input_tokens_seen": 73775272,
      "step": 127065
    },
    {
      "epoch": 18.92612451593685,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0002627678569111974,
      "loss": 0.7959,
      "num_input_tokens_seen": 73778376,
      "step": 127070
    },
    {
      "epoch": 18.92686922847781,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.00026240465042095873,
      "loss": 0.7917,
      "num_input_tokens_seen": 73781320,
      "step": 127075
    },
    {
      "epoch": 18.927613941018766,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0002620416929074126,
      "loss": 0.7934,
      "num_input_tokens_seen": 73784136,
      "step": 127080
    },
    {
      "epoch": 18.928358653559727,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.0002616789843766959,
      "loss": 0.8048,
      "num_input_tokens_seen": 73787272,
      "step": 127085
    },
    {
      "epoch": 18.929103366100684,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.00026131652483493196,
      "loss": 0.8005,
      "num_input_tokens_seen": 73790088,
      "step": 127090
    },
    {
      "epoch": 18.929848078641644,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.00026095431428824754,
      "loss": 0.7925,
      "num_input_tokens_seen": 73792904,
      "step": 127095
    },
    {
      "epoch": 18.930592791182605,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.00026059235274275783,
      "loss": 0.7958,
      "num_input_tokens_seen": 73795976,
      "step": 127100
    },
    {
      "epoch": 18.93133750372356,
      "grad_norm": 0.08984375,
      "learning_rate": 0.00026023064020457953,
      "loss": 0.7909,
      "num_input_tokens_seen": 73799080,
      "step": 127105
    },
    {
      "epoch": 18.932082216264522,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.0002598691766798261,
      "loss": 0.806,
      "num_input_tokens_seen": 73801992,
      "step": 127110
    },
    {
      "epoch": 18.932826928805483,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.00025950796217459935,
      "loss": 0.79,
      "num_input_tokens_seen": 73804872,
      "step": 127115
    },
    {
      "epoch": 18.93357164134644,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.0002591469966950044,
      "loss": 0.7831,
      "num_input_tokens_seen": 73807720,
      "step": 127120
    },
    {
      "epoch": 18.9343163538874,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0002587862802471397,
      "loss": 0.791,
      "num_input_tokens_seen": 73810568,
      "step": 127125
    },
    {
      "epoch": 18.935061066428357,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.00025842581283709863,
      "loss": 0.8003,
      "num_input_tokens_seen": 73813352,
      "step": 127130
    },
    {
      "epoch": 18.935805778969318,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.00025806559447096977,
      "loss": 0.7739,
      "num_input_tokens_seen": 73816456,
      "step": 127135
    },
    {
      "epoch": 18.93655049151028,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00025770562515483984,
      "loss": 0.7862,
      "num_input_tokens_seen": 73819656,
      "step": 127140
    },
    {
      "epoch": 18.937295204051235,
      "grad_norm": 0.047119140625,
      "learning_rate": 0.00025734590489479237,
      "loss": 0.7925,
      "num_input_tokens_seen": 73822440,
      "step": 127145
    },
    {
      "epoch": 18.938039916592196,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0002569864336968991,
      "loss": 0.8108,
      "num_input_tokens_seen": 73825256,
      "step": 127150
    },
    {
      "epoch": 18.938784629133156,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.00025662721156723687,
      "loss": 0.7948,
      "num_input_tokens_seen": 73828008,
      "step": 127155
    },
    {
      "epoch": 18.939529341674113,
      "grad_norm": 0.05224609375,
      "learning_rate": 0.0002562682385118742,
      "loss": 0.7829,
      "num_input_tokens_seen": 73831080,
      "step": 127160
    },
    {
      "epoch": 18.940274054215074,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.0002559095145368745,
      "loss": 0.8349,
      "num_input_tokens_seen": 73834344,
      "step": 127165
    },
    {
      "epoch": 18.94101876675603,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00025555103964829796,
      "loss": 0.781,
      "num_input_tokens_seen": 73837352,
      "step": 127170
    },
    {
      "epoch": 18.94176347929699,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.00025519281385220135,
      "loss": 0.8127,
      "num_input_tokens_seen": 73840616,
      "step": 127175
    },
    {
      "epoch": 18.942508191837952,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00025483483715463495,
      "loss": 0.8058,
      "num_input_tokens_seen": 73843240,
      "step": 127180
    },
    {
      "epoch": 18.94325290437891,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0002544771095616488,
      "loss": 0.805,
      "num_input_tokens_seen": 73846280,
      "step": 127185
    },
    {
      "epoch": 18.94399761691987,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.00025411963107928483,
      "loss": 0.7898,
      "num_input_tokens_seen": 73849320,
      "step": 127190
    },
    {
      "epoch": 18.94474232946083,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.00025376240171358144,
      "loss": 0.8005,
      "num_input_tokens_seen": 73852200,
      "step": 127195
    },
    {
      "epoch": 18.945487042001787,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.0002534054214705772,
      "loss": 0.8153,
      "num_input_tokens_seen": 73855368,
      "step": 127200
    },
    {
      "epoch": 18.946231754542747,
      "grad_norm": 0.03515625,
      "learning_rate": 0.00025304869035629895,
      "loss": 0.7757,
      "num_input_tokens_seen": 73858184,
      "step": 127205
    },
    {
      "epoch": 18.946976467083704,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0002526922083767769,
      "loss": 0.7919,
      "num_input_tokens_seen": 73861000,
      "step": 127210
    },
    {
      "epoch": 18.947721179624665,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.00025233597553802944,
      "loss": 0.8072,
      "num_input_tokens_seen": 73863848,
      "step": 127215
    },
    {
      "epoch": 18.948465892165625,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0002519799918460769,
      "loss": 0.7936,
      "num_input_tokens_seen": 73866728,
      "step": 127220
    },
    {
      "epoch": 18.949210604706582,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0002516242573069344,
      "loss": 0.8137,
      "num_input_tokens_seen": 73869384,
      "step": 127225
    },
    {
      "epoch": 18.949955317247543,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.00025126877192661044,
      "loss": 0.8095,
      "num_input_tokens_seen": 73872168,
      "step": 127230
    },
    {
      "epoch": 18.9507000297885,
      "grad_norm": 0.02197265625,
      "learning_rate": 0.0002509135357111103,
      "loss": 0.8048,
      "num_input_tokens_seen": 73874856,
      "step": 127235
    },
    {
      "epoch": 18.95144474232946,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.00025055854866643575,
      "loss": 0.8058,
      "num_input_tokens_seen": 73877960,
      "step": 127240
    },
    {
      "epoch": 18.95218945487042,
      "grad_norm": 0.03515625,
      "learning_rate": 0.00025020381079858377,
      "loss": 0.8076,
      "num_input_tokens_seen": 73880808,
      "step": 127245
    },
    {
      "epoch": 18.952934167411378,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.00024984932211354614,
      "loss": 0.793,
      "num_input_tokens_seen": 73883880,
      "step": 127250
    },
    {
      "epoch": 18.95367887995234,
      "grad_norm": 0.052734375,
      "learning_rate": 0.00024949508261731476,
      "loss": 0.7896,
      "num_input_tokens_seen": 73886792,
      "step": 127255
    },
    {
      "epoch": 18.9544235924933,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0002491410923158699,
      "loss": 0.7838,
      "num_input_tokens_seen": 73889352,
      "step": 127260
    },
    {
      "epoch": 18.955168305034256,
      "grad_norm": 0.078125,
      "learning_rate": 0.00024878735121519667,
      "loss": 0.7962,
      "num_input_tokens_seen": 73892296,
      "step": 127265
    },
    {
      "epoch": 18.955913017575217,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.0002484338593212687,
      "loss": 0.785,
      "num_input_tokens_seen": 73894888,
      "step": 127270
    },
    {
      "epoch": 18.956657730116174,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0002480806166400545,
      "loss": 0.786,
      "num_input_tokens_seen": 73897704,
      "step": 127275
    },
    {
      "epoch": 18.957402442657134,
      "grad_norm": 0.046875,
      "learning_rate": 0.0002477276231775294,
      "loss": 0.7774,
      "num_input_tokens_seen": 73900680,
      "step": 127280
    },
    {
      "epoch": 18.958147155198095,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00024737487893964846,
      "loss": 0.794,
      "num_input_tokens_seen": 73903688,
      "step": 127285
    },
    {
      "epoch": 18.95889186773905,
      "grad_norm": 0.0252685546875,
      "learning_rate": 0.00024702238393237874,
      "loss": 0.8113,
      "num_input_tokens_seen": 73906280,
      "step": 127290
    },
    {
      "epoch": 18.959636580280012,
      "grad_norm": 0.06298828125,
      "learning_rate": 0.00024667013816167036,
      "loss": 0.7932,
      "num_input_tokens_seen": 73909288,
      "step": 127295
    },
    {
      "epoch": 18.960381292820973,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00024631814163347533,
      "loss": 0.8101,
      "num_input_tokens_seen": 73912200,
      "step": 127300
    },
    {
      "epoch": 18.96112600536193,
      "grad_norm": 0.052734375,
      "learning_rate": 0.00024596639435374044,
      "loss": 0.8089,
      "num_input_tokens_seen": 73914760,
      "step": 127305
    },
    {
      "epoch": 18.96187071790289,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00024561489632840604,
      "loss": 0.7961,
      "num_input_tokens_seen": 73917576,
      "step": 127310
    },
    {
      "epoch": 18.962615430443847,
      "grad_norm": 0.037109375,
      "learning_rate": 0.000245263647563414,
      "loss": 0.7842,
      "num_input_tokens_seen": 73920456,
      "step": 127315
    },
    {
      "epoch": 18.963360142984808,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.00024491264806469627,
      "loss": 0.8098,
      "num_input_tokens_seen": 73923720,
      "step": 127320
    },
    {
      "epoch": 18.96410485552577,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.00024456189783818305,
      "loss": 0.7808,
      "num_input_tokens_seen": 73926504,
      "step": 127325
    },
    {
      "epoch": 18.964849568066725,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.00024421139688979965,
      "loss": 0.8053,
      "num_input_tokens_seen": 73929384,
      "step": 127330
    },
    {
      "epoch": 18.965594280607686,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.00024386114522546796,
      "loss": 0.7669,
      "num_input_tokens_seen": 73932328,
      "step": 127335
    },
    {
      "epoch": 18.966338993148646,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00024351114285110496,
      "loss": 0.7928,
      "num_input_tokens_seen": 73935176,
      "step": 127340
    },
    {
      "epoch": 18.967083705689603,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.00024316138977262258,
      "loss": 0.8033,
      "num_input_tokens_seen": 73937896,
      "step": 127345
    },
    {
      "epoch": 18.967828418230564,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.00024281188599593106,
      "loss": 0.7876,
      "num_input_tokens_seen": 73941064,
      "step": 127350
    },
    {
      "epoch": 18.96857313077152,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.00024246263152693403,
      "loss": 0.7903,
      "num_input_tokens_seen": 73943880,
      "step": 127355
    },
    {
      "epoch": 18.96931784331248,
      "grad_norm": 0.029052734375,
      "learning_rate": 0.00024211362637153175,
      "loss": 0.8012,
      "num_input_tokens_seen": 73946728,
      "step": 127360
    },
    {
      "epoch": 18.970062555853442,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.00024176487053562123,
      "loss": 0.8059,
      "num_input_tokens_seen": 73949672,
      "step": 127365
    },
    {
      "epoch": 18.9708072683944,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.0002414163640250927,
      "loss": 0.7799,
      "num_input_tokens_seen": 73952264,
      "step": 127370
    },
    {
      "epoch": 18.97155198093536,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0002410681068458348,
      "loss": 0.8156,
      "num_input_tokens_seen": 73955240,
      "step": 127375
    },
    {
      "epoch": 18.972296693476316,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.00024072009900373114,
      "loss": 0.8159,
      "num_input_tokens_seen": 73958120,
      "step": 127380
    },
    {
      "epoch": 18.973041406017277,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.00024037234050466037,
      "loss": 0.8091,
      "num_input_tokens_seen": 73961128,
      "step": 127385
    },
    {
      "epoch": 18.973786118558237,
      "grad_norm": 0.042236328125,
      "learning_rate": 0.0002400248313544978,
      "loss": 0.7899,
      "num_input_tokens_seen": 73964040,
      "step": 127390
    },
    {
      "epoch": 18.974530831099194,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.0002396775715591154,
      "loss": 0.8069,
      "num_input_tokens_seen": 73966888,
      "step": 127395
    },
    {
      "epoch": 18.975275543640155,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.00023933056112437844,
      "loss": 0.8041,
      "num_input_tokens_seen": 73969960,
      "step": 127400
    },
    {
      "epoch": 18.976020256181116,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.00023898380005614895,
      "loss": 0.784,
      "num_input_tokens_seen": 73972648,
      "step": 127405
    },
    {
      "epoch": 18.976764968722073,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.0002386372883602855,
      "loss": 0.7811,
      "num_input_tokens_seen": 73975496,
      "step": 127410
    },
    {
      "epoch": 18.977509681263033,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0002382910260426435,
      "loss": 0.8081,
      "num_input_tokens_seen": 73978440,
      "step": 127415
    },
    {
      "epoch": 18.97825439380399,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.0002379450131090699,
      "loss": 0.8153,
      "num_input_tokens_seen": 73981224,
      "step": 127420
    },
    {
      "epoch": 18.97899910634495,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.00023759924956541332,
      "loss": 0.7952,
      "num_input_tokens_seen": 73984456,
      "step": 127425
    },
    {
      "epoch": 18.97974381888591,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.00023725373541751414,
      "loss": 0.7761,
      "num_input_tokens_seen": 73987304,
      "step": 127430
    },
    {
      "epoch": 18.980488531426868,
      "grad_norm": 0.044921875,
      "learning_rate": 0.000236908470671206,
      "loss": 0.7894,
      "num_input_tokens_seen": 73990056,
      "step": 127435
    },
    {
      "epoch": 18.98123324396783,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.00023656345533232758,
      "loss": 0.798,
      "num_input_tokens_seen": 73993416,
      "step": 127440
    },
    {
      "epoch": 18.98197795650879,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.00023621868940670252,
      "loss": 0.8127,
      "num_input_tokens_seen": 73996104,
      "step": 127445
    },
    {
      "epoch": 18.982722669049746,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.00023587417290015953,
      "loss": 0.7754,
      "num_input_tokens_seen": 73998600,
      "step": 127450
    },
    {
      "epoch": 18.983467381590707,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.00023552990581851563,
      "loss": 0.7759,
      "num_input_tokens_seen": 74001256,
      "step": 127455
    },
    {
      "epoch": 18.984212094131664,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.00023518588816758777,
      "loss": 0.7966,
      "num_input_tokens_seen": 74004104,
      "step": 127460
    },
    {
      "epoch": 18.984956806672624,
      "grad_norm": 0.0390625,
      "learning_rate": 0.00023484211995318971,
      "loss": 0.7916,
      "num_input_tokens_seen": 74007208,
      "step": 127465
    },
    {
      "epoch": 18.985701519213585,
      "grad_norm": 0.0390625,
      "learning_rate": 0.00023449860118112507,
      "loss": 0.7937,
      "num_input_tokens_seen": 74009864,
      "step": 127470
    },
    {
      "epoch": 18.98644623175454,
      "grad_norm": 0.0322265625,
      "learning_rate": 0.00023415533185720092,
      "loss": 0.7963,
      "num_input_tokens_seen": 74013224,
      "step": 127475
    },
    {
      "epoch": 18.987190944295502,
      "grad_norm": 0.041015625,
      "learning_rate": 0.0002338123119872143,
      "loss": 0.7916,
      "num_input_tokens_seen": 74016264,
      "step": 127480
    },
    {
      "epoch": 18.987935656836463,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.0002334695415769622,
      "loss": 0.7777,
      "num_input_tokens_seen": 74019496,
      "step": 127485
    },
    {
      "epoch": 18.98868036937742,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.00023312702063223333,
      "loss": 0.8087,
      "num_input_tokens_seen": 74022120,
      "step": 127490
    },
    {
      "epoch": 18.98942508191838,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.00023278474915881475,
      "loss": 0.7816,
      "num_input_tokens_seen": 74024936,
      "step": 127495
    },
    {
      "epoch": 18.990169794459337,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.00023244272716249013,
      "loss": 0.7877,
      "num_input_tokens_seen": 74028008,
      "step": 127500
    },
    {
      "epoch": 18.990914507000298,
      "grad_norm": 0.0311279296875,
      "learning_rate": 0.00023210095464903656,
      "loss": 0.7967,
      "num_input_tokens_seen": 74030888,
      "step": 127505
    },
    {
      "epoch": 18.99165921954126,
      "grad_norm": 0.04150390625,
      "learning_rate": 0.00023175943162422773,
      "loss": 0.8136,
      "num_input_tokens_seen": 74033480,
      "step": 127510
    },
    {
      "epoch": 18.992403932082215,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.00023141815809383235,
      "loss": 0.7946,
      "num_input_tokens_seen": 74036424,
      "step": 127515
    },
    {
      "epoch": 18.993148644623176,
      "grad_norm": 0.0859375,
      "learning_rate": 0.00023107713406361752,
      "loss": 0.8115,
      "num_input_tokens_seen": 74039464,
      "step": 127520
    },
    {
      "epoch": 18.993893357164133,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.00023073635953934355,
      "loss": 0.7958,
      "num_input_tokens_seen": 74042056,
      "step": 127525
    },
    {
      "epoch": 18.994638069705093,
      "grad_norm": 0.059326171875,
      "learning_rate": 0.00023039583452676924,
      "loss": 0.8182,
      "num_input_tokens_seen": 74045096,
      "step": 127530
    },
    {
      "epoch": 18.995382782246054,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.00023005555903164498,
      "loss": 0.8084,
      "num_input_tokens_seen": 74048168,
      "step": 127535
    },
    {
      "epoch": 18.99612749478701,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0002297155330597228,
      "loss": 0.8185,
      "num_input_tokens_seen": 74050984,
      "step": 127540
    },
    {
      "epoch": 18.99687220732797,
      "grad_norm": 0.05859375,
      "learning_rate": 0.00022937575661674314,
      "loss": 0.7743,
      "num_input_tokens_seen": 74054344,
      "step": 127545
    },
    {
      "epoch": 18.997616919868932,
      "grad_norm": 0.0260009765625,
      "learning_rate": 0.00022903622970844804,
      "loss": 0.797,
      "num_input_tokens_seen": 74057288,
      "step": 127550
    },
    {
      "epoch": 18.99836163240989,
      "grad_norm": 0.34375,
      "learning_rate": 0.00022869695234057296,
      "loss": 0.8024,
      "num_input_tokens_seen": 74060264,
      "step": 127555
    },
    {
      "epoch": 18.99910634495085,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0002283579245188516,
      "loss": 0.8231,
      "num_input_tokens_seen": 74063336,
      "step": 127560
    },
    {
      "epoch": 18.999851057491806,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.00022801914624900942,
      "loss": 0.8225,
      "num_input_tokens_seen": 74065864,
      "step": 127565
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.799495279788971,
      "eval_runtime": 70.6051,
      "eval_samples_per_second": 42.263,
      "eval_steps_per_second": 10.566,
      "num_input_tokens_seen": 74065984,
      "step": 127566
    },
    {
      "epoch": 19.000595770032767,
      "grad_norm": 0.061279296875,
      "learning_rate": 0.00022768061753676848,
      "loss": 0.8254,
      "num_input_tokens_seen": 74068512,
      "step": 127570
    },
    {
      "epoch": 19.001340482573728,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0002273423383878509,
      "loss": 0.8153,
      "num_input_tokens_seen": 74071232,
      "step": 127575
    },
    {
      "epoch": 19.002085195114685,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00022700430880796872,
      "loss": 0.808,
      "num_input_tokens_seen": 74074112,
      "step": 127580
    },
    {
      "epoch": 19.002829907655645,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00022666652880283576,
      "loss": 0.7901,
      "num_input_tokens_seen": 74076992,
      "step": 127585
    },
    {
      "epoch": 19.003574620196606,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.00022632899837815744,
      "loss": 0.8185,
      "num_input_tokens_seen": 74080160,
      "step": 127590
    },
    {
      "epoch": 19.004319332737563,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.0002259917175396342,
      "loss": 0.7981,
      "num_input_tokens_seen": 74083264,
      "step": 127595
    },
    {
      "epoch": 19.005064045278523,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0002256546862929648,
      "loss": 0.7958,
      "num_input_tokens_seen": 74086176,
      "step": 127600
    },
    {
      "epoch": 19.00580875781948,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00022531790464384304,
      "loss": 0.7837,
      "num_input_tokens_seen": 74089184,
      "step": 127605
    },
    {
      "epoch": 19.00655347036044,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.00022498137259796102,
      "loss": 0.8297,
      "num_input_tokens_seen": 74092096,
      "step": 127610
    },
    {
      "epoch": 19.0072981829014,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00022464509016100086,
      "loss": 0.8101,
      "num_input_tokens_seen": 74094976,
      "step": 127615
    },
    {
      "epoch": 19.008042895442358,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.0002243090573386447,
      "loss": 0.7928,
      "num_input_tokens_seen": 74097856,
      "step": 127620
    },
    {
      "epoch": 19.00878760798332,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0002239732741365713,
      "loss": 0.7969,
      "num_input_tokens_seen": 74100768,
      "step": 127625
    },
    {
      "epoch": 19.00953232052428,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.00022363774056045115,
      "loss": 0.8086,
      "num_input_tokens_seen": 74103488,
      "step": 127630
    },
    {
      "epoch": 19.010277033065236,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.00022330245661595304,
      "loss": 0.8003,
      "num_input_tokens_seen": 74106400,
      "step": 127635
    },
    {
      "epoch": 19.011021745606197,
      "grad_norm": 0.056640625,
      "learning_rate": 0.00022296742230874076,
      "loss": 0.784,
      "num_input_tokens_seen": 74109184,
      "step": 127640
    },
    {
      "epoch": 19.011766458147154,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.00022263263764447648,
      "loss": 0.8033,
      "num_input_tokens_seen": 74112096,
      "step": 127645
    },
    {
      "epoch": 19.012511170688114,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.00022229810262881233,
      "loss": 0.7927,
      "num_input_tokens_seen": 74115040,
      "step": 127650
    },
    {
      "epoch": 19.013255883229075,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.00022196381726740543,
      "loss": 0.7871,
      "num_input_tokens_seen": 74118176,
      "step": 127655
    },
    {
      "epoch": 19.01400059577003,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.00022162978156589795,
      "loss": 0.8108,
      "num_input_tokens_seen": 74120800,
      "step": 127660
    },
    {
      "epoch": 19.014745308310992,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00022129599552993539,
      "loss": 0.7993,
      "num_input_tokens_seen": 74123680,
      "step": 127665
    },
    {
      "epoch": 19.015490020851953,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0002209624591651582,
      "loss": 0.7897,
      "num_input_tokens_seen": 74126528,
      "step": 127670
    },
    {
      "epoch": 19.01623473339291,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.00022062917247719693,
      "loss": 0.7991,
      "num_input_tokens_seen": 74129280,
      "step": 127675
    },
    {
      "epoch": 19.01697944593387,
      "grad_norm": 0.0263671875,
      "learning_rate": 0.00022029613547168702,
      "loss": 0.8117,
      "num_input_tokens_seen": 74132384,
      "step": 127680
    },
    {
      "epoch": 19.017724158474827,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.00021996334815425065,
      "loss": 0.7837,
      "num_input_tokens_seen": 74134976,
      "step": 127685
    },
    {
      "epoch": 19.018468871015788,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.00021963081053051335,
      "loss": 0.7863,
      "num_input_tokens_seen": 74138208,
      "step": 127690
    },
    {
      "epoch": 19.01921358355675,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00021929852260609229,
      "loss": 0.7841,
      "num_input_tokens_seen": 74141088,
      "step": 127695
    },
    {
      "epoch": 19.019958296097705,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.0002189664843865996,
      "loss": 0.7948,
      "num_input_tokens_seen": 74143936,
      "step": 127700
    },
    {
      "epoch": 19.020703008638666,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.00021863469587764416,
      "loss": 0.7821,
      "num_input_tokens_seen": 74146688,
      "step": 127705
    },
    {
      "epoch": 19.021447721179623,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.00021830315708483316,
      "loss": 0.7954,
      "num_input_tokens_seen": 74149472,
      "step": 127710
    },
    {
      "epoch": 19.022192433720583,
      "grad_norm": 0.037109375,
      "learning_rate": 0.0002179718680137671,
      "loss": 0.8013,
      "num_input_tokens_seen": 74152512,
      "step": 127715
    },
    {
      "epoch": 19.022937146261544,
      "grad_norm": 0.0654296875,
      "learning_rate": 0.00021764082867004153,
      "loss": 0.8049,
      "num_input_tokens_seen": 74155488,
      "step": 127720
    },
    {
      "epoch": 19.0236818588025,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.00021731003905925026,
      "loss": 0.8229,
      "num_input_tokens_seen": 74158240,
      "step": 127725
    },
    {
      "epoch": 19.02442657134346,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.00021697949918698055,
      "loss": 0.7845,
      "num_input_tokens_seen": 74160992,
      "step": 127730
    },
    {
      "epoch": 19.025171283884422,
      "grad_norm": 0.03076171875,
      "learning_rate": 0.00021664920905881956,
      "loss": 0.8069,
      "num_input_tokens_seen": 74164192,
      "step": 127735
    },
    {
      "epoch": 19.02591599642538,
      "grad_norm": 0.061767578125,
      "learning_rate": 0.00021631916868034283,
      "loss": 0.7855,
      "num_input_tokens_seen": 74167104,
      "step": 127740
    },
    {
      "epoch": 19.02666070896634,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0002159893780571309,
      "loss": 0.7783,
      "num_input_tokens_seen": 74169888,
      "step": 127745
    },
    {
      "epoch": 19.027405421507297,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.00021565983719475101,
      "loss": 0.7977,
      "num_input_tokens_seen": 74172960,
      "step": 127750
    },
    {
      "epoch": 19.028150134048257,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.00021533054609877366,
      "loss": 0.797,
      "num_input_tokens_seen": 74175712,
      "step": 127755
    },
    {
      "epoch": 19.028894846589218,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.00021500150477475944,
      "loss": 0.7994,
      "num_input_tokens_seen": 74178592,
      "step": 127760
    },
    {
      "epoch": 19.029639559130175,
      "grad_norm": 0.048583984375,
      "learning_rate": 0.00021467271322826552,
      "loss": 0.8268,
      "num_input_tokens_seen": 74181888,
      "step": 127765
    },
    {
      "epoch": 19.030384271671135,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.0002143441714648525,
      "loss": 0.8034,
      "num_input_tokens_seen": 74184672,
      "step": 127770
    },
    {
      "epoch": 19.031128984212096,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.0002140158794900643,
      "loss": 0.7946,
      "num_input_tokens_seen": 74187456,
      "step": 127775
    },
    {
      "epoch": 19.031873696753053,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.0002136878373094514,
      "loss": 0.8432,
      "num_input_tokens_seen": 74190496,
      "step": 127780
    },
    {
      "epoch": 19.032618409294013,
      "grad_norm": 0.083984375,
      "learning_rate": 0.00021336004492855442,
      "loss": 0.8042,
      "num_input_tokens_seen": 74193408,
      "step": 127785
    },
    {
      "epoch": 19.03336312183497,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.0002130325023529106,
      "loss": 0.8109,
      "num_input_tokens_seen": 74196096,
      "step": 127790
    },
    {
      "epoch": 19.03410783437593,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.00021270520958805548,
      "loss": 0.7996,
      "num_input_tokens_seen": 74199104,
      "step": 127795
    },
    {
      "epoch": 19.03485254691689,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00021237816663951468,
      "loss": 0.8019,
      "num_input_tokens_seen": 74201952,
      "step": 127800
    },
    {
      "epoch": 19.035597259457848,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.00021205137351281542,
      "loss": 0.7941,
      "num_input_tokens_seen": 74204640,
      "step": 127805
    },
    {
      "epoch": 19.03634197199881,
      "grad_norm": 0.028076171875,
      "learning_rate": 0.0002117248302134783,
      "loss": 0.806,
      "num_input_tokens_seen": 74207584,
      "step": 127810
    },
    {
      "epoch": 19.03708668453977,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00021139853674702225,
      "loss": 0.7772,
      "num_input_tokens_seen": 74210304,
      "step": 127815
    },
    {
      "epoch": 19.037831397080726,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.00021107249311895615,
      "loss": 0.7958,
      "num_input_tokens_seen": 74213184,
      "step": 127820
    },
    {
      "epoch": 19.038576109621687,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00021074669933478894,
      "loss": 0.8072,
      "num_input_tokens_seen": 74216512,
      "step": 127825
    },
    {
      "epoch": 19.039320822162644,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.00021042115540002626,
      "loss": 0.8072,
      "num_input_tokens_seen": 74219328,
      "step": 127830
    },
    {
      "epoch": 19.040065534703604,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.00021009586132016532,
      "loss": 0.7993,
      "num_input_tokens_seen": 74222432,
      "step": 127835
    },
    {
      "epoch": 19.040810247244565,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00020977081710070344,
      "loss": 0.8277,
      "num_input_tokens_seen": 74225376,
      "step": 127840
    },
    {
      "epoch": 19.041554959785522,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.0002094460227471312,
      "loss": 0.7847,
      "num_input_tokens_seen": 74228384,
      "step": 127845
    },
    {
      "epoch": 19.042299672326482,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.00020912147826493755,
      "loss": 0.8017,
      "num_input_tokens_seen": 74231424,
      "step": 127850
    },
    {
      "epoch": 19.043044384867443,
      "grad_norm": 0.056884765625,
      "learning_rate": 0.00020879718365960143,
      "loss": 0.8033,
      "num_input_tokens_seen": 74234720,
      "step": 127855
    },
    {
      "epoch": 19.0437890974084,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.00020847313893660512,
      "loss": 0.8078,
      "num_input_tokens_seen": 74237472,
      "step": 127860
    },
    {
      "epoch": 19.04453380994936,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00020814934410142094,
      "loss": 0.7968,
      "num_input_tokens_seen": 74240288,
      "step": 127865
    },
    {
      "epoch": 19.045278522490317,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00020782579915952113,
      "loss": 0.7992,
      "num_input_tokens_seen": 74242976,
      "step": 127870
    },
    {
      "epoch": 19.046023235031278,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.00020750250411636972,
      "loss": 0.7901,
      "num_input_tokens_seen": 74245664,
      "step": 127875
    },
    {
      "epoch": 19.04676794757224,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00020717945897742894,
      "loss": 0.8017,
      "num_input_tokens_seen": 74248576,
      "step": 127880
    },
    {
      "epoch": 19.047512660113195,
      "grad_norm": 0.0291748046875,
      "learning_rate": 0.00020685666374815613,
      "loss": 0.8025,
      "num_input_tokens_seen": 74251392,
      "step": 127885
    },
    {
      "epoch": 19.048257372654156,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.00020653411843400693,
      "loss": 0.7863,
      "num_input_tokens_seen": 74254432,
      "step": 127890
    },
    {
      "epoch": 19.049002085195113,
      "grad_norm": 0.0299072265625,
      "learning_rate": 0.00020621182304042694,
      "loss": 0.7938,
      "num_input_tokens_seen": 74257248,
      "step": 127895
    },
    {
      "epoch": 19.049746797736073,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00020588977757286185,
      "loss": 0.7958,
      "num_input_tokens_seen": 74260032,
      "step": 127900
    },
    {
      "epoch": 19.050491510277034,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0002055679820367523,
      "loss": 0.7928,
      "num_input_tokens_seen": 74263040,
      "step": 127905
    },
    {
      "epoch": 19.05123622281799,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.00020524643643753725,
      "loss": 0.8015,
      "num_input_tokens_seen": 74266176,
      "step": 127910
    },
    {
      "epoch": 19.05198093535895,
      "grad_norm": 0.0625,
      "learning_rate": 0.00020492514078064572,
      "loss": 0.7951,
      "num_input_tokens_seen": 74268992,
      "step": 127915
    },
    {
      "epoch": 19.052725647899912,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00020460409507150833,
      "loss": 0.816,
      "num_input_tokens_seen": 74271872,
      "step": 127920
    },
    {
      "epoch": 19.05347036044087,
      "grad_norm": 0.046875,
      "learning_rate": 0.00020428329931554412,
      "loss": 0.8078,
      "num_input_tokens_seen": 74274688,
      "step": 127925
    },
    {
      "epoch": 19.05421507298183,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.0002039627535181787,
      "loss": 0.7916,
      "num_input_tokens_seen": 74277824,
      "step": 127930
    },
    {
      "epoch": 19.054959785522787,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0002036424576848228,
      "loss": 0.7968,
      "num_input_tokens_seen": 74280768,
      "step": 127935
    },
    {
      "epoch": 19.055704498063747,
      "grad_norm": 0.0257568359375,
      "learning_rate": 0.00020332241182089206,
      "loss": 0.7985,
      "num_input_tokens_seen": 74283936,
      "step": 127940
    },
    {
      "epoch": 19.056449210604708,
      "grad_norm": 0.056640625,
      "learning_rate": 0.00020300261593178713,
      "loss": 0.8019,
      "num_input_tokens_seen": 74286976,
      "step": 127945
    },
    {
      "epoch": 19.057193923145665,
      "grad_norm": 0.020751953125,
      "learning_rate": 0.00020268307002291707,
      "loss": 0.7867,
      "num_input_tokens_seen": 74289888,
      "step": 127950
    },
    {
      "epoch": 19.057938635686625,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.00020236377409967588,
      "loss": 0.8003,
      "num_input_tokens_seen": 74292704,
      "step": 127955
    },
    {
      "epoch": 19.058683348227586,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0002020447281674592,
      "loss": 0.795,
      "num_input_tokens_seen": 74295968,
      "step": 127960
    },
    {
      "epoch": 19.059428060768543,
      "grad_norm": 0.06640625,
      "learning_rate": 0.00020172593223165614,
      "loss": 0.8012,
      "num_input_tokens_seen": 74299168,
      "step": 127965
    },
    {
      "epoch": 19.060172773309503,
      "grad_norm": 0.07421875,
      "learning_rate": 0.00020140738629765396,
      "loss": 0.7869,
      "num_input_tokens_seen": 74302112,
      "step": 127970
    },
    {
      "epoch": 19.06091748585046,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.00020108909037083343,
      "loss": 0.8064,
      "num_input_tokens_seen": 74304800,
      "step": 127975
    },
    {
      "epoch": 19.06166219839142,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.00020077104445657024,
      "loss": 0.7967,
      "num_input_tokens_seen": 74307648,
      "step": 127980
    },
    {
      "epoch": 19.06240691093238,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00020045324856024003,
      "loss": 0.7803,
      "num_input_tokens_seen": 74310496,
      "step": 127985
    },
    {
      "epoch": 19.06315162347334,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0002001357026872119,
      "loss": 0.7796,
      "num_input_tokens_seen": 74313600,
      "step": 127990
    },
    {
      "epoch": 19.0638963360143,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.00019981840684284657,
      "loss": 0.8018,
      "num_input_tokens_seen": 74316544,
      "step": 127995
    },
    {
      "epoch": 19.06464104855526,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0001995013610325097,
      "loss": 0.8006,
      "num_input_tokens_seen": 74319392,
      "step": 128000
    },
    {
      "epoch": 19.065385761096216,
      "grad_norm": 0.043701171875,
      "learning_rate": 0.00019918456526155204,
      "loss": 0.7995,
      "num_input_tokens_seen": 74322464,
      "step": 128005
    },
    {
      "epoch": 19.066130473637177,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.00019886801953533095,
      "loss": 0.7921,
      "num_input_tokens_seen": 74325376,
      "step": 128010
    },
    {
      "epoch": 19.066875186178134,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00019855172385918883,
      "loss": 0.7778,
      "num_input_tokens_seen": 74328224,
      "step": 128015
    },
    {
      "epoch": 19.067619898719094,
      "grad_norm": 0.08203125,
      "learning_rate": 0.00019823567823847477,
      "loss": 0.7902,
      "num_input_tokens_seen": 74331424,
      "step": 128020
    },
    {
      "epoch": 19.068364611260055,
      "grad_norm": 0.059814453125,
      "learning_rate": 0.0001979198826785228,
      "loss": 0.7865,
      "num_input_tokens_seen": 74334464,
      "step": 128025
    },
    {
      "epoch": 19.069109323801012,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.00019760433718467196,
      "loss": 0.8002,
      "num_input_tokens_seen": 74337504,
      "step": 128030
    },
    {
      "epoch": 19.069854036341972,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.00019728904176225136,
      "loss": 0.7925,
      "num_input_tokens_seen": 74340352,
      "step": 128035
    },
    {
      "epoch": 19.070598748882933,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00019697399641658506,
      "loss": 0.8008,
      "num_input_tokens_seen": 74343264,
      "step": 128040
    },
    {
      "epoch": 19.07134346142389,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0001966592011530005,
      "loss": 0.81,
      "num_input_tokens_seen": 74345952,
      "step": 128045
    },
    {
      "epoch": 19.07208817396485,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001963446559768117,
      "loss": 0.7987,
      "num_input_tokens_seen": 74348768,
      "step": 128050
    },
    {
      "epoch": 19.072832886505807,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.00019603036089333448,
      "loss": 0.8006,
      "num_input_tokens_seen": 74351776,
      "step": 128055
    },
    {
      "epoch": 19.073577599046768,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.00019571631590787784,
      "loss": 0.7961,
      "num_input_tokens_seen": 74354592,
      "step": 128060
    },
    {
      "epoch": 19.07432231158773,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00019540252102574928,
      "loss": 0.796,
      "num_input_tokens_seen": 74357664,
      "step": 128065
    },
    {
      "epoch": 19.075067024128685,
      "grad_norm": 0.034912109375,
      "learning_rate": 0.00019508897625224785,
      "loss": 0.7843,
      "num_input_tokens_seen": 74360480,
      "step": 128070
    },
    {
      "epoch": 19.075811736669646,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.000194775681592671,
      "loss": 0.8079,
      "num_input_tokens_seen": 74363584,
      "step": 128075
    },
    {
      "epoch": 19.076556449210603,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.00019446263705231114,
      "loss": 0.7991,
      "num_input_tokens_seen": 74366432,
      "step": 128080
    },
    {
      "epoch": 19.077301161751564,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.0001941498426364574,
      "loss": 0.7859,
      "num_input_tokens_seen": 74369472,
      "step": 128085
    },
    {
      "epoch": 19.078045874292524,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.00019383729835039553,
      "loss": 0.807,
      "num_input_tokens_seen": 74372288,
      "step": 128090
    },
    {
      "epoch": 19.07879058683348,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00019352500419940133,
      "loss": 0.7953,
      "num_input_tokens_seen": 74375040,
      "step": 128095
    },
    {
      "epoch": 19.07953529937444,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00019321296018875555,
      "loss": 0.819,
      "num_input_tokens_seen": 74377792,
      "step": 128100
    },
    {
      "epoch": 19.080280011915402,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00019290116632372732,
      "loss": 0.8107,
      "num_input_tokens_seen": 74380864,
      "step": 128105
    },
    {
      "epoch": 19.08102472445636,
      "grad_norm": 0.0380859375,
      "learning_rate": 0.00019258962260958578,
      "loss": 0.8398,
      "num_input_tokens_seen": 74384160,
      "step": 128110
    },
    {
      "epoch": 19.08176943699732,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.00019227832905159168,
      "loss": 0.7876,
      "num_input_tokens_seen": 74387072,
      "step": 128115
    },
    {
      "epoch": 19.082514149538277,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.00019196728565500586,
      "loss": 0.7899,
      "num_input_tokens_seen": 74389920,
      "step": 128120
    },
    {
      "epoch": 19.083258862079237,
      "grad_norm": 0.212890625,
      "learning_rate": 0.0001916564924250824,
      "loss": 0.8274,
      "num_input_tokens_seen": 74392896,
      "step": 128125
    },
    {
      "epoch": 19.084003574620198,
      "grad_norm": 0.062255859375,
      "learning_rate": 0.0001913459493670705,
      "loss": 0.8116,
      "num_input_tokens_seen": 74395776,
      "step": 128130
    },
    {
      "epoch": 19.084748287161155,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.0001910356564862209,
      "loss": 0.795,
      "num_input_tokens_seen": 74398624,
      "step": 128135
    },
    {
      "epoch": 19.085492999702115,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0001907256137877694,
      "loss": 0.801,
      "num_input_tokens_seen": 74401568,
      "step": 128140
    },
    {
      "epoch": 19.086237712243076,
      "grad_norm": 0.0390625,
      "learning_rate": 0.00019041582127696022,
      "loss": 0.8215,
      "num_input_tokens_seen": 74404544,
      "step": 128145
    },
    {
      "epoch": 19.086982424784033,
      "grad_norm": 0.0283203125,
      "learning_rate": 0.00019010627895902244,
      "loss": 0.7945,
      "num_input_tokens_seen": 74407392,
      "step": 128150
    },
    {
      "epoch": 19.087727137324993,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00018979698683918854,
      "loss": 0.8114,
      "num_input_tokens_seen": 74410112,
      "step": 128155
    },
    {
      "epoch": 19.08847184986595,
      "grad_norm": 0.04345703125,
      "learning_rate": 0.00018948794492268105,
      "loss": 0.7794,
      "num_input_tokens_seen": 74412832,
      "step": 128160
    },
    {
      "epoch": 19.08921656240691,
      "grad_norm": 0.0625,
      "learning_rate": 0.0001891791532147208,
      "loss": 0.8009,
      "num_input_tokens_seen": 74415648,
      "step": 128165
    },
    {
      "epoch": 19.08996127494787,
      "grad_norm": 0.044677734375,
      "learning_rate": 0.00018887061172052688,
      "loss": 0.8126,
      "num_input_tokens_seen": 74418592,
      "step": 128170
    },
    {
      "epoch": 19.09070598748883,
      "grad_norm": 0.03369140625,
      "learning_rate": 0.00018856232044531018,
      "loss": 0.8162,
      "num_input_tokens_seen": 74421632,
      "step": 128175
    },
    {
      "epoch": 19.09145070002979,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.00018825427939427985,
      "loss": 0.7847,
      "num_input_tokens_seen": 74424576,
      "step": 128180
    },
    {
      "epoch": 19.09219541257075,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.0001879464885726384,
      "loss": 0.7852,
      "num_input_tokens_seen": 74427296,
      "step": 128185
    },
    {
      "epoch": 19.092940125111706,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.00018763894798558834,
      "loss": 0.8124,
      "num_input_tokens_seen": 74430048,
      "step": 128190
    },
    {
      "epoch": 19.093684837652667,
      "grad_norm": 0.09375,
      "learning_rate": 0.00018733165763832382,
      "loss": 0.8524,
      "num_input_tokens_seen": 74432992,
      "step": 128195
    },
    {
      "epoch": 19.094429550193624,
      "grad_norm": 0.05859375,
      "learning_rate": 0.00018702461753603404,
      "loss": 0.8098,
      "num_input_tokens_seen": 74436032,
      "step": 128200
    },
    {
      "epoch": 19.095174262734584,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.00018671782768390986,
      "loss": 0.7878,
      "num_input_tokens_seen": 74439008,
      "step": 128205
    },
    {
      "epoch": 19.095918975275545,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.00018641128808713047,
      "loss": 0.7952,
      "num_input_tokens_seen": 74441760,
      "step": 128210
    },
    {
      "epoch": 19.096663687816502,
      "grad_norm": 0.0546875,
      "learning_rate": 0.00018610499875087837,
      "loss": 0.827,
      "num_input_tokens_seen": 74444544,
      "step": 128215
    },
    {
      "epoch": 19.097408400357462,
      "grad_norm": 0.06884765625,
      "learning_rate": 0.0001857989596803261,
      "loss": 0.7851,
      "num_input_tokens_seen": 74447456,
      "step": 128220
    },
    {
      "epoch": 19.098153112898423,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.00018549317088064287,
      "loss": 0.7959,
      "num_input_tokens_seen": 74450304,
      "step": 128225
    },
    {
      "epoch": 19.09889782543938,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001851876323569962,
      "loss": 0.7768,
      "num_input_tokens_seen": 74453184,
      "step": 128230
    },
    {
      "epoch": 19.09964253798034,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.00018488234411454695,
      "loss": 0.7997,
      "num_input_tokens_seen": 74456096,
      "step": 128235
    },
    {
      "epoch": 19.100387250521297,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00018457730615845436,
      "loss": 0.7843,
      "num_input_tokens_seen": 74458816,
      "step": 128240
    },
    {
      "epoch": 19.101131963062258,
      "grad_norm": 0.0693359375,
      "learning_rate": 0.00018427251849386926,
      "loss": 0.7902,
      "num_input_tokens_seen": 74461888,
      "step": 128245
    },
    {
      "epoch": 19.10187667560322,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00018396798112594258,
      "loss": 0.8022,
      "num_input_tokens_seen": 74464960,
      "step": 128250
    },
    {
      "epoch": 19.102621388144176,
      "grad_norm": 0.027587890625,
      "learning_rate": 0.00018366369405981687,
      "loss": 0.8085,
      "num_input_tokens_seen": 74468000,
      "step": 128255
    },
    {
      "epoch": 19.103366100685136,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0001833596573006363,
      "loss": 0.8058,
      "num_input_tokens_seen": 74470976,
      "step": 128260
    },
    {
      "epoch": 19.104110813226093,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.00018305587085353515,
      "loss": 0.7899,
      "num_input_tokens_seen": 74473920,
      "step": 128265
    },
    {
      "epoch": 19.104855525767054,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00018275233472364594,
      "loss": 0.8055,
      "num_input_tokens_seen": 74476544,
      "step": 128270
    },
    {
      "epoch": 19.105600238308014,
      "grad_norm": 0.054931640625,
      "learning_rate": 0.00018244904891609625,
      "loss": 0.7782,
      "num_input_tokens_seen": 74479424,
      "step": 128275
    },
    {
      "epoch": 19.10634495084897,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00018214601343600865,
      "loss": 0.7882,
      "num_input_tokens_seen": 74482656,
      "step": 128280
    },
    {
      "epoch": 19.10708966338993,
      "grad_norm": 0.057373046875,
      "learning_rate": 0.00018184322828850575,
      "loss": 0.7863,
      "num_input_tokens_seen": 74485408,
      "step": 128285
    },
    {
      "epoch": 19.107834375930892,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.00018154069347870006,
      "loss": 0.7954,
      "num_input_tokens_seen": 74488800,
      "step": 128290
    },
    {
      "epoch": 19.10857908847185,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0001812384090117042,
      "loss": 0.8325,
      "num_input_tokens_seen": 74491904,
      "step": 128295
    },
    {
      "epoch": 19.10932380101281,
      "grad_norm": 0.0289306640625,
      "learning_rate": 0.00018093637489262403,
      "loss": 0.7968,
      "num_input_tokens_seen": 74494496,
      "step": 128300
    },
    {
      "epoch": 19.110068513553767,
      "grad_norm": 0.052490234375,
      "learning_rate": 0.00018063459112656222,
      "loss": 0.798,
      "num_input_tokens_seen": 74497472,
      "step": 128305
    },
    {
      "epoch": 19.110813226094727,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0001803330577186163,
      "loss": 0.7914,
      "num_input_tokens_seen": 74500288,
      "step": 128310
    },
    {
      "epoch": 19.111557938635688,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.00018003177467388386,
      "loss": 0.7889,
      "num_input_tokens_seen": 74504576,
      "step": 128315
    },
    {
      "epoch": 19.112302651176645,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.00017973074199745086,
      "loss": 0.8044,
      "num_input_tokens_seen": 74507360,
      "step": 128320
    },
    {
      "epoch": 19.113047363717605,
      "grad_norm": 0.05908203125,
      "learning_rate": 0.00017942995969440323,
      "loss": 0.7932,
      "num_input_tokens_seen": 74510112,
      "step": 128325
    },
    {
      "epoch": 19.113792076258566,
      "grad_norm": 0.111328125,
      "learning_rate": 0.00017912942776982522,
      "loss": 0.8119,
      "num_input_tokens_seen": 74512960,
      "step": 128330
    },
    {
      "epoch": 19.114536788799523,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00017882914622878942,
      "loss": 0.8058,
      "num_input_tokens_seen": 74515936,
      "step": 128335
    },
    {
      "epoch": 19.115281501340483,
      "grad_norm": 0.05517578125,
      "learning_rate": 0.00017852911507637513,
      "loss": 0.8082,
      "num_input_tokens_seen": 74518816,
      "step": 128340
    },
    {
      "epoch": 19.11602621388144,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00017822933431764498,
      "loss": 0.8361,
      "num_input_tokens_seen": 74521472,
      "step": 128345
    },
    {
      "epoch": 19.1167709264224,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.00017792980395766822,
      "loss": 0.7928,
      "num_input_tokens_seen": 74524352,
      "step": 128350
    },
    {
      "epoch": 19.11751563896336,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00017763052400150247,
      "loss": 0.7933,
      "num_input_tokens_seen": 74527264,
      "step": 128355
    },
    {
      "epoch": 19.11826035150432,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.0001773314944542037,
      "loss": 0.8001,
      "num_input_tokens_seen": 74529952,
      "step": 128360
    },
    {
      "epoch": 19.11900506404528,
      "grad_norm": 0.080078125,
      "learning_rate": 0.00017703271532082453,
      "loss": 0.7926,
      "num_input_tokens_seen": 74532896,
      "step": 128365
    },
    {
      "epoch": 19.11974977658624,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00017673418660641094,
      "loss": 0.8039,
      "num_input_tokens_seen": 74535872,
      "step": 128370
    },
    {
      "epoch": 19.120494489127196,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00017643590831600885,
      "loss": 0.7992,
      "num_input_tokens_seen": 74538912,
      "step": 128375
    },
    {
      "epoch": 19.121239201668157,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.00017613788045465594,
      "loss": 0.7981,
      "num_input_tokens_seen": 74541760,
      "step": 128380
    },
    {
      "epoch": 19.121983914209114,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0001758401030273865,
      "loss": 0.8015,
      "num_input_tokens_seen": 74544896,
      "step": 128385
    },
    {
      "epoch": 19.122728626750074,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0001755425760392315,
      "loss": 0.7834,
      "num_input_tokens_seen": 74547616,
      "step": 128390
    },
    {
      "epoch": 19.123473339291035,
      "grad_norm": 0.041748046875,
      "learning_rate": 0.00017524529949522027,
      "loss": 0.7976,
      "num_input_tokens_seen": 74550432,
      "step": 128395
    },
    {
      "epoch": 19.124218051831992,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00017494827340037045,
      "loss": 0.795,
      "num_input_tokens_seen": 74553408,
      "step": 128400
    },
    {
      "epoch": 19.124962764372953,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.000174651497759703,
      "loss": 0.7911,
      "num_input_tokens_seen": 74556480,
      "step": 128405
    },
    {
      "epoch": 19.12570747691391,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00017435497257823062,
      "loss": 0.8006,
      "num_input_tokens_seen": 74559296,
      "step": 128410
    },
    {
      "epoch": 19.12645218945487,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.00017405869786096261,
      "loss": 0.8229,
      "num_input_tokens_seen": 74562400,
      "step": 128415
    },
    {
      "epoch": 19.12719690199583,
      "grad_norm": 0.06396484375,
      "learning_rate": 0.00017376267361290498,
      "loss": 0.7762,
      "num_input_tokens_seen": 74565344,
      "step": 128420
    },
    {
      "epoch": 19.127941614536788,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.00017346689983905705,
      "loss": 0.8204,
      "num_input_tokens_seen": 74568320,
      "step": 128425
    },
    {
      "epoch": 19.128686327077748,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.00017317137654441816,
      "loss": 0.809,
      "num_input_tokens_seen": 74571296,
      "step": 128430
    },
    {
      "epoch": 19.12943103961871,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0001728761037339793,
      "loss": 0.7712,
      "num_input_tokens_seen": 74574240,
      "step": 128435
    },
    {
      "epoch": 19.130175752159666,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0001725810814127282,
      "loss": 0.8091,
      "num_input_tokens_seen": 74576992,
      "step": 128440
    },
    {
      "epoch": 19.130920464700626,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.00017228630958564917,
      "loss": 0.8158,
      "num_input_tokens_seen": 74579808,
      "step": 128445
    },
    {
      "epoch": 19.131665177241583,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.0001719917882577232,
      "loss": 0.7944,
      "num_input_tokens_seen": 74582784,
      "step": 128450
    },
    {
      "epoch": 19.132409889782544,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.00017169751743392636,
      "loss": 0.7933,
      "num_input_tokens_seen": 74585632,
      "step": 128455
    },
    {
      "epoch": 19.133154602323504,
      "grad_norm": 0.03515625,
      "learning_rate": 0.00017140349711922964,
      "loss": 0.8119,
      "num_input_tokens_seen": 74588288,
      "step": 128460
    },
    {
      "epoch": 19.13389931486446,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.00017110972731859908,
      "loss": 0.7956,
      "num_input_tokens_seen": 74591424,
      "step": 128465
    },
    {
      "epoch": 19.13464402740542,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.00017081620803699736,
      "loss": 0.7986,
      "num_input_tokens_seen": 74594368,
      "step": 128470
    },
    {
      "epoch": 19.135388739946382,
      "grad_norm": 0.07080078125,
      "learning_rate": 0.00017052293927938554,
      "loss": 0.7878,
      "num_input_tokens_seen": 74597056,
      "step": 128475
    },
    {
      "epoch": 19.13613345248734,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.0001702299210507163,
      "loss": 0.8242,
      "num_input_tokens_seen": 74600000,
      "step": 128480
    },
    {
      "epoch": 19.1368781650283,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.00016993715335593905,
      "loss": 0.8113,
      "num_input_tokens_seen": 74602944,
      "step": 128485
    },
    {
      "epoch": 19.137622877569257,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00016964463620000312,
      "loss": 0.7895,
      "num_input_tokens_seen": 74605760,
      "step": 128490
    },
    {
      "epoch": 19.138367590110217,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.00016935236958784628,
      "loss": 0.7923,
      "num_input_tokens_seen": 74608800,
      "step": 128495
    },
    {
      "epoch": 19.139112302651178,
      "grad_norm": 0.09765625,
      "learning_rate": 0.00016906035352440784,
      "loss": 0.7943,
      "num_input_tokens_seen": 74611648,
      "step": 128500
    },
    {
      "epoch": 19.139857015192135,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.00016876858801462223,
      "loss": 0.807,
      "num_input_tokens_seen": 74614720,
      "step": 128505
    },
    {
      "epoch": 19.140601727733095,
      "grad_norm": 0.06640625,
      "learning_rate": 0.00016847707306341718,
      "loss": 0.8162,
      "num_input_tokens_seen": 74617696,
      "step": 128510
    },
    {
      "epoch": 19.141346440274056,
      "grad_norm": 0.041015625,
      "learning_rate": 0.00016818580867571873,
      "loss": 0.8045,
      "num_input_tokens_seen": 74620512,
      "step": 128515
    },
    {
      "epoch": 19.142091152815013,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00016789479485644464,
      "loss": 0.7806,
      "num_input_tokens_seen": 74623424,
      "step": 128520
    },
    {
      "epoch": 19.142835865355973,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.00016760403161051595,
      "loss": 0.8044,
      "num_input_tokens_seen": 74626272,
      "step": 128525
    },
    {
      "epoch": 19.14358057789693,
      "grad_norm": 0.287109375,
      "learning_rate": 0.00016731351894283875,
      "loss": 0.814,
      "num_input_tokens_seen": 74628960,
      "step": 128530
    },
    {
      "epoch": 19.14432529043789,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.00016702325685832742,
      "loss": 0.8052,
      "num_input_tokens_seen": 74631936,
      "step": 128535
    },
    {
      "epoch": 19.14507000297885,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.00016673324536188138,
      "loss": 0.7944,
      "num_input_tokens_seen": 74634976,
      "step": 128540
    },
    {
      "epoch": 19.14581471551981,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.00016644348445840173,
      "loss": 0.7974,
      "num_input_tokens_seen": 74637888,
      "step": 128545
    },
    {
      "epoch": 19.14655942806077,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.00016615397415278287,
      "loss": 0.7972,
      "num_input_tokens_seen": 74640768,
      "step": 128550
    },
    {
      "epoch": 19.14730414060173,
      "grad_norm": 0.04443359375,
      "learning_rate": 0.00016586471444991757,
      "loss": 0.798,
      "num_input_tokens_seen": 74643648,
      "step": 128555
    },
    {
      "epoch": 19.148048853142686,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.00016557570535469024,
      "loss": 0.797,
      "num_input_tokens_seen": 74646720,
      "step": 128560
    },
    {
      "epoch": 19.148793565683647,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00016528694687198363,
      "loss": 0.7935,
      "num_input_tokens_seen": 74649312,
      "step": 128565
    },
    {
      "epoch": 19.149538278224604,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.0001649984390066772,
      "loss": 0.8174,
      "num_input_tokens_seen": 74652256,
      "step": 128570
    },
    {
      "epoch": 19.150282990765565,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.0001647101817636437,
      "loss": 0.7764,
      "num_input_tokens_seen": 74654976,
      "step": 128575
    },
    {
      "epoch": 19.151027703306525,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.00016442217514775424,
      "loss": 0.8051,
      "num_input_tokens_seen": 74657856,
      "step": 128580
    },
    {
      "epoch": 19.151772415847482,
      "grad_norm": 0.05419921875,
      "learning_rate": 0.00016413441916387494,
      "loss": 0.8122,
      "num_input_tokens_seen": 74660832,
      "step": 128585
    },
    {
      "epoch": 19.152517128388443,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.0001638469138168652,
      "loss": 0.7799,
      "num_input_tokens_seen": 74663488,
      "step": 128590
    },
    {
      "epoch": 19.1532618409294,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.00016355965911158288,
      "loss": 0.8063,
      "num_input_tokens_seen": 74666432,
      "step": 128595
    },
    {
      "epoch": 19.15400655347036,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.00016327265505288068,
      "loss": 0.7942,
      "num_input_tokens_seen": 74669184,
      "step": 128600
    },
    {
      "epoch": 19.15475126601132,
      "grad_norm": 0.052734375,
      "learning_rate": 0.00016298590164560978,
      "loss": 0.7974,
      "num_input_tokens_seen": 74671968,
      "step": 128605
    },
    {
      "epoch": 19.155495978552278,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.0001626993988946096,
      "loss": 0.8111,
      "num_input_tokens_seen": 74674816,
      "step": 128610
    },
    {
      "epoch": 19.156240691093238,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.00016241314680472463,
      "loss": 0.7904,
      "num_input_tokens_seen": 74677728,
      "step": 128615
    },
    {
      "epoch": 19.1569854036342,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.0001621271453807893,
      "loss": 0.7901,
      "num_input_tokens_seen": 74680672,
      "step": 128620
    },
    {
      "epoch": 19.157730116175156,
      "grad_norm": 0.0576171875,
      "learning_rate": 0.00016184139462763647,
      "loss": 0.8083,
      "num_input_tokens_seen": 74683584,
      "step": 128625
    },
    {
      "epoch": 19.158474828716116,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00016155589455009054,
      "loss": 0.7842,
      "num_input_tokens_seen": 74686368,
      "step": 128630
    },
    {
      "epoch": 19.159219541257073,
      "grad_norm": 0.023681640625,
      "learning_rate": 0.00016127064515297773,
      "loss": 0.8035,
      "num_input_tokens_seen": 74688992,
      "step": 128635
    },
    {
      "epoch": 19.159964253798034,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.0001609856464411158,
      "loss": 0.8193,
      "num_input_tokens_seen": 74691904,
      "step": 128640
    },
    {
      "epoch": 19.160708966338994,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.00016070089841931757,
      "loss": 0.7972,
      "num_input_tokens_seen": 74694848,
      "step": 128645
    },
    {
      "epoch": 19.16145367887995,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00016041640109239918,
      "loss": 0.8342,
      "num_input_tokens_seen": 74697952,
      "step": 128650
    },
    {
      "epoch": 19.16219839142091,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.00016013215446516015,
      "loss": 0.8265,
      "num_input_tokens_seen": 74700736,
      "step": 128655
    },
    {
      "epoch": 19.162943103961872,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.0001598481585424083,
      "loss": 0.8066,
      "num_input_tokens_seen": 74703520,
      "step": 128660
    },
    {
      "epoch": 19.16368781650283,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.00015956441332893643,
      "loss": 0.7938,
      "num_input_tokens_seen": 74706336,
      "step": 128665
    },
    {
      "epoch": 19.16443252904379,
      "grad_norm": 0.044921875,
      "learning_rate": 0.00015928091882954243,
      "loss": 0.7916,
      "num_input_tokens_seen": 74709088,
      "step": 128670
    },
    {
      "epoch": 19.165177241584747,
      "grad_norm": 0.033203125,
      "learning_rate": 0.00015899767504901408,
      "loss": 0.7966,
      "num_input_tokens_seen": 74712224,
      "step": 128675
    },
    {
      "epoch": 19.165921954125707,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0001587146819921359,
      "loss": 0.8019,
      "num_input_tokens_seen": 74715200,
      "step": 128680
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.00015843193966368906,
      "loss": 0.7838,
      "num_input_tokens_seen": 74718176,
      "step": 128685
    },
    {
      "epoch": 19.167411379207625,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00015814944806845143,
      "loss": 0.8092,
      "num_input_tokens_seen": 74720928,
      "step": 128690
    },
    {
      "epoch": 19.168156091748585,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00015786720721119252,
      "loss": 0.8068,
      "num_input_tokens_seen": 74723936,
      "step": 128695
    },
    {
      "epoch": 19.168900804289546,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0001575852170966835,
      "loss": 0.7948,
      "num_input_tokens_seen": 74726784,
      "step": 128700
    },
    {
      "epoch": 19.169645516830503,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0001573034777296872,
      "loss": 0.8148,
      "num_input_tokens_seen": 74729408,
      "step": 128705
    },
    {
      "epoch": 19.170390229371463,
      "grad_norm": 0.189453125,
      "learning_rate": 0.00015702198911496157,
      "loss": 0.8254,
      "num_input_tokens_seen": 74732640,
      "step": 128710
    },
    {
      "epoch": 19.17113494191242,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.00015674075125726437,
      "loss": 0.7913,
      "num_input_tokens_seen": 74735296,
      "step": 128715
    },
    {
      "epoch": 19.17187965445338,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.00015645976416134687,
      "loss": 0.8031,
      "num_input_tokens_seen": 74738112,
      "step": 128720
    },
    {
      "epoch": 19.17262436699434,
      "grad_norm": 0.025390625,
      "learning_rate": 0.00015617902783195524,
      "loss": 0.8035,
      "num_input_tokens_seen": 74740864,
      "step": 128725
    },
    {
      "epoch": 19.1733690795353,
      "grad_norm": 0.06787109375,
      "learning_rate": 0.00015589854227383238,
      "loss": 0.8023,
      "num_input_tokens_seen": 74743520,
      "step": 128730
    },
    {
      "epoch": 19.17411379207626,
      "grad_norm": 0.0439453125,
      "learning_rate": 0.00015561830749171612,
      "loss": 0.7863,
      "num_input_tokens_seen": 74746400,
      "step": 128735
    },
    {
      "epoch": 19.17485850461722,
      "grad_norm": 0.091796875,
      "learning_rate": 0.00015533832349034105,
      "loss": 0.7936,
      "num_input_tokens_seen": 74749280,
      "step": 128740
    },
    {
      "epoch": 19.175603217158177,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00015505859027443834,
      "loss": 0.7776,
      "num_input_tokens_seen": 74752192,
      "step": 128745
    },
    {
      "epoch": 19.176347929699137,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.00015477910784873427,
      "loss": 0.8084,
      "num_input_tokens_seen": 74755360,
      "step": 128750
    },
    {
      "epoch": 19.177092642240094,
      "grad_norm": 0.0625,
      "learning_rate": 0.0001544998762179467,
      "loss": 0.8093,
      "num_input_tokens_seen": 74758304,
      "step": 128755
    },
    {
      "epoch": 19.177837354781055,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00015422089538679685,
      "loss": 0.8136,
      "num_input_tokens_seen": 74761856,
      "step": 128760
    },
    {
      "epoch": 19.178582067322015,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.00015394216535999594,
      "loss": 0.7974,
      "num_input_tokens_seen": 74764864,
      "step": 128765
    },
    {
      "epoch": 19.179326779862972,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00015366368614225356,
      "loss": 0.7901,
      "num_input_tokens_seen": 74767488,
      "step": 128770
    },
    {
      "epoch": 19.180071492403933,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.00015338545773827428,
      "loss": 0.8021,
      "num_input_tokens_seen": 74770368,
      "step": 128775
    },
    {
      "epoch": 19.18081620494489,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00015310748015275765,
      "loss": 0.8021,
      "num_input_tokens_seen": 74773120,
      "step": 128780
    },
    {
      "epoch": 19.18156091748585,
      "grad_norm": 0.06005859375,
      "learning_rate": 0.0001528297533904016,
      "loss": 0.788,
      "num_input_tokens_seen": 74776000,
      "step": 128785
    },
    {
      "epoch": 19.18230563002681,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.00015255227745589406,
      "loss": 0.8082,
      "num_input_tokens_seen": 74778720,
      "step": 128790
    },
    {
      "epoch": 19.183050342567768,
      "grad_norm": 0.05712890625,
      "learning_rate": 0.00015227505235392792,
      "loss": 0.7885,
      "num_input_tokens_seen": 74781824,
      "step": 128795
    },
    {
      "epoch": 19.183795055108728,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.0001519980780891844,
      "loss": 0.796,
      "num_input_tokens_seen": 74784800,
      "step": 128800
    },
    {
      "epoch": 19.18453976764969,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0001517213546663415,
      "loss": 0.7813,
      "num_input_tokens_seen": 74787872,
      "step": 128805
    },
    {
      "epoch": 19.185284480190646,
      "grad_norm": 0.05859375,
      "learning_rate": 0.00015144488209007544,
      "loss": 0.7794,
      "num_input_tokens_seen": 74790496,
      "step": 128810
    },
    {
      "epoch": 19.186029192731606,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00015116866036505748,
      "loss": 0.8014,
      "num_input_tokens_seen": 74793120,
      "step": 128815
    },
    {
      "epoch": 19.186773905272563,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.00015089268949595392,
      "loss": 0.794,
      "num_input_tokens_seen": 74795808,
      "step": 128820
    },
    {
      "epoch": 19.187518617813524,
      "grad_norm": 0.04248046875,
      "learning_rate": 0.0001506169694874243,
      "loss": 0.8031,
      "num_input_tokens_seen": 74798688,
      "step": 128825
    },
    {
      "epoch": 19.188263330354484,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.00015034150034412997,
      "loss": 0.8012,
      "num_input_tokens_seen": 74801248,
      "step": 128830
    },
    {
      "epoch": 19.18900804289544,
      "grad_norm": 0.033203125,
      "learning_rate": 0.00015006628207072214,
      "loss": 0.783,
      "num_input_tokens_seen": 74803936,
      "step": 128835
    },
    {
      "epoch": 19.189752755436402,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.0001497913146718538,
      "loss": 0.8133,
      "num_input_tokens_seen": 74807008,
      "step": 128840
    },
    {
      "epoch": 19.190497467977362,
      "grad_norm": 0.02587890625,
      "learning_rate": 0.0001495165981521662,
      "loss": 0.7861,
      "num_input_tokens_seen": 74809952,
      "step": 128845
    },
    {
      "epoch": 19.19124218051832,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00014924213251630235,
      "loss": 0.7878,
      "num_input_tokens_seen": 74813024,
      "step": 128850
    },
    {
      "epoch": 19.19198689305928,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00014896791776890017,
      "loss": 0.8017,
      "num_input_tokens_seen": 74816032,
      "step": 128855
    },
    {
      "epoch": 19.192731605600237,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.00014869395391459093,
      "loss": 0.7955,
      "num_input_tokens_seen": 74819200,
      "step": 128860
    },
    {
      "epoch": 19.193476318141197,
      "grad_norm": 0.0269775390625,
      "learning_rate": 0.00014842024095800265,
      "loss": 0.7809,
      "num_input_tokens_seen": 74821888,
      "step": 128865
    },
    {
      "epoch": 19.194221030682158,
      "grad_norm": 0.03759765625,
      "learning_rate": 0.00014814677890375992,
      "loss": 0.7868,
      "num_input_tokens_seen": 74825952,
      "step": 128870
    },
    {
      "epoch": 19.194965743223115,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.0001478735677564824,
      "loss": 0.7959,
      "num_input_tokens_seen": 74828896,
      "step": 128875
    },
    {
      "epoch": 19.195710455764075,
      "grad_norm": 0.051025390625,
      "learning_rate": 0.00014760060752078808,
      "loss": 0.806,
      "num_input_tokens_seen": 74831648,
      "step": 128880
    },
    {
      "epoch": 19.196455168305036,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.00014732789820128488,
      "loss": 0.8071,
      "num_input_tokens_seen": 74834592,
      "step": 128885
    },
    {
      "epoch": 19.197199880845993,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.00014705543980258085,
      "loss": 0.7814,
      "num_input_tokens_seen": 74837504,
      "step": 128890
    },
    {
      "epoch": 19.197944593386953,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001467832323292789,
      "loss": 0.7638,
      "num_input_tokens_seen": 74840352,
      "step": 128895
    },
    {
      "epoch": 19.19868930592791,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00014651127578597876,
      "loss": 0.7925,
      "num_input_tokens_seen": 74843456,
      "step": 128900
    },
    {
      "epoch": 19.19943401846887,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0001462395701772734,
      "loss": 0.8455,
      "num_input_tokens_seen": 74846240,
      "step": 128905
    },
    {
      "epoch": 19.20017873100983,
      "grad_norm": 0.083984375,
      "learning_rate": 0.00014596811550775411,
      "loss": 0.7873,
      "num_input_tokens_seen": 74849184,
      "step": 128910
    },
    {
      "epoch": 19.20092344355079,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00014569691178200727,
      "loss": 0.8201,
      "num_input_tokens_seen": 74852160,
      "step": 128915
    },
    {
      "epoch": 19.20166815609175,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0001454259590046142,
      "loss": 0.8406,
      "num_input_tokens_seen": 74855136,
      "step": 128920
    },
    {
      "epoch": 19.202412868632706,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0001451552571801512,
      "loss": 0.7838,
      "num_input_tokens_seen": 74858272,
      "step": 128925
    },
    {
      "epoch": 19.203157581173667,
      "grad_norm": 0.05029296875,
      "learning_rate": 0.00014488480631319134,
      "loss": 0.7904,
      "num_input_tokens_seen": 74861440,
      "step": 128930
    },
    {
      "epoch": 19.203902293714627,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.00014461460640830592,
      "loss": 0.8275,
      "num_input_tokens_seen": 74864160,
      "step": 128935
    },
    {
      "epoch": 19.204647006255584,
      "grad_norm": 0.055419921875,
      "learning_rate": 0.00014434465747005797,
      "loss": 0.7898,
      "num_input_tokens_seen": 74866976,
      "step": 128940
    },
    {
      "epoch": 19.205391718796545,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.00014407495950300884,
      "loss": 0.8023,
      "num_input_tokens_seen": 74869664,
      "step": 128945
    },
    {
      "epoch": 19.206136431337505,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00014380551251171492,
      "loss": 0.7927,
      "num_input_tokens_seen": 74872480,
      "step": 128950
    },
    {
      "epoch": 19.206881143878462,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0001435363165007275,
      "loss": 0.8095,
      "num_input_tokens_seen": 74875232,
      "step": 128955
    },
    {
      "epoch": 19.207625856419423,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00014326737147459466,
      "loss": 0.7875,
      "num_input_tokens_seen": 74878080,
      "step": 128960
    },
    {
      "epoch": 19.20837056896038,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00014299867743785777,
      "loss": 0.7895,
      "num_input_tokens_seen": 74880992,
      "step": 128965
    },
    {
      "epoch": 19.20911528150134,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.00014273023439505982,
      "loss": 0.8062,
      "num_input_tokens_seen": 74884288,
      "step": 128970
    },
    {
      "epoch": 19.2098599940423,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0001424620423507339,
      "loss": 0.7921,
      "num_input_tokens_seen": 74887008,
      "step": 128975
    },
    {
      "epoch": 19.210604706583258,
      "grad_norm": 0.052001953125,
      "learning_rate": 0.00014219410130941135,
      "loss": 0.7849,
      "num_input_tokens_seen": 74889856,
      "step": 128980
    },
    {
      "epoch": 19.21134941912422,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.00014192641127561855,
      "loss": 0.7778,
      "num_input_tokens_seen": 74892544,
      "step": 128985
    },
    {
      "epoch": 19.21209413166518,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.00014165897225387858,
      "loss": 0.7846,
      "num_input_tokens_seen": 74895360,
      "step": 128990
    },
    {
      "epoch": 19.212838844206136,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.0001413917842487078,
      "loss": 0.7889,
      "num_input_tokens_seen": 74898240,
      "step": 128995
    },
    {
      "epoch": 19.213583556747096,
      "grad_norm": 0.055908203125,
      "learning_rate": 0.0001411248472646226,
      "loss": 0.79,
      "num_input_tokens_seen": 74901152,
      "step": 129000
    },
    {
      "epoch": 19.214328269288053,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.00014085816130613105,
      "loss": 0.7905,
      "num_input_tokens_seen": 74903968,
      "step": 129005
    },
    {
      "epoch": 19.215072981829014,
      "grad_norm": 0.0301513671875,
      "learning_rate": 0.00014059172637773788,
      "loss": 0.7998,
      "num_input_tokens_seen": 74906816,
      "step": 129010
    },
    {
      "epoch": 19.215817694369974,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00014032554248394612,
      "loss": 0.7988,
      "num_input_tokens_seen": 74909632,
      "step": 129015
    },
    {
      "epoch": 19.21656240691093,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.00014005960962925056,
      "loss": 0.7908,
      "num_input_tokens_seen": 74912224,
      "step": 129020
    },
    {
      "epoch": 19.217307119451892,
      "grad_norm": 0.050537109375,
      "learning_rate": 0.00013979392781814592,
      "loss": 0.7896,
      "num_input_tokens_seen": 74915200,
      "step": 129025
    },
    {
      "epoch": 19.218051831992852,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.0001395284970551186,
      "loss": 0.7943,
      "num_input_tokens_seen": 74917984,
      "step": 129030
    },
    {
      "epoch": 19.21879654453381,
      "grad_norm": 0.031982421875,
      "learning_rate": 0.00013926331734465501,
      "loss": 0.7946,
      "num_input_tokens_seen": 74920928,
      "step": 129035
    },
    {
      "epoch": 19.21954125707477,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.00013899838869123327,
      "loss": 0.8115,
      "num_input_tokens_seen": 74923520,
      "step": 129040
    },
    {
      "epoch": 19.220285969615727,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.00013873371109932974,
      "loss": 0.782,
      "num_input_tokens_seen": 74926528,
      "step": 129045
    },
    {
      "epoch": 19.221030682156687,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001384692845734159,
      "loss": 0.8154,
      "num_input_tokens_seen": 74929152,
      "step": 129050
    },
    {
      "epoch": 19.221775394697648,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.00013820510911795813,
      "loss": 0.7981,
      "num_input_tokens_seen": 74932160,
      "step": 129055
    },
    {
      "epoch": 19.222520107238605,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00013794118473742123,
      "loss": 0.806,
      "num_input_tokens_seen": 74934688,
      "step": 129060
    },
    {
      "epoch": 19.223264819779565,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00013767751143626162,
      "loss": 0.8143,
      "num_input_tokens_seen": 74937568,
      "step": 129065
    },
    {
      "epoch": 19.224009532320526,
      "grad_norm": 0.06689453125,
      "learning_rate": 0.00013741408921893572,
      "loss": 0.8036,
      "num_input_tokens_seen": 74940864,
      "step": 129070
    },
    {
      "epoch": 19.224754244861483,
      "grad_norm": 0.037353515625,
      "learning_rate": 0.0001371509180898933,
      "loss": 0.8106,
      "num_input_tokens_seen": 74943776,
      "step": 129075
    },
    {
      "epoch": 19.225498957402444,
      "grad_norm": 0.046875,
      "learning_rate": 0.00013688799805358087,
      "loss": 0.8095,
      "num_input_tokens_seen": 74946592,
      "step": 129080
    },
    {
      "epoch": 19.2262436699434,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.00013662532911443815,
      "loss": 0.7978,
      "num_input_tokens_seen": 74949696,
      "step": 129085
    },
    {
      "epoch": 19.22698838248436,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.0001363629112769049,
      "loss": 0.7697,
      "num_input_tokens_seen": 74952512,
      "step": 129090
    },
    {
      "epoch": 19.22773309502532,
      "grad_norm": 0.03271484375,
      "learning_rate": 0.0001361007445454143,
      "loss": 0.7871,
      "num_input_tokens_seen": 74955296,
      "step": 129095
    },
    {
      "epoch": 19.22847780756628,
      "grad_norm": 0.04931640625,
      "learning_rate": 0.0001358388289243928,
      "loss": 0.8085,
      "num_input_tokens_seen": 74958208,
      "step": 129100
    },
    {
      "epoch": 19.22922252010724,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.0001355771644182685,
      "loss": 0.8117,
      "num_input_tokens_seen": 74961056,
      "step": 129105
    },
    {
      "epoch": 19.229967232648196,
      "grad_norm": 0.05322265625,
      "learning_rate": 0.00013531575103145954,
      "loss": 0.7992,
      "num_input_tokens_seen": 74964064,
      "step": 129110
    },
    {
      "epoch": 19.230711945189157,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.00013505458876838405,
      "loss": 0.8018,
      "num_input_tokens_seen": 74966848,
      "step": 129115
    },
    {
      "epoch": 19.231456657730117,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0001347936776334535,
      "loss": 0.7985,
      "num_input_tokens_seen": 74969664,
      "step": 129120
    },
    {
      "epoch": 19.232201370271074,
      "grad_norm": 0.047607421875,
      "learning_rate": 0.00013453301763107273,
      "loss": 0.7951,
      "num_input_tokens_seen": 74972384,
      "step": 129125
    },
    {
      "epoch": 19.232946082812035,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.0001342726087656515,
      "loss": 0.7919,
      "num_input_tokens_seen": 74975360,
      "step": 129130
    },
    {
      "epoch": 19.233690795352995,
      "grad_norm": 0.05126953125,
      "learning_rate": 0.000134012451041583,
      "loss": 0.8068,
      "num_input_tokens_seen": 74978272,
      "step": 129135
    },
    {
      "epoch": 19.234435507893952,
      "grad_norm": 0.0279541015625,
      "learning_rate": 0.000133752544463267,
      "loss": 0.807,
      "num_input_tokens_seen": 74981088,
      "step": 129140
    },
    {
      "epoch": 19.235180220434913,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001334928890350917,
      "loss": 0.814,
      "num_input_tokens_seen": 74983840,
      "step": 129145
    },
    {
      "epoch": 19.23592493297587,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0001332334847614436,
      "loss": 0.801,
      "num_input_tokens_seen": 74986784,
      "step": 129150
    },
    {
      "epoch": 19.23666964551683,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.00013297433164670747,
      "loss": 0.7979,
      "num_input_tokens_seen": 74989472,
      "step": 129155
    },
    {
      "epoch": 19.23741435805779,
      "grad_norm": 0.05078125,
      "learning_rate": 0.00013271542969525818,
      "loss": 0.7987,
      "num_input_tokens_seen": 74992416,
      "step": 129160
    },
    {
      "epoch": 19.238159070598748,
      "grad_norm": 0.0400390625,
      "learning_rate": 0.00013245677891147222,
      "loss": 0.7949,
      "num_input_tokens_seen": 74995392,
      "step": 129165
    },
    {
      "epoch": 19.23890378313971,
      "grad_norm": 0.041015625,
      "learning_rate": 0.00013219837929971778,
      "loss": 0.7922,
      "num_input_tokens_seen": 74998112,
      "step": 129170
    },
    {
      "epoch": 19.23964849568067,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00013194023086436133,
      "loss": 0.7865,
      "num_input_tokens_seen": 75001024,
      "step": 129175
    },
    {
      "epoch": 19.240393208221626,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.00013168233360976277,
      "loss": 0.7955,
      "num_input_tokens_seen": 75004256,
      "step": 129180
    },
    {
      "epoch": 19.241137920762586,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.00013142468754028024,
      "loss": 0.7797,
      "num_input_tokens_seen": 75007040,
      "step": 129185
    },
    {
      "epoch": 19.241882633303543,
      "grad_norm": 0.035888671875,
      "learning_rate": 0.00013116729266026528,
      "loss": 0.7867,
      "num_input_tokens_seen": 75009888,
      "step": 129190
    },
    {
      "epoch": 19.242627345844504,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.0001309101489740677,
      "loss": 0.8038,
      "num_input_tokens_seen": 75012448,
      "step": 129195
    },
    {
      "epoch": 19.243372058385464,
      "grad_norm": 0.040771484375,
      "learning_rate": 0.00013065325648603076,
      "loss": 0.8028,
      "num_input_tokens_seen": 75015360,
      "step": 129200
    },
    {
      "epoch": 19.24411677092642,
      "grad_norm": 0.036376953125,
      "learning_rate": 0.00013039661520049427,
      "loss": 0.8018,
      "num_input_tokens_seen": 75018464,
      "step": 129205
    },
    {
      "epoch": 19.244861483467382,
      "grad_norm": 0.04296875,
      "learning_rate": 0.0001301402251217948,
      "loss": 0.7961,
      "num_input_tokens_seen": 75021760,
      "step": 129210
    },
    {
      "epoch": 19.245606196008342,
      "grad_norm": 0.0498046875,
      "learning_rate": 0.00012988408625426217,
      "loss": 0.8046,
      "num_input_tokens_seen": 75024864,
      "step": 129215
    },
    {
      "epoch": 19.2463509085493,
      "grad_norm": 0.051513671875,
      "learning_rate": 0.00012962819860222463,
      "loss": 0.7812,
      "num_input_tokens_seen": 75027296,
      "step": 129220
    },
    {
      "epoch": 19.24709562109026,
      "grad_norm": 0.06201171875,
      "learning_rate": 0.00012937256217000537,
      "loss": 0.7727,
      "num_input_tokens_seen": 75030144,
      "step": 129225
    },
    {
      "epoch": 19.247840333631217,
      "grad_norm": 0.032470703125,
      "learning_rate": 0.00012911717696192427,
      "loss": 0.8139,
      "num_input_tokens_seen": 75033088,
      "step": 129230
    },
    {
      "epoch": 19.248585046172177,
      "grad_norm": 0.0537109375,
      "learning_rate": 0.0001288620429822912,
      "loss": 0.7939,
      "num_input_tokens_seen": 75035968,
      "step": 129235
    },
    {
      "epoch": 19.249329758713138,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.00012860716023542273,
      "loss": 0.7785,
      "num_input_tokens_seen": 75038944,
      "step": 129240
    },
    {
      "epoch": 19.250074471254095,
      "grad_norm": 0.058349609375,
      "learning_rate": 0.00012835252872562042,
      "loss": 0.7943,
      "num_input_tokens_seen": 75041984,
      "step": 129245
    },
    {
      "epoch": 19.250819183795056,
      "grad_norm": 0.046630859375,
      "learning_rate": 0.0001280981484571858,
      "loss": 0.8077,
      "num_input_tokens_seen": 75044736,
      "step": 129250
    },
    {
      "epoch": 19.251563896336016,
      "grad_norm": 0.0625,
      "learning_rate": 0.00012784401943442047,
      "loss": 0.8036,
      "num_input_tokens_seen": 75047808,
      "step": 129255
    },
    {
      "epoch": 19.252308608876973,
      "grad_norm": 0.045654296875,
      "learning_rate": 0.00012759014166161263,
      "loss": 0.8073,
      "num_input_tokens_seen": 75050688,
      "step": 129260
    },
    {
      "epoch": 19.253053321417934,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.00012733651514305555,
      "loss": 0.8086,
      "num_input_tokens_seen": 75053760,
      "step": 129265
    },
    {
      "epoch": 19.25379803395889,
      "grad_norm": 0.064453125,
      "learning_rate": 0.00012708313988303078,
      "loss": 0.7939,
      "num_input_tokens_seen": 75056544,
      "step": 129270
    },
    {
      "epoch": 19.25454274649985,
      "grad_norm": 0.060791015625,
      "learning_rate": 0.00012683001588582155,
      "loss": 0.7955,
      "num_input_tokens_seen": 75059520,
      "step": 129275
    },
    {
      "epoch": 19.25528745904081,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.0001265771431557028,
      "loss": 0.7926,
      "num_input_tokens_seen": 75062208,
      "step": 129280
    },
    {
      "epoch": 19.25603217158177,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00012632452169694607,
      "loss": 0.7783,
      "num_input_tokens_seen": 75065184,
      "step": 129285
    },
    {
      "epoch": 19.25677688412273,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.00012607215151381966,
      "loss": 0.7982,
      "num_input_tokens_seen": 75067904,
      "step": 129290
    },
    {
      "epoch": 19.257521596663686,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.00012582003261058682,
      "loss": 0.8093,
      "num_input_tokens_seen": 75070848,
      "step": 129295
    },
    {
      "epoch": 19.258266309204647,
      "grad_norm": 0.04736328125,
      "learning_rate": 0.00012556816499150746,
      "loss": 0.7808,
      "num_input_tokens_seen": 75073984,
      "step": 129300
    },
    {
      "epoch": 19.259011021745607,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.00012531654866083485,
      "loss": 0.8046,
      "num_input_tokens_seen": 75076768,
      "step": 129305
    },
    {
      "epoch": 19.259755734286564,
      "grad_norm": 0.0390625,
      "learning_rate": 0.00012506518362282227,
      "loss": 0.8015,
      "num_input_tokens_seen": 75079968,
      "step": 129310
    },
    {
      "epoch": 19.260500446827525,
      "grad_norm": 0.037109375,
      "learning_rate": 0.00012481406988171462,
      "loss": 0.7875,
      "num_input_tokens_seen": 75082592,
      "step": 129315
    },
    {
      "epoch": 19.261245159368485,
      "grad_norm": 0.048095703125,
      "learning_rate": 0.00012456320744175687,
      "loss": 0.8142,
      "num_input_tokens_seen": 75085696,
      "step": 129320
    },
    {
      "epoch": 19.261989871909442,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00012431259630718227,
      "loss": 0.7838,
      "num_input_tokens_seen": 75089184,
      "step": 129325
    },
    {
      "epoch": 19.262734584450403,
      "grad_norm": 0.04638671875,
      "learning_rate": 0.00012406223648222912,
      "loss": 0.8058,
      "num_input_tokens_seen": 75091872,
      "step": 129330
    },
    {
      "epoch": 19.26347929699136,
      "grad_norm": 0.03173828125,
      "learning_rate": 0.00012381212797112572,
      "loss": 0.8059,
      "num_input_tokens_seen": 75094944,
      "step": 129335
    },
    {
      "epoch": 19.26422400953232,
      "grad_norm": 0.037841796875,
      "learning_rate": 0.00012356227077809533,
      "loss": 0.7953,
      "num_input_tokens_seen": 75097568,
      "step": 129340
    },
    {
      "epoch": 19.26496872207328,
      "grad_norm": 0.04833984375,
      "learning_rate": 0.00012331266490736125,
      "loss": 0.7809,
      "num_input_tokens_seen": 75100160,
      "step": 129345
    },
    {
      "epoch": 19.265713434614238,
      "grad_norm": 0.03955078125,
      "learning_rate": 0.0001230633103631401,
      "loss": 0.7956,
      "num_input_tokens_seen": 75103232,
      "step": 129350
    },
    {
      "epoch": 19.2664581471552,
      "grad_norm": 0.045166015625,
      "learning_rate": 0.00012281420714964353,
      "loss": 0.7971,
      "num_input_tokens_seen": 75106656,
      "step": 129355
    },
    {
      "epoch": 19.26720285969616,
      "grad_norm": 0.025634765625,
      "learning_rate": 0.00012256535527108148,
      "loss": 0.8313,
      "num_input_tokens_seen": 75109504,
      "step": 129360
    },
    {
      "epoch": 19.267947572237116,
      "grad_norm": 0.064453125,
      "learning_rate": 0.0001223167547316556,
      "loss": 0.7762,
      "num_input_tokens_seen": 75112384,
      "step": 129365
    },
    {
      "epoch": 19.268692284778076,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.00012206840553556753,
      "loss": 0.8116,
      "num_input_tokens_seen": 75115648,
      "step": 129370
    },
    {
      "epoch": 19.269436997319033,
      "grad_norm": 0.033935546875,
      "learning_rate": 0.00012182030768701223,
      "loss": 0.7923,
      "num_input_tokens_seen": 75118496,
      "step": 129375
    },
    {
      "epoch": 19.270181709859994,
      "grad_norm": 0.032958984375,
      "learning_rate": 0.00012157246119018139,
      "loss": 0.799,
      "num_input_tokens_seen": 75121088,
      "step": 129380
    },
    {
      "epoch": 19.270926422400954,
      "grad_norm": 0.07421875,
      "learning_rate": 0.00012132486604926329,
      "loss": 0.7971,
      "num_input_tokens_seen": 75123776,
      "step": 129385
    },
    {
      "epoch": 19.27167113494191,
      "grad_norm": 0.053466796875,
      "learning_rate": 0.00012107752226843792,
      "loss": 0.7996,
      "num_input_tokens_seen": 75126912,
      "step": 129390
    },
    {
      "epoch": 19.272415847482872,
      "grad_norm": 0.027099609375,
      "learning_rate": 0.00012083042985188696,
      "loss": 0.8045,
      "num_input_tokens_seen": 75129696,
      "step": 129395
    },
    {
      "epoch": 19.273160560023832,
      "grad_norm": 0.026123046875,
      "learning_rate": 0.00012058358880378205,
      "loss": 0.8028,
      "num_input_tokens_seen": 75132512,
      "step": 129400
    },
    {
      "epoch": 19.27390527256479,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.00012033699912829653,
      "loss": 0.7883,
      "num_input_tokens_seen": 75135360,
      "step": 129405
    },
    {
      "epoch": 19.27464998510575,
      "grad_norm": 0.05859375,
      "learning_rate": 0.00012009066082959373,
      "loss": 0.8094,
      "num_input_tokens_seen": 75138240,
      "step": 129410
    },
    {
      "epoch": 19.275394697646707,
      "grad_norm": 0.033447265625,
      "learning_rate": 0.00011984457391183367,
      "loss": 0.7738,
      "num_input_tokens_seen": 75141152,
      "step": 129415
    },
    {
      "epoch": 19.276139410187668,
      "grad_norm": 0.0458984375,
      "learning_rate": 0.000119598738379178,
      "loss": 0.7824,
      "num_input_tokens_seen": 75143840,
      "step": 129420
    },
    {
      "epoch": 19.276884122728628,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.00011935315423577841,
      "loss": 0.7731,
      "num_input_tokens_seen": 75146592,
      "step": 129425
    },
    {
      "epoch": 19.277628835269585,
      "grad_norm": 0.0634765625,
      "learning_rate": 0.00011910782148578158,
      "loss": 0.7863,
      "num_input_tokens_seen": 75149216,
      "step": 129430
    },
    {
      "epoch": 19.278373547810546,
      "grad_norm": 0.03857421875,
      "learning_rate": 0.00011886274013333586,
      "loss": 0.786,
      "num_input_tokens_seen": 75151744,
      "step": 129435
    },
    {
      "epoch": 19.279118260351503,
      "grad_norm": 0.0361328125,
      "learning_rate": 0.00011861791018257794,
      "loss": 0.7894,
      "num_input_tokens_seen": 75154688,
      "step": 129440
    },
    {
      "epoch": 19.279862972892463,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.00011837333163764618,
      "loss": 0.7889,
      "num_input_tokens_seen": 75157408,
      "step": 129445
    },
    {
      "epoch": 19.280607685433424,
      "grad_norm": 0.044921875,
      "learning_rate": 0.00011812900450267228,
      "loss": 0.7952,
      "num_input_tokens_seen": 75160352,
      "step": 129450
    },
    {
      "epoch": 19.28135239797438,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.00011788492878178291,
      "loss": 0.79,
      "num_input_tokens_seen": 75163328,
      "step": 129455
    },
    {
      "epoch": 19.28209711051534,
      "grad_norm": 0.052734375,
      "learning_rate": 0.00011764110447910314,
      "loss": 0.7929,
      "num_input_tokens_seen": 75166176,
      "step": 129460
    },
    {
      "epoch": 19.2828418230563,
      "grad_norm": 0.025390625,
      "learning_rate": 0.00011739753159874965,
      "loss": 0.792,
      "num_input_tokens_seen": 75168896,
      "step": 129465
    },
    {
      "epoch": 19.28358653559726,
      "grad_norm": 0.046875,
      "learning_rate": 0.00011715421014484084,
      "loss": 0.7908,
      "num_input_tokens_seen": 75172160,
      "step": 129470
    },
    {
      "epoch": 19.28433124813822,
      "grad_norm": 0.369140625,
      "learning_rate": 0.00011691114012148506,
      "loss": 0.8161,
      "num_input_tokens_seen": 75175072,
      "step": 129475
    },
    {
      "epoch": 19.285075960679176,
      "grad_norm": 0.038330078125,
      "learning_rate": 0.00011666832153278906,
      "loss": 0.8095,
      "num_input_tokens_seen": 75178144,
      "step": 129480
    },
    {
      "epoch": 19.285820673220137,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.00011642575438285451,
      "loss": 0.7796,
      "num_input_tokens_seen": 75180800,
      "step": 129485
    },
    {
      "epoch": 19.286565385761097,
      "grad_norm": 0.0419921875,
      "learning_rate": 0.00011618343867577818,
      "loss": 0.7947,
      "num_input_tokens_seen": 75183360,
      "step": 129490
    },
    {
      "epoch": 19.287310098302054,
      "grad_norm": 0.040283203125,
      "learning_rate": 0.00011594137441565844,
      "loss": 0.8096,
      "num_input_tokens_seen": 75186240,
      "step": 129495
    },
    {
      "epoch": 19.288054810843015,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.00011569956160657868,
      "loss": 0.7949,
      "num_input_tokens_seen": 75189088,
      "step": 129500
    },
    {
      "epoch": 19.288799523383975,
      "grad_norm": 0.04296875,
      "learning_rate": 0.00011545800025262898,
      "loss": 0.8041,
      "num_input_tokens_seen": 75192064,
      "step": 129505
    },
    {
      "epoch": 19.289544235924932,
      "grad_norm": 0.0341796875,
      "learning_rate": 0.00011521669035788772,
      "loss": 0.8078,
      "num_input_tokens_seen": 75195136,
      "step": 129510
    },
    {
      "epoch": 19.290288948465893,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.00011497563192643334,
      "loss": 0.8017,
      "num_input_tokens_seen": 75198080,
      "step": 129515
    },
    {
      "epoch": 19.29103366100685,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.00011473482496233589,
      "loss": 0.801,
      "num_input_tokens_seen": 75201184,
      "step": 129520
    },
    {
      "epoch": 19.29177837354781,
      "grad_norm": 0.031494140625,
      "learning_rate": 0.00011449426946966545,
      "loss": 0.8088,
      "num_input_tokens_seen": 75204064,
      "step": 129525
    },
    {
      "epoch": 19.29252308608877,
      "grad_norm": 0.057861328125,
      "learning_rate": 0.00011425396545248545,
      "loss": 0.7945,
      "num_input_tokens_seen": 75207328,
      "step": 129530
    },
    {
      "epoch": 19.293267798629728,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.00011401391291485596,
      "loss": 0.8012,
      "num_input_tokens_seen": 75210400,
      "step": 129535
    },
    {
      "epoch": 19.29401251117069,
      "grad_norm": 0.07177734375,
      "learning_rate": 0.00011377411186083208,
      "loss": 0.8014,
      "num_input_tokens_seen": 75213248,
      "step": 129540
    },
    {
      "epoch": 19.29475722371165,
      "grad_norm": 0.03515625,
      "learning_rate": 0.0001135345622944639,
      "loss": 0.7998,
      "num_input_tokens_seen": 75215936,
      "step": 129545
    },
    {
      "epoch": 19.295501936252606,
      "grad_norm": 0.058837890625,
      "learning_rate": 0.00011329526421979985,
      "loss": 0.7949,
      "num_input_tokens_seen": 75218464,
      "step": 129550
    },
    {
      "epoch": 19.296246648793566,
      "grad_norm": 0.06591796875,
      "learning_rate": 0.00011305621764088169,
      "loss": 0.7924,
      "num_input_tokens_seen": 75221024,
      "step": 129555
    },
    {
      "epoch": 19.296991361334523,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.00011281742256174954,
      "loss": 0.7895,
      "num_input_tokens_seen": 75223808,
      "step": 129560
    },
    {
      "epoch": 19.297736073875484,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.00011257887898643514,
      "loss": 0.782,
      "num_input_tokens_seen": 75226784,
      "step": 129565
    },
    {
      "epoch": 19.298480786416444,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.00011234058691896864,
      "loss": 0.798,
      "num_input_tokens_seen": 75229632,
      "step": 129570
    },
    {
      "epoch": 19.2992254989574,
      "grad_norm": 0.04052734375,
      "learning_rate": 0.00011210254636337845,
      "loss": 0.8014,
      "num_input_tokens_seen": 75232608,
      "step": 129575
    },
    {
      "epoch": 19.299970211498362,
      "grad_norm": 0.036865234375,
      "learning_rate": 0.00011186475732368472,
      "loss": 0.8041,
      "num_input_tokens_seen": 75235264,
      "step": 129580
    },
    {
      "epoch": 19.300714924039323,
      "grad_norm": 0.043212890625,
      "learning_rate": 0.00011162721980390421,
      "loss": 0.815,
      "num_input_tokens_seen": 75238080,
      "step": 129585
    },
    {
      "epoch": 19.30145963658028,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.00011138993380804874,
      "loss": 0.8007,
      "num_input_tokens_seen": 75240832,
      "step": 129590
    },
    {
      "epoch": 19.30220434912124,
      "grad_norm": 0.0556640625,
      "learning_rate": 0.00011115289934012839,
      "loss": 0.7985,
      "num_input_tokens_seen": 75243744,
      "step": 129595
    },
    {
      "epoch": 19.302949061662197,
      "grad_norm": 0.029541015625,
      "learning_rate": 0.00011091611640414833,
      "loss": 0.7933,
      "num_input_tokens_seen": 75246688,
      "step": 129600
    },
    {
      "epoch": 19.303693774203158,
      "grad_norm": 0.034423828125,
      "learning_rate": 0.00011067958500410701,
      "loss": 0.8066,
      "num_input_tokens_seen": 75249856,
      "step": 129605
    },
    {
      "epoch": 19.304438486744118,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.0001104433051440029,
      "loss": 0.8037,
      "num_input_tokens_seen": 75252640,
      "step": 129610
    },
    {
      "epoch": 19.305183199285075,
      "grad_norm": 0.05615234375,
      "learning_rate": 0.0001102072768278245,
      "loss": 0.8167,
      "num_input_tokens_seen": 75255648,
      "step": 129615
    },
    {
      "epoch": 19.305927911826036,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.00010997150005956191,
      "loss": 0.7907,
      "num_input_tokens_seen": 75258592,
      "step": 129620
    },
    {
      "epoch": 19.306672624366993,
      "grad_norm": 0.0546875,
      "learning_rate": 0.0001097359748431953,
      "loss": 0.8336,
      "num_input_tokens_seen": 75261472,
      "step": 129625
    },
    {
      "epoch": 19.307417336907953,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.00010950070118270816,
      "loss": 0.8076,
      "num_input_tokens_seen": 75264288,
      "step": 129630
    },
    {
      "epoch": 19.308162049448914,
      "grad_norm": 0.041015625,
      "learning_rate": 0.00010926567908207062,
      "loss": 0.7928,
      "num_input_tokens_seen": 75267328,
      "step": 129635
    },
    {
      "epoch": 19.30890676198987,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001090309085452562,
      "loss": 0.801,
      "num_input_tokens_seen": 75270272,
      "step": 129640
    },
    {
      "epoch": 19.30965147453083,
      "grad_norm": 0.053955078125,
      "learning_rate": 0.00010879638957623005,
      "loss": 0.7891,
      "num_input_tokens_seen": 75273120,
      "step": 129645
    },
    {
      "epoch": 19.31039618707179,
      "grad_norm": 0.072265625,
      "learning_rate": 0.000108562122178954,
      "loss": 0.7973,
      "num_input_tokens_seen": 75276352,
      "step": 129650
    },
    {
      "epoch": 19.31114089961275,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.00010832810635738654,
      "loss": 0.786,
      "num_input_tokens_seen": 75279232,
      "step": 129655
    },
    {
      "epoch": 19.31188561215371,
      "grad_norm": 0.06982421875,
      "learning_rate": 0.00010809434211547953,
      "loss": 0.7878,
      "num_input_tokens_seen": 75281984,
      "step": 129660
    },
    {
      "epoch": 19.312630324694666,
      "grad_norm": 0.0517578125,
      "learning_rate": 0.00010786082945718311,
      "loss": 0.8081,
      "num_input_tokens_seen": 75284832,
      "step": 129665
    },
    {
      "epoch": 19.313375037235627,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.00010762756838644249,
      "loss": 0.785,
      "num_input_tokens_seen": 75287680,
      "step": 129670
    },
    {
      "epoch": 19.314119749776587,
      "grad_norm": 0.04296875,
      "learning_rate": 0.00010739455890719785,
      "loss": 0.8012,
      "num_input_tokens_seen": 75290400,
      "step": 129675
    },
    {
      "epoch": 19.314864462317544,
      "grad_norm": 0.050048828125,
      "learning_rate": 0.00010716180102338767,
      "loss": 0.796,
      "num_input_tokens_seen": 75293472,
      "step": 129680
    },
    {
      "epoch": 19.315609174858505,
      "grad_norm": 0.03564453125,
      "learning_rate": 0.00010692929473894052,
      "loss": 0.7985,
      "num_input_tokens_seen": 75296352,
      "step": 129685
    },
    {
      "epoch": 19.316353887399465,
      "grad_norm": 0.03466796875,
      "learning_rate": 0.00010669704005778823,
      "loss": 0.8024,
      "num_input_tokens_seen": 75299456,
      "step": 129690
    },
    {
      "epoch": 19.317098599940422,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.000106465036983851,
      "loss": 0.7936,
      "num_input_tokens_seen": 75302272,
      "step": 129695
    },
    {
      "epoch": 19.317843312481383,
      "grad_norm": 0.046142578125,
      "learning_rate": 0.00010623328552105072,
      "loss": 0.7961,
      "num_input_tokens_seen": 75305184,
      "step": 129700
    },
    {
      "epoch": 19.31858802502234,
      "grad_norm": 0.052734375,
      "learning_rate": 0.00010600178567329921,
      "loss": 0.7932,
      "num_input_tokens_seen": 75307936,
      "step": 129705
    },
    {
      "epoch": 19.3193327375633,
      "grad_norm": 0.06494140625,
      "learning_rate": 0.00010577053744451337,
      "loss": 0.8193,
      "num_input_tokens_seen": 75310848,
      "step": 129710
    },
    {
      "epoch": 19.32007745010426,
      "grad_norm": 0.05810546875,
      "learning_rate": 0.0001055395408385934,
      "loss": 0.802,
      "num_input_tokens_seen": 75313664,
      "step": 129715
    },
    {
      "epoch": 19.320822162645218,
      "grad_norm": 0.038818359375,
      "learning_rate": 0.00010530879585944786,
      "loss": 0.7769,
      "num_input_tokens_seen": 75316640,
      "step": 129720
    },
    {
      "epoch": 19.32156687518618,
      "grad_norm": 0.049072265625,
      "learning_rate": 0.00010507830251097028,
      "loss": 0.8007,
      "num_input_tokens_seen": 75319680,
      "step": 129725
    },
    {
      "epoch": 19.32231158772714,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.00010484806079705588,
      "loss": 0.8097,
      "num_input_tokens_seen": 75322912,
      "step": 129730
    },
    {
      "epoch": 19.323056300268096,
      "grad_norm": 0.06103515625,
      "learning_rate": 0.00010461807072159657,
      "loss": 0.8025,
      "num_input_tokens_seen": 75326272,
      "step": 129735
    },
    {
      "epoch": 19.323801012809056,
      "grad_norm": 0.0703125,
      "learning_rate": 0.00010438833228847422,
      "loss": 0.8062,
      "num_input_tokens_seen": 75329248,
      "step": 129740
    },
    {
      "epoch": 19.324545725350013,
      "grad_norm": 0.04541015625,
      "learning_rate": 0.00010415884550157239,
      "loss": 0.8146,
      "num_input_tokens_seen": 75332128,
      "step": 129745
    },
    {
      "epoch": 19.325290437890974,
      "grad_norm": 0.060302734375,
      "learning_rate": 0.00010392961036476799,
      "loss": 0.7818,
      "num_input_tokens_seen": 75334784,
      "step": 129750
    },
    {
      "epoch": 19.326035150431935,
      "grad_norm": 0.052978515625,
      "learning_rate": 0.00010370062688193293,
      "loss": 0.8022,
      "num_input_tokens_seen": 75337888,
      "step": 129755
    },
    {
      "epoch": 19.32677986297289,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.0001034718950569341,
      "loss": 0.7956,
      "num_input_tokens_seen": 75340672,
      "step": 129760
    },
    {
      "epoch": 19.327524575513852,
      "grad_norm": 0.025390625,
      "learning_rate": 0.00010324341489364007,
      "loss": 0.8147,
      "num_input_tokens_seen": 75343520,
      "step": 129765
    },
    {
      "epoch": 19.328269288054813,
      "grad_norm": 0.048828125,
      "learning_rate": 0.00010301518639590612,
      "loss": 0.8186,
      "num_input_tokens_seen": 75346304,
      "step": 129770
    },
    {
      "epoch": 19.32901400059577,
      "grad_norm": 0.039794921875,
      "learning_rate": 0.0001027872095675908,
      "loss": 0.8002,
      "num_input_tokens_seen": 75349024,
      "step": 129775
    },
    {
      "epoch": 19.32975871313673,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.00010255948441254436,
      "loss": 0.7937,
      "num_input_tokens_seen": 75352064,
      "step": 129780
    },
    {
      "epoch": 19.330503425677687,
      "grad_norm": 0.033203125,
      "learning_rate": 0.0001023320109346154,
      "loss": 0.7948,
      "num_input_tokens_seen": 75354880,
      "step": 129785
    },
    {
      "epoch": 19.331248138218648,
      "grad_norm": 0.03662109375,
      "learning_rate": 0.0001021047891376442,
      "loss": 0.8143,
      "num_input_tokens_seen": 75357600,
      "step": 129790
    },
    {
      "epoch": 19.331992850759608,
      "grad_norm": 0.044189453125,
      "learning_rate": 0.00010187781902547266,
      "loss": 0.8049,
      "num_input_tokens_seen": 75360384,
      "step": 129795
    },
    {
      "epoch": 19.332737563300565,
      "grad_norm": 0.041259765625,
      "learning_rate": 0.00010165110060193272,
      "loss": 0.7978,
      "num_input_tokens_seen": 75363296,
      "step": 129800
    },
    {
      "epoch": 19.333482275841526,
      "grad_norm": 0.103515625,
      "learning_rate": 0.00010142463387085465,
      "loss": 0.7877,
      "num_input_tokens_seen": 75366080,
      "step": 129805
    },
    {
      "epoch": 19.334226988382483,
      "grad_norm": 0.042724609375,
      "learning_rate": 0.00010119841883606705,
      "loss": 0.7727,
      "num_input_tokens_seen": 75369216,
      "step": 129810
    },
    {
      "epoch": 19.334971700923443,
      "grad_norm": 0.054443359375,
      "learning_rate": 0.00010097245550138856,
      "loss": 0.7878,
      "num_input_tokens_seen": 75371968,
      "step": 129815
    },
    {
      "epoch": 19.335716413464404,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.00010074674387063775,
      "loss": 0.7886,
      "num_input_tokens_seen": 75374784,
      "step": 129820
    },
    {
      "epoch": 19.33646112600536,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.00010052128394762826,
      "loss": 0.795,
      "num_input_tokens_seen": 75377888,
      "step": 129825
    },
    {
      "epoch": 19.33720583854632,
      "grad_norm": 0.0478515625,
      "learning_rate": 0.00010029607573616706,
      "loss": 0.8055,
      "num_input_tokens_seen": 75380672,
      "step": 129830
    },
    {
      "epoch": 19.337950551087282,
      "grad_norm": 0.039306640625,
      "learning_rate": 0.00010007111924006273,
      "loss": 0.7971,
      "num_input_tokens_seen": 75383552,
      "step": 129835
    },
    {
      "epoch": 19.33869526362824,
      "grad_norm": 0.04345703125,
      "learning_rate": 9.98464144631106e-05,
      "loss": 0.8049,
      "num_input_tokens_seen": 75386688,
      "step": 129840
    },
    {
      "epoch": 19.3394399761692,
      "grad_norm": 0.042236328125,
      "learning_rate": 9.962196140911095e-05,
      "loss": 0.8039,
      "num_input_tokens_seen": 75389248,
      "step": 129845
    },
    {
      "epoch": 19.340184688710156,
      "grad_norm": 0.0537109375,
      "learning_rate": 9.939776008185408e-05,
      "loss": 0.7969,
      "num_input_tokens_seen": 75392192,
      "step": 129850
    },
    {
      "epoch": 19.340929401251117,
      "grad_norm": 0.0810546875,
      "learning_rate": 9.917381048512863e-05,
      "loss": 0.805,
      "num_input_tokens_seen": 75395072,
      "step": 129855
    },
    {
      "epoch": 19.341674113792077,
      "grad_norm": 0.04931640625,
      "learning_rate": 9.895011262271491e-05,
      "loss": 0.8144,
      "num_input_tokens_seen": 75398112,
      "step": 129860
    },
    {
      "epoch": 19.342418826333034,
      "grad_norm": 0.07958984375,
      "learning_rate": 9.872666649839656e-05,
      "loss": 0.8187,
      "num_input_tokens_seen": 75400960,
      "step": 129865
    },
    {
      "epoch": 19.343163538873995,
      "grad_norm": 0.095703125,
      "learning_rate": 9.85034721159439e-05,
      "loss": 0.7836,
      "num_input_tokens_seen": 75404032,
      "step": 129870
    },
    {
      "epoch": 19.343908251414955,
      "grad_norm": 0.046875,
      "learning_rate": 9.828052947913224e-05,
      "loss": 0.7963,
      "num_input_tokens_seen": 75406816,
      "step": 129875
    },
    {
      "epoch": 19.344652963955912,
      "grad_norm": 0.025634765625,
      "learning_rate": 9.805783859172524e-05,
      "loss": 0.7992,
      "num_input_tokens_seen": 75409952,
      "step": 129880
    },
    {
      "epoch": 19.345397676496873,
      "grad_norm": 0.0361328125,
      "learning_rate": 9.783539945748487e-05,
      "loss": 0.8014,
      "num_input_tokens_seen": 75412640,
      "step": 129885
    },
    {
      "epoch": 19.34614238903783,
      "grad_norm": 0.049560546875,
      "learning_rate": 9.761321208016815e-05,
      "loss": 0.7809,
      "num_input_tokens_seen": 75415488,
      "step": 129890
    },
    {
      "epoch": 19.34688710157879,
      "grad_norm": 0.037353515625,
      "learning_rate": 9.739127646353207e-05,
      "loss": 0.7995,
      "num_input_tokens_seen": 75418240,
      "step": 129895
    },
    {
      "epoch": 19.34763181411975,
      "grad_norm": 0.038330078125,
      "learning_rate": 9.716959261132196e-05,
      "loss": 0.7983,
      "num_input_tokens_seen": 75421440,
      "step": 129900
    },
    {
      "epoch": 19.348376526660708,
      "grad_norm": 0.031982421875,
      "learning_rate": 9.69481605272865e-05,
      "loss": 0.8035,
      "num_input_tokens_seen": 75424256,
      "step": 129905
    },
    {
      "epoch": 19.34912123920167,
      "grad_norm": 0.050048828125,
      "learning_rate": 9.672698021516602e-05,
      "loss": 0.8055,
      "num_input_tokens_seen": 75427072,
      "step": 129910
    },
    {
      "epoch": 19.34986595174263,
      "grad_norm": 0.045654296875,
      "learning_rate": 9.650605167869419e-05,
      "loss": 0.7957,
      "num_input_tokens_seen": 75429568,
      "step": 129915
    },
    {
      "epoch": 19.350610664283586,
      "grad_norm": 0.11474609375,
      "learning_rate": 9.628537492160638e-05,
      "loss": 0.8016,
      "num_input_tokens_seen": 75432416,
      "step": 129920
    },
    {
      "epoch": 19.351355376824547,
      "grad_norm": 0.041015625,
      "learning_rate": 9.60649499476296e-05,
      "loss": 0.7964,
      "num_input_tokens_seen": 75435424,
      "step": 129925
    },
    {
      "epoch": 19.352100089365504,
      "grad_norm": 0.050537109375,
      "learning_rate": 9.584477676048753e-05,
      "loss": 0.7927,
      "num_input_tokens_seen": 75438336,
      "step": 129930
    },
    {
      "epoch": 19.352844801906464,
      "grad_norm": 0.0272216796875,
      "learning_rate": 9.562485536390052e-05,
      "loss": 0.8039,
      "num_input_tokens_seen": 75441024,
      "step": 129935
    },
    {
      "epoch": 19.353589514447425,
      "grad_norm": 0.0419921875,
      "learning_rate": 9.540518576158396e-05,
      "loss": 0.7834,
      "num_input_tokens_seen": 75444000,
      "step": 129940
    },
    {
      "epoch": 19.35433422698838,
      "grad_norm": 0.060302734375,
      "learning_rate": 9.51857679572482e-05,
      "loss": 0.8099,
      "num_input_tokens_seen": 75446944,
      "step": 129945
    },
    {
      "epoch": 19.355078939529342,
      "grad_norm": 0.05615234375,
      "learning_rate": 9.496660195460027e-05,
      "loss": 0.8014,
      "num_input_tokens_seen": 75449888,
      "step": 129950
    },
    {
      "epoch": 19.3558236520703,
      "grad_norm": 0.050537109375,
      "learning_rate": 9.474768775734055e-05,
      "loss": 0.8048,
      "num_input_tokens_seen": 75452768,
      "step": 129955
    },
    {
      "epoch": 19.35656836461126,
      "grad_norm": 0.034423828125,
      "learning_rate": 9.452902536917273e-05,
      "loss": 0.7901,
      "num_input_tokens_seen": 75455584,
      "step": 129960
    },
    {
      "epoch": 19.35731307715222,
      "grad_norm": 0.0341796875,
      "learning_rate": 9.431061479378722e-05,
      "loss": 0.8077,
      "num_input_tokens_seen": 75458176,
      "step": 129965
    },
    {
      "epoch": 19.358057789693177,
      "grad_norm": 0.035400390625,
      "learning_rate": 9.40924560348727e-05,
      "loss": 0.7897,
      "num_input_tokens_seen": 75461056,
      "step": 129970
    },
    {
      "epoch": 19.358802502234138,
      "grad_norm": 0.052001953125,
      "learning_rate": 9.387454909611792e-05,
      "loss": 0.8092,
      "num_input_tokens_seen": 75463808,
      "step": 129975
    },
    {
      "epoch": 19.359547214775098,
      "grad_norm": 0.0654296875,
      "learning_rate": 9.365689398120158e-05,
      "loss": 0.8025,
      "num_input_tokens_seen": 75467040,
      "step": 129980
    },
    {
      "epoch": 19.360291927316055,
      "grad_norm": 0.1162109375,
      "learning_rate": 9.343949069380242e-05,
      "loss": 0.7914,
      "num_input_tokens_seen": 75469824,
      "step": 129985
    },
    {
      "epoch": 19.361036639857016,
      "grad_norm": 0.036865234375,
      "learning_rate": 9.322233923759248e-05,
      "loss": 0.7945,
      "num_input_tokens_seen": 75472800,
      "step": 129990
    },
    {
      "epoch": 19.361781352397973,
      "grad_norm": 0.05078125,
      "learning_rate": 9.300543961624053e-05,
      "loss": 0.7896,
      "num_input_tokens_seen": 75475680,
      "step": 129995
    },
    {
      "epoch": 19.362526064938933,
      "grad_norm": 0.11279296875,
      "learning_rate": 9.278879183341193e-05,
      "loss": 0.8081,
      "num_input_tokens_seen": 75478368,
      "step": 130000
    },
    {
      "epoch": 19.363270777479894,
      "grad_norm": 0.03955078125,
      "learning_rate": 9.257239589276545e-05,
      "loss": 0.7963,
      "num_input_tokens_seen": 75481024,
      "step": 130005
    },
    {
      "epoch": 19.36401549002085,
      "grad_norm": 0.09326171875,
      "learning_rate": 9.235625179795648e-05,
      "loss": 0.8168,
      "num_input_tokens_seen": 75483936,
      "step": 130010
    },
    {
      "epoch": 19.36476020256181,
      "grad_norm": 0.03759765625,
      "learning_rate": 9.21403595526371e-05,
      "loss": 0.7959,
      "num_input_tokens_seen": 75487008,
      "step": 130015
    },
    {
      "epoch": 19.365504915102772,
      "grad_norm": 0.04833984375,
      "learning_rate": 9.192471916045608e-05,
      "loss": 0.7671,
      "num_input_tokens_seen": 75489824,
      "step": 130020
    },
    {
      "epoch": 19.36624962764373,
      "grad_norm": 0.04296875,
      "learning_rate": 9.170933062505382e-05,
      "loss": 0.7967,
      "num_input_tokens_seen": 75492640,
      "step": 130025
    },
    {
      "epoch": 19.36699434018469,
      "grad_norm": 0.040771484375,
      "learning_rate": 9.149419395006907e-05,
      "loss": 0.7923,
      "num_input_tokens_seen": 75495520,
      "step": 130030
    },
    {
      "epoch": 19.367739052725646,
      "grad_norm": 0.050048828125,
      "learning_rate": 9.127930913913895e-05,
      "loss": 0.7899,
      "num_input_tokens_seen": 75498784,
      "step": 130035
    },
    {
      "epoch": 19.368483765266607,
      "grad_norm": 0.045654296875,
      "learning_rate": 9.10646761958922e-05,
      "loss": 0.8038,
      "num_input_tokens_seen": 75501600,
      "step": 130040
    },
    {
      "epoch": 19.369228477807567,
      "grad_norm": 0.07568359375,
      "learning_rate": 9.085029512395593e-05,
      "loss": 0.7885,
      "num_input_tokens_seen": 75504736,
      "step": 130045
    },
    {
      "epoch": 19.369973190348524,
      "grad_norm": 0.04833984375,
      "learning_rate": 9.063616592695056e-05,
      "loss": 0.7986,
      "num_input_tokens_seen": 75507520,
      "step": 130050
    },
    {
      "epoch": 19.370717902889485,
      "grad_norm": 0.06884765625,
      "learning_rate": 9.042228860849322e-05,
      "loss": 0.7954,
      "num_input_tokens_seen": 75510400,
      "step": 130055
    },
    {
      "epoch": 19.371462615430445,
      "grad_norm": 0.048583984375,
      "learning_rate": 9.020866317219766e-05,
      "loss": 0.8362,
      "num_input_tokens_seen": 75513408,
      "step": 130060
    },
    {
      "epoch": 19.372207327971402,
      "grad_norm": 0.0458984375,
      "learning_rate": 8.999528962167435e-05,
      "loss": 0.7962,
      "num_input_tokens_seen": 75516256,
      "step": 130065
    },
    {
      "epoch": 19.372952040512363,
      "grad_norm": 0.0654296875,
      "learning_rate": 8.97821679605254e-05,
      "loss": 0.8181,
      "num_input_tokens_seen": 75519328,
      "step": 130070
    },
    {
      "epoch": 19.37369675305332,
      "grad_norm": 0.0380859375,
      "learning_rate": 8.956929819235293e-05,
      "loss": 0.8134,
      "num_input_tokens_seen": 75522528,
      "step": 130075
    },
    {
      "epoch": 19.37444146559428,
      "grad_norm": 0.0439453125,
      "learning_rate": 8.935668032075405e-05,
      "loss": 0.7993,
      "num_input_tokens_seen": 75525440,
      "step": 130080
    },
    {
      "epoch": 19.37518617813524,
      "grad_norm": 0.06982421875,
      "learning_rate": 8.914431434931924e-05,
      "loss": 0.7964,
      "num_input_tokens_seen": 75528288,
      "step": 130085
    },
    {
      "epoch": 19.375930890676198,
      "grad_norm": 0.0439453125,
      "learning_rate": 8.893220028163562e-05,
      "loss": 0.8048,
      "num_input_tokens_seen": 75531168,
      "step": 130090
    },
    {
      "epoch": 19.37667560321716,
      "grad_norm": 0.044921875,
      "learning_rate": 8.872033812128698e-05,
      "loss": 0.7864,
      "num_input_tokens_seen": 75533920,
      "step": 130095
    },
    {
      "epoch": 19.37742031575812,
      "grad_norm": 0.04541015625,
      "learning_rate": 8.850872787185548e-05,
      "loss": 0.8123,
      "num_input_tokens_seen": 75536736,
      "step": 130100
    },
    {
      "epoch": 19.378165028299076,
      "grad_norm": 0.0263671875,
      "learning_rate": 8.829736953691158e-05,
      "loss": 0.8043,
      "num_input_tokens_seen": 75539488,
      "step": 130105
    },
    {
      "epoch": 19.378909740840037,
      "grad_norm": 0.0380859375,
      "learning_rate": 8.80862631200291e-05,
      "loss": 0.7939,
      "num_input_tokens_seen": 75542656,
      "step": 130110
    },
    {
      "epoch": 19.379654453380994,
      "grad_norm": 0.033935546875,
      "learning_rate": 8.787540862477183e-05,
      "loss": 0.7747,
      "num_input_tokens_seen": 75545248,
      "step": 130115
    },
    {
      "epoch": 19.380399165921954,
      "grad_norm": 0.029296875,
      "learning_rate": 8.766480605470527e-05,
      "loss": 0.8041,
      "num_input_tokens_seen": 75547872,
      "step": 130120
    },
    {
      "epoch": 19.381143878462915,
      "grad_norm": 0.0634765625,
      "learning_rate": 8.745445541338492e-05,
      "loss": 0.798,
      "num_input_tokens_seen": 75550976,
      "step": 130125
    },
    {
      "epoch": 19.38188859100387,
      "grad_norm": 0.04296875,
      "learning_rate": 8.724435670436458e-05,
      "loss": 0.793,
      "num_input_tokens_seen": 75553920,
      "step": 130130
    },
    {
      "epoch": 19.382633303544832,
      "grad_norm": 0.039794921875,
      "learning_rate": 8.703450993119644e-05,
      "loss": 0.8073,
      "num_input_tokens_seen": 75557216,
      "step": 130135
    },
    {
      "epoch": 19.38337801608579,
      "grad_norm": 0.038818359375,
      "learning_rate": 8.682491509742096e-05,
      "loss": 0.7957,
      "num_input_tokens_seen": 75559904,
      "step": 130140
    },
    {
      "epoch": 19.38412272862675,
      "grad_norm": 0.049560546875,
      "learning_rate": 8.661557220658367e-05,
      "loss": 0.7918,
      "num_input_tokens_seen": 75562816,
      "step": 130145
    },
    {
      "epoch": 19.38486744116771,
      "grad_norm": 0.061279296875,
      "learning_rate": 8.640648126221672e-05,
      "loss": 0.7783,
      "num_input_tokens_seen": 75565600,
      "step": 130150
    },
    {
      "epoch": 19.385612153708667,
      "grad_norm": 0.0361328125,
      "learning_rate": 8.619764226785564e-05,
      "loss": 0.7878,
      "num_input_tokens_seen": 75568640,
      "step": 130155
    },
    {
      "epoch": 19.386356866249628,
      "grad_norm": 0.033935546875,
      "learning_rate": 8.59890552270276e-05,
      "loss": 0.8185,
      "num_input_tokens_seen": 75571712,
      "step": 130160
    },
    {
      "epoch": 19.38710157879059,
      "grad_norm": 0.0281982421875,
      "learning_rate": 8.578072014325644e-05,
      "loss": 0.8013,
      "num_input_tokens_seen": 75574592,
      "step": 130165
    },
    {
      "epoch": 19.387846291331545,
      "grad_norm": 0.06640625,
      "learning_rate": 8.557263702006268e-05,
      "loss": 0.7943,
      "num_input_tokens_seen": 75577376,
      "step": 130170
    },
    {
      "epoch": 19.388591003872506,
      "grad_norm": 0.05712890625,
      "learning_rate": 8.536480586096017e-05,
      "loss": 0.801,
      "num_input_tokens_seen": 75580064,
      "step": 130175
    },
    {
      "epoch": 19.389335716413463,
      "grad_norm": 0.05517578125,
      "learning_rate": 8.515722666946113e-05,
      "loss": 0.7911,
      "num_input_tokens_seen": 75582720,
      "step": 130180
    },
    {
      "epoch": 19.390080428954423,
      "grad_norm": 0.0556640625,
      "learning_rate": 8.494989944907105e-05,
      "loss": 0.8094,
      "num_input_tokens_seen": 75585760,
      "step": 130185
    },
    {
      "epoch": 19.390825141495384,
      "grad_norm": 0.0478515625,
      "learning_rate": 8.474282420329381e-05,
      "loss": 0.7964,
      "num_input_tokens_seen": 75588544,
      "step": 130190
    },
    {
      "epoch": 19.39156985403634,
      "grad_norm": 0.03076171875,
      "learning_rate": 8.453600093562829e-05,
      "loss": 0.8145,
      "num_input_tokens_seen": 75591360,
      "step": 130195
    },
    {
      "epoch": 19.3923145665773,
      "grad_norm": 0.047119140625,
      "learning_rate": 8.432942964956668e-05,
      "loss": 0.7861,
      "num_input_tokens_seen": 75594112,
      "step": 130200
    },
    {
      "epoch": 19.393059279118262,
      "grad_norm": 0.03662109375,
      "learning_rate": 8.412311034859954e-05,
      "loss": 0.7966,
      "num_input_tokens_seen": 75596896,
      "step": 130205
    },
    {
      "epoch": 19.39380399165922,
      "grad_norm": 0.056884765625,
      "learning_rate": 8.391704303621405e-05,
      "loss": 0.8048,
      "num_input_tokens_seen": 75599648,
      "step": 130210
    },
    {
      "epoch": 19.39454870420018,
      "grad_norm": 0.0625,
      "learning_rate": 8.371122771588913e-05,
      "loss": 0.7753,
      "num_input_tokens_seen": 75602400,
      "step": 130215
    },
    {
      "epoch": 19.395293416741136,
      "grad_norm": 0.054931640625,
      "learning_rate": 8.350566439110196e-05,
      "loss": 0.8062,
      "num_input_tokens_seen": 75605184,
      "step": 130220
    },
    {
      "epoch": 19.396038129282097,
      "grad_norm": 0.060791015625,
      "learning_rate": 8.330035306532812e-05,
      "loss": 0.7944,
      "num_input_tokens_seen": 75608224,
      "step": 130225
    },
    {
      "epoch": 19.396782841823057,
      "grad_norm": 0.0303955078125,
      "learning_rate": 8.309529374203317e-05,
      "loss": 0.7892,
      "num_input_tokens_seen": 75611136,
      "step": 130230
    },
    {
      "epoch": 19.397527554364014,
      "grad_norm": 0.055419921875,
      "learning_rate": 8.2890486424681e-05,
      "loss": 0.8272,
      "num_input_tokens_seen": 75614112,
      "step": 130235
    },
    {
      "epoch": 19.398272266904975,
      "grad_norm": 0.0289306640625,
      "learning_rate": 8.268593111673383e-05,
      "loss": 0.7894,
      "num_input_tokens_seen": 75616736,
      "step": 130240
    },
    {
      "epoch": 19.399016979445936,
      "grad_norm": 0.0498046875,
      "learning_rate": 8.248162782164725e-05,
      "loss": 0.781,
      "num_input_tokens_seen": 75619712,
      "step": 130245
    },
    {
      "epoch": 19.399761691986892,
      "grad_norm": 0.031982421875,
      "learning_rate": 8.227757654287349e-05,
      "loss": 0.8142,
      "num_input_tokens_seen": 75622560,
      "step": 130250
    },
    {
      "epoch": 19.400506404527853,
      "grad_norm": 0.046142578125,
      "learning_rate": 8.207377728385478e-05,
      "loss": 0.8066,
      "num_input_tokens_seen": 75625504,
      "step": 130255
    },
    {
      "epoch": 19.40125111706881,
      "grad_norm": 0.07666015625,
      "learning_rate": 8.187023004804005e-05,
      "loss": 0.7961,
      "num_input_tokens_seen": 75628096,
      "step": 130260
    },
    {
      "epoch": 19.40199582960977,
      "grad_norm": 0.049072265625,
      "learning_rate": 8.166693483886489e-05,
      "loss": 0.7933,
      "num_input_tokens_seen": 75630848,
      "step": 130265
    },
    {
      "epoch": 19.40274054215073,
      "grad_norm": 0.048095703125,
      "learning_rate": 8.146389165976487e-05,
      "loss": 0.8066,
      "num_input_tokens_seen": 75633664,
      "step": 130270
    },
    {
      "epoch": 19.403485254691688,
      "grad_norm": 0.03662109375,
      "learning_rate": 8.126110051417056e-05,
      "loss": 0.7948,
      "num_input_tokens_seen": 75636288,
      "step": 130275
    },
    {
      "epoch": 19.40422996723265,
      "grad_norm": 0.0693359375,
      "learning_rate": 8.10585614055076e-05,
      "loss": 0.7985,
      "num_input_tokens_seen": 75639328,
      "step": 130280
    },
    {
      "epoch": 19.40497467977361,
      "grad_norm": 0.06005859375,
      "learning_rate": 8.085627433719489e-05,
      "loss": 0.7826,
      "num_input_tokens_seen": 75642048,
      "step": 130285
    },
    {
      "epoch": 19.405719392314566,
      "grad_norm": 0.045166015625,
      "learning_rate": 8.06542393126547e-05,
      "loss": 0.7999,
      "num_input_tokens_seen": 75644608,
      "step": 130290
    },
    {
      "epoch": 19.406464104855527,
      "grad_norm": 0.029541015625,
      "learning_rate": 8.045245633529596e-05,
      "loss": 0.8025,
      "num_input_tokens_seen": 75647488,
      "step": 130295
    },
    {
      "epoch": 19.407208817396484,
      "grad_norm": 0.07470703125,
      "learning_rate": 8.025092540853096e-05,
      "loss": 0.7861,
      "num_input_tokens_seen": 75650560,
      "step": 130300
    },
    {
      "epoch": 19.407953529937444,
      "grad_norm": 0.054931640625,
      "learning_rate": 8.004964653576196e-05,
      "loss": 0.8037,
      "num_input_tokens_seen": 75653824,
      "step": 130305
    },
    {
      "epoch": 19.408698242478405,
      "grad_norm": 0.040283203125,
      "learning_rate": 7.984861972038959e-05,
      "loss": 0.7967,
      "num_input_tokens_seen": 75656864,
      "step": 130310
    },
    {
      "epoch": 19.40944295501936,
      "grad_norm": 0.111328125,
      "learning_rate": 7.964784496581112e-05,
      "loss": 0.7875,
      "num_input_tokens_seen": 75659616,
      "step": 130315
    },
    {
      "epoch": 19.410187667560322,
      "grad_norm": 0.06494140625,
      "learning_rate": 7.944732227541884e-05,
      "loss": 0.803,
      "num_input_tokens_seen": 75662240,
      "step": 130320
    },
    {
      "epoch": 19.41093238010128,
      "grad_norm": 0.049072265625,
      "learning_rate": 7.924705165259837e-05,
      "loss": 0.7994,
      "num_input_tokens_seen": 75665280,
      "step": 130325
    },
    {
      "epoch": 19.41167709264224,
      "grad_norm": 0.07666015625,
      "learning_rate": 7.904703310073535e-05,
      "loss": 0.7877,
      "num_input_tokens_seen": 75668000,
      "step": 130330
    },
    {
      "epoch": 19.4124218051832,
      "grad_norm": 0.045166015625,
      "learning_rate": 7.884726662320706e-05,
      "loss": 0.7866,
      "num_input_tokens_seen": 75670784,
      "step": 130335
    },
    {
      "epoch": 19.413166517724157,
      "grad_norm": 0.08544921875,
      "learning_rate": 7.864775222338915e-05,
      "loss": 0.8006,
      "num_input_tokens_seen": 75673792,
      "step": 130340
    },
    {
      "epoch": 19.413911230265118,
      "grad_norm": 0.04443359375,
      "learning_rate": 7.844848990465225e-05,
      "loss": 0.7851,
      "num_input_tokens_seen": 75676448,
      "step": 130345
    },
    {
      "epoch": 19.41465594280608,
      "grad_norm": 0.0361328125,
      "learning_rate": 7.8249479670362e-05,
      "loss": 0.7935,
      "num_input_tokens_seen": 75679424,
      "step": 130350
    },
    {
      "epoch": 19.415400655347035,
      "grad_norm": 0.07421875,
      "learning_rate": 7.805072152388071e-05,
      "loss": 0.8068,
      "num_input_tokens_seen": 75682208,
      "step": 130355
    },
    {
      "epoch": 19.416145367887996,
      "grad_norm": 0.03857421875,
      "learning_rate": 7.785221546856734e-05,
      "loss": 0.8127,
      "num_input_tokens_seen": 75685216,
      "step": 130360
    },
    {
      "epoch": 19.416890080428953,
      "grad_norm": 0.03857421875,
      "learning_rate": 7.765396150777426e-05,
      "loss": 0.8256,
      "num_input_tokens_seen": 75688000,
      "step": 130365
    },
    {
      "epoch": 19.417634792969913,
      "grad_norm": 0.0537109375,
      "learning_rate": 7.745595964485041e-05,
      "loss": 0.8012,
      "num_input_tokens_seen": 75690752,
      "step": 130370
    },
    {
      "epoch": 19.418379505510874,
      "grad_norm": 0.05126953125,
      "learning_rate": 7.725820988314313e-05,
      "loss": 0.8008,
      "num_input_tokens_seen": 75693376,
      "step": 130375
    },
    {
      "epoch": 19.41912421805183,
      "grad_norm": 0.054443359375,
      "learning_rate": 7.706071222598976e-05,
      "loss": 0.7954,
      "num_input_tokens_seen": 75696384,
      "step": 130380
    },
    {
      "epoch": 19.41986893059279,
      "grad_norm": 0.0576171875,
      "learning_rate": 7.686346667672927e-05,
      "loss": 0.7771,
      "num_input_tokens_seen": 75699072,
      "step": 130385
    },
    {
      "epoch": 19.420613643133752,
      "grad_norm": 0.0439453125,
      "learning_rate": 7.666647323869401e-05,
      "loss": 0.7791,
      "num_input_tokens_seen": 75701792,
      "step": 130390
    },
    {
      "epoch": 19.42135835567471,
      "grad_norm": 0.05419921875,
      "learning_rate": 7.646973191521133e-05,
      "loss": 0.8005,
      "num_input_tokens_seen": 75704416,
      "step": 130395
    },
    {
      "epoch": 19.42210306821567,
      "grad_norm": 0.0419921875,
      "learning_rate": 7.627324270960356e-05,
      "loss": 0.794,
      "num_input_tokens_seen": 75707200,
      "step": 130400
    },
    {
      "epoch": 19.422847780756626,
      "grad_norm": 0.07763671875,
      "learning_rate": 7.607700562519303e-05,
      "loss": 0.7993,
      "num_input_tokens_seen": 75710208,
      "step": 130405
    },
    {
      "epoch": 19.423592493297587,
      "grad_norm": 0.09521484375,
      "learning_rate": 7.588102066529212e-05,
      "loss": 0.7877,
      "num_input_tokens_seen": 75713024,
      "step": 130410
    },
    {
      "epoch": 19.424337205838548,
      "grad_norm": 0.042724609375,
      "learning_rate": 7.568528783321481e-05,
      "loss": 0.7987,
      "num_input_tokens_seen": 75715776,
      "step": 130415
    },
    {
      "epoch": 19.425081918379504,
      "grad_norm": 0.03955078125,
      "learning_rate": 7.548980713226517e-05,
      "loss": 0.7782,
      "num_input_tokens_seen": 75718720,
      "step": 130420
    },
    {
      "epoch": 19.425826630920465,
      "grad_norm": 0.05078125,
      "learning_rate": 7.529457856574717e-05,
      "loss": 0.7983,
      "num_input_tokens_seen": 75721568,
      "step": 130425
    },
    {
      "epoch": 19.426571343461426,
      "grad_norm": 0.040771484375,
      "learning_rate": 7.509960213695821e-05,
      "loss": 0.7787,
      "num_input_tokens_seen": 75724448,
      "step": 130430
    },
    {
      "epoch": 19.427316056002383,
      "grad_norm": 0.0771484375,
      "learning_rate": 7.49048778491923e-05,
      "loss": 0.7833,
      "num_input_tokens_seen": 75727424,
      "step": 130435
    },
    {
      "epoch": 19.428060768543343,
      "grad_norm": 0.052001953125,
      "learning_rate": 7.47104057057385e-05,
      "loss": 0.7814,
      "num_input_tokens_seen": 75730528,
      "step": 130440
    },
    {
      "epoch": 19.4288054810843,
      "grad_norm": 0.036865234375,
      "learning_rate": 7.451618570988416e-05,
      "loss": 0.8045,
      "num_input_tokens_seen": 75733472,
      "step": 130445
    },
    {
      "epoch": 19.42955019362526,
      "grad_norm": 0.051513671875,
      "learning_rate": 7.432221786491e-05,
      "loss": 0.8056,
      "num_input_tokens_seen": 75736800,
      "step": 130450
    },
    {
      "epoch": 19.43029490616622,
      "grad_norm": 0.056884765625,
      "learning_rate": 7.412850217409006e-05,
      "loss": 0.7885,
      "num_input_tokens_seen": 75740640,
      "step": 130455
    },
    {
      "epoch": 19.431039618707178,
      "grad_norm": 0.0673828125,
      "learning_rate": 7.393503864070006e-05,
      "loss": 0.779,
      "num_input_tokens_seen": 75743648,
      "step": 130460
    },
    {
      "epoch": 19.43178433124814,
      "grad_norm": 0.08642578125,
      "learning_rate": 7.374182726800904e-05,
      "loss": 0.7993,
      "num_input_tokens_seen": 75746368,
      "step": 130465
    },
    {
      "epoch": 19.432529043789096,
      "grad_norm": 0.04833984375,
      "learning_rate": 7.354886805927773e-05,
      "loss": 0.7929,
      "num_input_tokens_seen": 75749152,
      "step": 130470
    },
    {
      "epoch": 19.433273756330056,
      "grad_norm": 0.0419921875,
      "learning_rate": 7.335616101776854e-05,
      "loss": 0.7873,
      "num_input_tokens_seen": 75752352,
      "step": 130475
    },
    {
      "epoch": 19.434018468871017,
      "grad_norm": 0.045654296875,
      "learning_rate": 7.31637061467355e-05,
      "loss": 0.8013,
      "num_input_tokens_seen": 75755328,
      "step": 130480
    },
    {
      "epoch": 19.434763181411974,
      "grad_norm": 0.052978515625,
      "learning_rate": 7.297150344943103e-05,
      "loss": 0.8024,
      "num_input_tokens_seen": 75758368,
      "step": 130485
    },
    {
      "epoch": 19.435507893952934,
      "grad_norm": 0.0244140625,
      "learning_rate": 7.277955292910088e-05,
      "loss": 0.8058,
      "num_input_tokens_seen": 75761280,
      "step": 130490
    },
    {
      "epoch": 19.436252606493895,
      "grad_norm": 0.03759765625,
      "learning_rate": 7.258785458898908e-05,
      "loss": 0.8141,
      "num_input_tokens_seen": 75764192,
      "step": 130495
    },
    {
      "epoch": 19.43699731903485,
      "grad_norm": 0.06591796875,
      "learning_rate": 7.239640843233475e-05,
      "loss": 0.8109,
      "num_input_tokens_seen": 75766944,
      "step": 130500
    },
    {
      "epoch": 19.437742031575812,
      "grad_norm": 0.047119140625,
      "learning_rate": 7.220521446237194e-05,
      "loss": 0.7903,
      "num_input_tokens_seen": 75769888,
      "step": 130505
    },
    {
      "epoch": 19.43848674411677,
      "grad_norm": 0.06591796875,
      "learning_rate": 7.201427268232974e-05,
      "loss": 0.7957,
      "num_input_tokens_seen": 75772640,
      "step": 130510
    },
    {
      "epoch": 19.43923145665773,
      "grad_norm": 0.058349609375,
      "learning_rate": 7.182358309543224e-05,
      "loss": 0.7872,
      "num_input_tokens_seen": 75775456,
      "step": 130515
    },
    {
      "epoch": 19.43997616919869,
      "grad_norm": 0.06689453125,
      "learning_rate": 7.16331457049052e-05,
      "loss": 0.7733,
      "num_input_tokens_seen": 75778112,
      "step": 130520
    },
    {
      "epoch": 19.440720881739647,
      "grad_norm": 0.046875,
      "learning_rate": 7.14429605139627e-05,
      "loss": 0.7928,
      "num_input_tokens_seen": 75781056,
      "step": 130525
    },
    {
      "epoch": 19.441465594280608,
      "grad_norm": 0.0269775390625,
      "learning_rate": 7.125302752581885e-05,
      "loss": 0.8163,
      "num_input_tokens_seen": 75784128,
      "step": 130530
    },
    {
      "epoch": 19.44221030682157,
      "grad_norm": 0.05419921875,
      "learning_rate": 7.106334674368275e-05,
      "loss": 0.8094,
      "num_input_tokens_seen": 75786848,
      "step": 130535
    },
    {
      "epoch": 19.442955019362525,
      "grad_norm": 0.046875,
      "learning_rate": 7.08739181707585e-05,
      "loss": 0.7964,
      "num_input_tokens_seen": 75789984,
      "step": 130540
    },
    {
      "epoch": 19.443699731903486,
      "grad_norm": 0.043212890625,
      "learning_rate": 7.068474181024519e-05,
      "loss": 0.8107,
      "num_input_tokens_seen": 75792736,
      "step": 130545
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 0.044921875,
      "learning_rate": 7.04958176653403e-05,
      "loss": 0.8024,
      "num_input_tokens_seen": 75795936,
      "step": 130550
    },
    {
      "epoch": 19.445189156985403,
      "grad_norm": 0.04541015625,
      "learning_rate": 7.030714573923458e-05,
      "loss": 0.7945,
      "num_input_tokens_seen": 75798848,
      "step": 130555
    },
    {
      "epoch": 19.445933869526364,
      "grad_norm": 0.056884765625,
      "learning_rate": 7.011872603511549e-05,
      "loss": 0.8038,
      "num_input_tokens_seen": 75801664,
      "step": 130560
    },
    {
      "epoch": 19.44667858206732,
      "grad_norm": 0.08642578125,
      "learning_rate": 6.993055855616714e-05,
      "loss": 0.8189,
      "num_input_tokens_seen": 75804352,
      "step": 130565
    },
    {
      "epoch": 19.44742329460828,
      "grad_norm": 0.030029296875,
      "learning_rate": 6.9742643305567e-05,
      "loss": 0.8018,
      "num_input_tokens_seen": 75807328,
      "step": 130570
    },
    {
      "epoch": 19.448168007149242,
      "grad_norm": 0.036376953125,
      "learning_rate": 6.955498028649088e-05,
      "loss": 0.8078,
      "num_input_tokens_seen": 75810368,
      "step": 130575
    },
    {
      "epoch": 19.4489127196902,
      "grad_norm": 0.0556640625,
      "learning_rate": 6.936756950210954e-05,
      "loss": 0.8038,
      "num_input_tokens_seen": 75813120,
      "step": 130580
    },
    {
      "epoch": 19.44965743223116,
      "grad_norm": 0.046630859375,
      "learning_rate": 6.918041095558713e-05,
      "loss": 0.8102,
      "num_input_tokens_seen": 75816096,
      "step": 130585
    },
    {
      "epoch": 19.450402144772116,
      "grad_norm": 0.041015625,
      "learning_rate": 6.899350465008613e-05,
      "loss": 0.8123,
      "num_input_tokens_seen": 75819136,
      "step": 130590
    },
    {
      "epoch": 19.451146857313077,
      "grad_norm": 0.09765625,
      "learning_rate": 6.880685058876567e-05,
      "loss": 0.8016,
      "num_input_tokens_seen": 75821824,
      "step": 130595
    },
    {
      "epoch": 19.451891569854038,
      "grad_norm": 0.058349609375,
      "learning_rate": 6.862044877477823e-05,
      "loss": 0.7976,
      "num_input_tokens_seen": 75824576,
      "step": 130600
    },
    {
      "epoch": 19.452636282394995,
      "grad_norm": 0.11279296875,
      "learning_rate": 6.843429921127298e-05,
      "loss": 0.8095,
      "num_input_tokens_seen": 75827680,
      "step": 130605
    },
    {
      "epoch": 19.453380994935955,
      "grad_norm": 0.06640625,
      "learning_rate": 6.824840190139403e-05,
      "loss": 0.7837,
      "num_input_tokens_seen": 75831040,
      "step": 130610
    },
    {
      "epoch": 19.454125707476916,
      "grad_norm": 0.041748046875,
      "learning_rate": 6.806275684828222e-05,
      "loss": 0.794,
      "num_input_tokens_seen": 75833888,
      "step": 130615
    },
    {
      "epoch": 19.454870420017873,
      "grad_norm": 0.0260009765625,
      "learning_rate": 6.787736405507505e-05,
      "loss": 0.8072,
      "num_input_tokens_seen": 75836864,
      "step": 130620
    },
    {
      "epoch": 19.455615132558833,
      "grad_norm": 0.035400390625,
      "learning_rate": 6.769222352490167e-05,
      "loss": 0.7784,
      "num_input_tokens_seen": 75839712,
      "step": 130625
    },
    {
      "epoch": 19.45635984509979,
      "grad_norm": 0.051513671875,
      "learning_rate": 6.750733526089292e-05,
      "loss": 0.7981,
      "num_input_tokens_seen": 75842752,
      "step": 130630
    },
    {
      "epoch": 19.45710455764075,
      "grad_norm": 0.0380859375,
      "learning_rate": 6.732269926616963e-05,
      "loss": 0.8102,
      "num_input_tokens_seen": 75845952,
      "step": 130635
    },
    {
      "epoch": 19.45784927018171,
      "grad_norm": 0.06982421875,
      "learning_rate": 6.713831554385429e-05,
      "loss": 0.7992,
      "num_input_tokens_seen": 75849088,
      "step": 130640
    },
    {
      "epoch": 19.458593982722668,
      "grad_norm": 0.036865234375,
      "learning_rate": 6.695418409705944e-05,
      "loss": 0.8351,
      "num_input_tokens_seen": 75852096,
      "step": 130645
    },
    {
      "epoch": 19.45933869526363,
      "grad_norm": 0.040283203125,
      "learning_rate": 6.677030492889591e-05,
      "loss": 0.795,
      "num_input_tokens_seen": 75854752,
      "step": 130650
    },
    {
      "epoch": 19.46008340780459,
      "grad_norm": 0.04443359375,
      "learning_rate": 6.65866780424712e-05,
      "loss": 0.8003,
      "num_input_tokens_seen": 75857888,
      "step": 130655
    },
    {
      "epoch": 19.460828120345546,
      "grad_norm": 0.1328125,
      "learning_rate": 6.640330344088618e-05,
      "loss": 0.8403,
      "num_input_tokens_seen": 75860608,
      "step": 130660
    },
    {
      "epoch": 19.461572832886507,
      "grad_norm": 0.039794921875,
      "learning_rate": 6.622018112724004e-05,
      "loss": 0.8192,
      "num_input_tokens_seen": 75863872,
      "step": 130665
    },
    {
      "epoch": 19.462317545427464,
      "grad_norm": 0.03857421875,
      "learning_rate": 6.603731110462529e-05,
      "loss": 0.7937,
      "num_input_tokens_seen": 75866848,
      "step": 130670
    },
    {
      "epoch": 19.463062257968424,
      "grad_norm": 0.032958984375,
      "learning_rate": 6.58546933761328e-05,
      "loss": 0.8072,
      "num_input_tokens_seen": 75869856,
      "step": 130675
    },
    {
      "epoch": 19.463806970509385,
      "grad_norm": 0.03662109375,
      "learning_rate": 6.567232794484678e-05,
      "loss": 0.8098,
      "num_input_tokens_seen": 75872480,
      "step": 130680
    },
    {
      "epoch": 19.464551683050342,
      "grad_norm": 0.04052734375,
      "learning_rate": 6.549021481384809e-05,
      "loss": 0.8056,
      "num_input_tokens_seen": 75875680,
      "step": 130685
    },
    {
      "epoch": 19.465296395591302,
      "grad_norm": 0.03662109375,
      "learning_rate": 6.53083539862126e-05,
      "loss": 0.7664,
      "num_input_tokens_seen": 75878560,
      "step": 130690
    },
    {
      "epoch": 19.46604110813226,
      "grad_norm": 0.056396484375,
      "learning_rate": 6.51267454650145e-05,
      "loss": 0.7905,
      "num_input_tokens_seen": 75881728,
      "step": 130695
    },
    {
      "epoch": 19.46678582067322,
      "grad_norm": 0.03955078125,
      "learning_rate": 6.494538925332138e-05,
      "loss": 0.7843,
      "num_input_tokens_seen": 75884896,
      "step": 130700
    },
    {
      "epoch": 19.46753053321418,
      "grad_norm": 0.058837890625,
      "learning_rate": 6.476428535419576e-05,
      "loss": 0.7977,
      "num_input_tokens_seen": 75887904,
      "step": 130705
    },
    {
      "epoch": 19.468275245755137,
      "grad_norm": 0.0400390625,
      "learning_rate": 6.458343377069687e-05,
      "loss": 0.8199,
      "num_input_tokens_seen": 75890560,
      "step": 130710
    },
    {
      "epoch": 19.469019958296098,
      "grad_norm": 0.03173828125,
      "learning_rate": 6.440283450588391e-05,
      "loss": 0.7867,
      "num_input_tokens_seen": 75893280,
      "step": 130715
    },
    {
      "epoch": 19.46976467083706,
      "grad_norm": 0.0252685546875,
      "learning_rate": 6.42224875628028e-05,
      "loss": 0.8111,
      "num_input_tokens_seen": 75896128,
      "step": 130720
    },
    {
      "epoch": 19.470509383378015,
      "grad_norm": 0.08447265625,
      "learning_rate": 6.404239294450442e-05,
      "loss": 0.7974,
      "num_input_tokens_seen": 75899168,
      "step": 130725
    },
    {
      "epoch": 19.471254095918976,
      "grad_norm": 0.04248046875,
      "learning_rate": 6.386255065402801e-05,
      "loss": 0.8096,
      "num_input_tokens_seen": 75902272,
      "step": 130730
    },
    {
      "epoch": 19.471998808459933,
      "grad_norm": 0.047607421875,
      "learning_rate": 6.368296069441447e-05,
      "loss": 0.7988,
      "num_input_tokens_seen": 75905216,
      "step": 130735
    },
    {
      "epoch": 19.472743521000893,
      "grad_norm": 0.0849609375,
      "learning_rate": 6.350362306869639e-05,
      "loss": 0.7925,
      "num_input_tokens_seen": 75908224,
      "step": 130740
    },
    {
      "epoch": 19.473488233541854,
      "grad_norm": 0.064453125,
      "learning_rate": 6.332453777990465e-05,
      "loss": 0.7883,
      "num_input_tokens_seen": 75910880,
      "step": 130745
    },
    {
      "epoch": 19.47423294608281,
      "grad_norm": 0.035400390625,
      "learning_rate": 6.314570483106352e-05,
      "loss": 0.7838,
      "num_input_tokens_seen": 75913888,
      "step": 130750
    },
    {
      "epoch": 19.47497765862377,
      "grad_norm": 0.041259765625,
      "learning_rate": 6.29671242251939e-05,
      "loss": 0.8097,
      "num_input_tokens_seen": 75917216,
      "step": 130755
    },
    {
      "epoch": 19.475722371164732,
      "grad_norm": 0.055419921875,
      "learning_rate": 6.278879596531339e-05,
      "loss": 0.796,
      "num_input_tokens_seen": 75920128,
      "step": 130760
    },
    {
      "epoch": 19.47646708370569,
      "grad_norm": 0.06298828125,
      "learning_rate": 6.261072005443457e-05,
      "loss": 0.8068,
      "num_input_tokens_seen": 75923296,
      "step": 130765
    },
    {
      "epoch": 19.47721179624665,
      "grad_norm": 0.060546875,
      "learning_rate": 6.243289649556671e-05,
      "loss": 0.7978,
      "num_input_tokens_seen": 75926208,
      "step": 130770
    },
    {
      "epoch": 19.477956508787607,
      "grad_norm": 0.035888671875,
      "learning_rate": 6.22553252917124e-05,
      "loss": 0.8065,
      "num_input_tokens_seen": 75929152,
      "step": 130775
    },
    {
      "epoch": 19.478701221328567,
      "grad_norm": 0.056640625,
      "learning_rate": 6.207800644587257e-05,
      "loss": 0.7891,
      "num_input_tokens_seen": 75931968,
      "step": 130780
    },
    {
      "epoch": 19.479445933869528,
      "grad_norm": 0.039306640625,
      "learning_rate": 6.190093996104318e-05,
      "loss": 0.7926,
      "num_input_tokens_seen": 75935072,
      "step": 130785
    },
    {
      "epoch": 19.480190646410485,
      "grad_norm": 0.050048828125,
      "learning_rate": 6.172412584021514e-05,
      "loss": 0.7696,
      "num_input_tokens_seen": 75937856,
      "step": 130790
    },
    {
      "epoch": 19.480935358951445,
      "grad_norm": 0.05078125,
      "learning_rate": 6.154756408637441e-05,
      "loss": 0.8122,
      "num_input_tokens_seen": 75940864,
      "step": 130795
    },
    {
      "epoch": 19.481680071492406,
      "grad_norm": 0.045166015625,
      "learning_rate": 6.137125470250525e-05,
      "loss": 0.7993,
      "num_input_tokens_seen": 75943488,
      "step": 130800
    },
    {
      "epoch": 19.482424784033363,
      "grad_norm": 0.05517578125,
      "learning_rate": 6.119519769158699e-05,
      "loss": 0.8019,
      "num_input_tokens_seen": 75946272,
      "step": 130805
    },
    {
      "epoch": 19.483169496574323,
      "grad_norm": 0.061767578125,
      "learning_rate": 6.1019393056590544e-05,
      "loss": 0.7971,
      "num_input_tokens_seen": 75949152,
      "step": 130810
    },
    {
      "epoch": 19.48391420911528,
      "grad_norm": 0.0400390625,
      "learning_rate": 6.084384080049021e-05,
      "loss": 0.7886,
      "num_input_tokens_seen": 75952064,
      "step": 130815
    },
    {
      "epoch": 19.48465892165624,
      "grad_norm": 0.043212890625,
      "learning_rate": 6.066854092624863e-05,
      "loss": 0.803,
      "num_input_tokens_seen": 75954944,
      "step": 130820
    },
    {
      "epoch": 19.4854036341972,
      "grad_norm": 0.03564453125,
      "learning_rate": 6.0493493436828417e-05,
      "loss": 0.8,
      "num_input_tokens_seen": 75957536,
      "step": 130825
    },
    {
      "epoch": 19.486148346738158,
      "grad_norm": 0.05419921875,
      "learning_rate": 6.031869833518555e-05,
      "loss": 0.7803,
      "num_input_tokens_seen": 75960640,
      "step": 130830
    },
    {
      "epoch": 19.48689305927912,
      "grad_norm": 0.041748046875,
      "learning_rate": 6.0144155624276e-05,
      "loss": 0.7958,
      "num_input_tokens_seen": 75963680,
      "step": 130835
    },
    {
      "epoch": 19.487637771820076,
      "grad_norm": 0.04736328125,
      "learning_rate": 5.996986530704573e-05,
      "loss": 0.7862,
      "num_input_tokens_seen": 75966496,
      "step": 130840
    },
    {
      "epoch": 19.488382484361036,
      "grad_norm": 0.040771484375,
      "learning_rate": 5.979582738644073e-05,
      "loss": 0.7824,
      "num_input_tokens_seen": 75969184,
      "step": 130845
    },
    {
      "epoch": 19.489127196901997,
      "grad_norm": 0.038330078125,
      "learning_rate": 5.9622041865401987e-05,
      "loss": 0.7917,
      "num_input_tokens_seen": 75972288,
      "step": 130850
    },
    {
      "epoch": 19.489871909442954,
      "grad_norm": 0.03857421875,
      "learning_rate": 5.944850874686213e-05,
      "loss": 0.7973,
      "num_input_tokens_seen": 75974976,
      "step": 130855
    },
    {
      "epoch": 19.490616621983914,
      "grad_norm": 0.038818359375,
      "learning_rate": 5.927522803375551e-05,
      "loss": 0.7929,
      "num_input_tokens_seen": 75977920,
      "step": 130860
    },
    {
      "epoch": 19.491361334524875,
      "grad_norm": 0.057373046875,
      "learning_rate": 5.9102199729009764e-05,
      "loss": 0.7828,
      "num_input_tokens_seen": 75980704,
      "step": 130865
    },
    {
      "epoch": 19.492106047065832,
      "grad_norm": 0.05615234375,
      "learning_rate": 5.892942383554589e-05,
      "loss": 0.8,
      "num_input_tokens_seen": 75983552,
      "step": 130870
    },
    {
      "epoch": 19.492850759606792,
      "grad_norm": 0.053955078125,
      "learning_rate": 5.875690035628489e-05,
      "loss": 0.7986,
      "num_input_tokens_seen": 75986880,
      "step": 130875
    },
    {
      "epoch": 19.49359547214775,
      "grad_norm": 0.03955078125,
      "learning_rate": 5.858462929413943e-05,
      "loss": 0.8182,
      "num_input_tokens_seen": 75989760,
      "step": 130880
    },
    {
      "epoch": 19.49434018468871,
      "grad_norm": 0.0277099609375,
      "learning_rate": 5.841261065202052e-05,
      "loss": 0.8087,
      "num_input_tokens_seen": 75992864,
      "step": 130885
    },
    {
      "epoch": 19.49508489722967,
      "grad_norm": 0.056396484375,
      "learning_rate": 5.8240844432835835e-05,
      "loss": 0.7886,
      "num_input_tokens_seen": 75995744,
      "step": 130890
    },
    {
      "epoch": 19.495829609770627,
      "grad_norm": 0.0294189453125,
      "learning_rate": 5.806933063948638e-05,
      "loss": 0.8134,
      "num_input_tokens_seen": 75998656,
      "step": 130895
    },
    {
      "epoch": 19.496574322311588,
      "grad_norm": 0.0390625,
      "learning_rate": 5.789806927486818e-05,
      "loss": 0.8107,
      "num_input_tokens_seen": 76001696,
      "step": 130900
    },
    {
      "epoch": 19.49731903485255,
      "grad_norm": 0.041748046875,
      "learning_rate": 5.772706034187558e-05,
      "loss": 0.792,
      "num_input_tokens_seen": 76004288,
      "step": 130905
    },
    {
      "epoch": 19.498063747393505,
      "grad_norm": 0.057861328125,
      "learning_rate": 5.7556303843397934e-05,
      "loss": 0.7956,
      "num_input_tokens_seen": 76007104,
      "step": 130910
    },
    {
      "epoch": 19.498808459934466,
      "grad_norm": 0.052490234375,
      "learning_rate": 5.738579978231961e-05,
      "loss": 0.7916,
      "num_input_tokens_seen": 76009760,
      "step": 130915
    },
    {
      "epoch": 19.499553172475423,
      "grad_norm": 0.050537109375,
      "learning_rate": 5.7215548161521634e-05,
      "loss": 0.7873,
      "num_input_tokens_seen": 76012384,
      "step": 130920
    },
    {
      "epoch": 19.500297885016384,
      "grad_norm": 0.047119140625,
      "learning_rate": 5.704554898387837e-05,
      "loss": 0.818,
      "num_input_tokens_seen": 76015392,
      "step": 130925
    },
    {
      "epoch": 19.501042597557344,
      "grad_norm": 0.037353515625,
      "learning_rate": 5.687580225226418e-05,
      "loss": 0.7819,
      "num_input_tokens_seen": 76018240,
      "step": 130930
    },
    {
      "epoch": 19.5017873100983,
      "grad_norm": 0.032470703125,
      "learning_rate": 5.6706307969546785e-05,
      "loss": 0.8048,
      "num_input_tokens_seen": 76021088,
      "step": 130935
    },
    {
      "epoch": 19.50253202263926,
      "grad_norm": 0.0260009765625,
      "learning_rate": 5.6537066138588886e-05,
      "loss": 0.7852,
      "num_input_tokens_seen": 76024000,
      "step": 130940
    },
    {
      "epoch": 19.503276735180222,
      "grad_norm": 0.123046875,
      "learning_rate": 5.6368076762248196e-05,
      "loss": 0.801,
      "num_input_tokens_seen": 76027136,
      "step": 130945
    },
    {
      "epoch": 19.50402144772118,
      "grad_norm": 0.05712890625,
      "learning_rate": 5.619933984338077e-05,
      "loss": 0.7963,
      "num_input_tokens_seen": 76030016,
      "step": 130950
    },
    {
      "epoch": 19.50476616026214,
      "grad_norm": 0.0927734375,
      "learning_rate": 5.603085538483765e-05,
      "loss": 0.7764,
      "num_input_tokens_seen": 76032800,
      "step": 130955
    },
    {
      "epoch": 19.505510872803097,
      "grad_norm": 0.0517578125,
      "learning_rate": 5.5862623389463234e-05,
      "loss": 0.7991,
      "num_input_tokens_seen": 76035616,
      "step": 130960
    },
    {
      "epoch": 19.506255585344057,
      "grad_norm": 0.0478515625,
      "learning_rate": 5.5694643860103585e-05,
      "loss": 0.793,
      "num_input_tokens_seen": 76038432,
      "step": 130965
    },
    {
      "epoch": 19.507000297885018,
      "grad_norm": 0.044189453125,
      "learning_rate": 5.5526916799593094e-05,
      "loss": 0.7859,
      "num_input_tokens_seen": 76041120,
      "step": 130970
    },
    {
      "epoch": 19.507745010425975,
      "grad_norm": 0.045654296875,
      "learning_rate": 5.5359442210766163e-05,
      "loss": 0.7969,
      "num_input_tokens_seen": 76043904,
      "step": 130975
    },
    {
      "epoch": 19.508489722966935,
      "grad_norm": 0.03955078125,
      "learning_rate": 5.5192220096453855e-05,
      "loss": 0.7967,
      "num_input_tokens_seen": 76046880,
      "step": 130980
    },
    {
      "epoch": 19.509234435507892,
      "grad_norm": 0.058837890625,
      "learning_rate": 5.5025250459478924e-05,
      "loss": 0.7764,
      "num_input_tokens_seen": 76049600,
      "step": 130985
    },
    {
      "epoch": 19.509979148048853,
      "grad_norm": 0.0947265625,
      "learning_rate": 5.4858533302662436e-05,
      "loss": 0.7788,
      "num_input_tokens_seen": 76052480,
      "step": 130990
    },
    {
      "epoch": 19.510723860589813,
      "grad_norm": 0.037841796875,
      "learning_rate": 5.469206862882214e-05,
      "loss": 0.7893,
      "num_input_tokens_seen": 76055552,
      "step": 130995
    },
    {
      "epoch": 19.51146857313077,
      "grad_norm": 0.035888671875,
      "learning_rate": 5.4525856440769124e-05,
      "loss": 0.8087,
      "num_input_tokens_seen": 76058464,
      "step": 131000
    },
    {
      "epoch": 19.51221328567173,
      "grad_norm": 0.072265625,
      "learning_rate": 5.4359896741312804e-05,
      "loss": 0.8263,
      "num_input_tokens_seen": 76061408,
      "step": 131005
    },
    {
      "epoch": 19.51295799821269,
      "grad_norm": 0.0595703125,
      "learning_rate": 5.4194189533254275e-05,
      "loss": 0.8044,
      "num_input_tokens_seen": 76064512,
      "step": 131010
    },
    {
      "epoch": 19.51370271075365,
      "grad_norm": 0.05419921875,
      "learning_rate": 5.4028734819396295e-05,
      "loss": 0.7855,
      "num_input_tokens_seen": 76067456,
      "step": 131015
    },
    {
      "epoch": 19.51444742329461,
      "grad_norm": 0.09326171875,
      "learning_rate": 5.3863532602531626e-05,
      "loss": 0.7743,
      "num_input_tokens_seen": 76070144,
      "step": 131020
    },
    {
      "epoch": 19.515192135835566,
      "grad_norm": 0.059814453125,
      "learning_rate": 5.369858288545304e-05,
      "loss": 0.799,
      "num_input_tokens_seen": 76072576,
      "step": 131025
    },
    {
      "epoch": 19.515936848376526,
      "grad_norm": 0.0274658203125,
      "learning_rate": 5.353388567094497e-05,
      "loss": 0.7833,
      "num_input_tokens_seen": 76075552,
      "step": 131030
    },
    {
      "epoch": 19.516681560917487,
      "grad_norm": 0.04345703125,
      "learning_rate": 5.336944096179186e-05,
      "loss": 0.8025,
      "num_input_tokens_seen": 76078080,
      "step": 131035
    },
    {
      "epoch": 19.517426273458444,
      "grad_norm": 0.0478515625,
      "learning_rate": 5.320524876076982e-05,
      "loss": 0.7932,
      "num_input_tokens_seen": 76081152,
      "step": 131040
    },
    {
      "epoch": 19.518170985999404,
      "grad_norm": 0.06005859375,
      "learning_rate": 5.304130907065496e-05,
      "loss": 0.7928,
      "num_input_tokens_seen": 76084192,
      "step": 131045
    },
    {
      "epoch": 19.518915698540365,
      "grad_norm": 0.03271484375,
      "learning_rate": 5.2877621894215075e-05,
      "loss": 0.7981,
      "num_input_tokens_seen": 76087776,
      "step": 131050
    },
    {
      "epoch": 19.519660411081322,
      "grad_norm": 0.048095703125,
      "learning_rate": 5.271418723421628e-05,
      "loss": 0.8027,
      "num_input_tokens_seen": 76090912,
      "step": 131055
    },
    {
      "epoch": 19.520405123622282,
      "grad_norm": 0.04931640625,
      "learning_rate": 5.2551005093419697e-05,
      "loss": 0.7871,
      "num_input_tokens_seen": 76093824,
      "step": 131060
    },
    {
      "epoch": 19.52114983616324,
      "grad_norm": 0.047119140625,
      "learning_rate": 5.238807547458146e-05,
      "loss": 0.8118,
      "num_input_tokens_seen": 76096736,
      "step": 131065
    },
    {
      "epoch": 19.5218945487042,
      "grad_norm": 0.043701171875,
      "learning_rate": 5.2225398380456033e-05,
      "loss": 0.8319,
      "num_input_tokens_seen": 76099552,
      "step": 131070
    },
    {
      "epoch": 19.52263926124516,
      "grad_norm": 0.0478515625,
      "learning_rate": 5.2062973813789546e-05,
      "loss": 0.8269,
      "num_input_tokens_seen": 76102336,
      "step": 131075
    },
    {
      "epoch": 19.523383973786117,
      "grad_norm": 0.053466796875,
      "learning_rate": 5.190080177732648e-05,
      "loss": 0.801,
      "num_input_tokens_seen": 76105024,
      "step": 131080
    },
    {
      "epoch": 19.524128686327078,
      "grad_norm": 0.09814453125,
      "learning_rate": 5.17388822738063e-05,
      "loss": 0.7957,
      "num_input_tokens_seen": 76107840,
      "step": 131085
    },
    {
      "epoch": 19.52487339886804,
      "grad_norm": 0.03662109375,
      "learning_rate": 5.157721530596515e-05,
      "loss": 0.8122,
      "num_input_tokens_seen": 76110944,
      "step": 131090
    },
    {
      "epoch": 19.525618111408996,
      "grad_norm": 0.1005859375,
      "learning_rate": 5.1415800876534187e-05,
      "loss": 0.7871,
      "num_input_tokens_seen": 76113824,
      "step": 131095
    },
    {
      "epoch": 19.526362823949956,
      "grad_norm": 0.031982421875,
      "learning_rate": 5.1254638988241227e-05,
      "loss": 0.7989,
      "num_input_tokens_seen": 76117120,
      "step": 131100
    },
    {
      "epoch": 19.527107536490913,
      "grad_norm": 0.11767578125,
      "learning_rate": 5.109372964380742e-05,
      "loss": 0.7757,
      "num_input_tokens_seen": 76119840,
      "step": 131105
    },
    {
      "epoch": 19.527852249031874,
      "grad_norm": 0.048095703125,
      "learning_rate": 5.093307284595061e-05,
      "loss": 0.7879,
      "num_input_tokens_seen": 76122784,
      "step": 131110
    },
    {
      "epoch": 19.528596961572834,
      "grad_norm": 0.036865234375,
      "learning_rate": 5.0772668597386935e-05,
      "loss": 0.7931,
      "num_input_tokens_seen": 76125408,
      "step": 131115
    },
    {
      "epoch": 19.52934167411379,
      "grad_norm": 0.048583984375,
      "learning_rate": 5.061251690082424e-05,
      "loss": 0.7885,
      "num_input_tokens_seen": 76128384,
      "step": 131120
    },
    {
      "epoch": 19.53008638665475,
      "grad_norm": 0.0284423828125,
      "learning_rate": 5.045261775897036e-05,
      "loss": 0.7954,
      "num_input_tokens_seen": 76131200,
      "step": 131125
    },
    {
      "epoch": 19.530831099195712,
      "grad_norm": 0.029052734375,
      "learning_rate": 5.02929711745248e-05,
      "loss": 0.8041,
      "num_input_tokens_seen": 76134112,
      "step": 131130
    },
    {
      "epoch": 19.53157581173667,
      "grad_norm": 0.03515625,
      "learning_rate": 5.013357715018707e-05,
      "loss": 0.7929,
      "num_input_tokens_seen": 76137248,
      "step": 131135
    },
    {
      "epoch": 19.53232052427763,
      "grad_norm": 0.039794921875,
      "learning_rate": 4.997443568864668e-05,
      "loss": 0.8057,
      "num_input_tokens_seen": 76140224,
      "step": 131140
    },
    {
      "epoch": 19.533065236818587,
      "grad_norm": 0.037841796875,
      "learning_rate": 4.981554679259315e-05,
      "loss": 0.791,
      "num_input_tokens_seen": 76143264,
      "step": 131145
    },
    {
      "epoch": 19.533809949359547,
      "grad_norm": 0.035888671875,
      "learning_rate": 4.9656910464712656e-05,
      "loss": 0.7956,
      "num_input_tokens_seen": 76146368,
      "step": 131150
    },
    {
      "epoch": 19.534554661900508,
      "grad_norm": 0.05517578125,
      "learning_rate": 4.949852670768306e-05,
      "loss": 0.7954,
      "num_input_tokens_seen": 76149024,
      "step": 131155
    },
    {
      "epoch": 19.535299374441465,
      "grad_norm": 0.04833984375,
      "learning_rate": 4.9340395524180565e-05,
      "loss": 0.8006,
      "num_input_tokens_seen": 76151904,
      "step": 131160
    },
    {
      "epoch": 19.536044086982425,
      "grad_norm": 0.04052734375,
      "learning_rate": 4.918251691687636e-05,
      "loss": 0.7687,
      "num_input_tokens_seen": 76154592,
      "step": 131165
    },
    {
      "epoch": 19.536788799523386,
      "grad_norm": 0.044677734375,
      "learning_rate": 4.9024890888439975e-05,
      "loss": 0.8052,
      "num_input_tokens_seen": 76157440,
      "step": 131170
    },
    {
      "epoch": 19.537533512064343,
      "grad_norm": 0.06396484375,
      "learning_rate": 4.8867517441530946e-05,
      "loss": 0.7823,
      "num_input_tokens_seen": 76160192,
      "step": 131175
    },
    {
      "epoch": 19.538278224605303,
      "grad_norm": 0.039794921875,
      "learning_rate": 4.871039657881049e-05,
      "loss": 0.798,
      "num_input_tokens_seen": 76163264,
      "step": 131180
    },
    {
      "epoch": 19.53902293714626,
      "grad_norm": 0.06396484375,
      "learning_rate": 4.8553528302931466e-05,
      "loss": 0.7939,
      "num_input_tokens_seen": 76166240,
      "step": 131185
    },
    {
      "epoch": 19.53976764968722,
      "grad_norm": 0.032958984375,
      "learning_rate": 4.839691261654511e-05,
      "loss": 0.7969,
      "num_input_tokens_seen": 76168992,
      "step": 131190
    },
    {
      "epoch": 19.54051236222818,
      "grad_norm": 0.04345703125,
      "learning_rate": 4.8240549522295946e-05,
      "loss": 0.7869,
      "num_input_tokens_seen": 76172000,
      "step": 131195
    },
    {
      "epoch": 19.54125707476914,
      "grad_norm": 0.03271484375,
      "learning_rate": 4.8084439022826884e-05,
      "loss": 0.8034,
      "num_input_tokens_seen": 76174880,
      "step": 131200
    },
    {
      "epoch": 19.5420017873101,
      "grad_norm": 0.072265625,
      "learning_rate": 4.79285811207758e-05,
      "loss": 0.8132,
      "num_input_tokens_seen": 76177856,
      "step": 131205
    },
    {
      "epoch": 19.542746499851056,
      "grad_norm": 0.0517578125,
      "learning_rate": 4.777297581877393e-05,
      "loss": 0.7972,
      "num_input_tokens_seen": 76180896,
      "step": 131210
    },
    {
      "epoch": 19.543491212392016,
      "grad_norm": 0.039794921875,
      "learning_rate": 4.7617623119450835e-05,
      "loss": 0.7834,
      "num_input_tokens_seen": 76183872,
      "step": 131215
    },
    {
      "epoch": 19.544235924932977,
      "grad_norm": 0.1201171875,
      "learning_rate": 4.746252302543108e-05,
      "loss": 0.7926,
      "num_input_tokens_seen": 76186656,
      "step": 131220
    },
    {
      "epoch": 19.544980637473934,
      "grad_norm": 0.061767578125,
      "learning_rate": 4.7307675539334236e-05,
      "loss": 0.804,
      "num_input_tokens_seen": 76189568,
      "step": 131225
    },
    {
      "epoch": 19.545725350014894,
      "grad_norm": 0.05322265625,
      "learning_rate": 4.715308066377821e-05,
      "loss": 0.786,
      "num_input_tokens_seen": 76192320,
      "step": 131230
    },
    {
      "epoch": 19.546470062555855,
      "grad_norm": 0.044677734375,
      "learning_rate": 4.6998738401372585e-05,
      "loss": 0.7877,
      "num_input_tokens_seen": 76194880,
      "step": 131235
    },
    {
      "epoch": 19.547214775096812,
      "grad_norm": 0.03515625,
      "learning_rate": 4.6844648754726936e-05,
      "loss": 0.7968,
      "num_input_tokens_seen": 76197696,
      "step": 131240
    },
    {
      "epoch": 19.547959487637772,
      "grad_norm": 0.05029296875,
      "learning_rate": 4.6690811726440847e-05,
      "loss": 0.8083,
      "num_input_tokens_seen": 76200416,
      "step": 131245
    },
    {
      "epoch": 19.54870420017873,
      "grad_norm": 0.044677734375,
      "learning_rate": 4.653722731911724e-05,
      "loss": 0.7999,
      "num_input_tokens_seen": 76203200,
      "step": 131250
    },
    {
      "epoch": 19.54944891271969,
      "grad_norm": 0.04345703125,
      "learning_rate": 4.638389553534905e-05,
      "loss": 0.7861,
      "num_input_tokens_seen": 76205984,
      "step": 131255
    },
    {
      "epoch": 19.55019362526065,
      "grad_norm": 0.1337890625,
      "learning_rate": 4.623081637772752e-05,
      "loss": 0.8009,
      "num_input_tokens_seen": 76208992,
      "step": 131260
    },
    {
      "epoch": 19.550938337801608,
      "grad_norm": 0.059326171875,
      "learning_rate": 4.607798984883726e-05,
      "loss": 0.8239,
      "num_input_tokens_seen": 76211776,
      "step": 131265
    },
    {
      "epoch": 19.551683050342568,
      "grad_norm": 0.078125,
      "learning_rate": 4.5925415951261205e-05,
      "loss": 0.7906,
      "num_input_tokens_seen": 76214688,
      "step": 131270
    },
    {
      "epoch": 19.55242776288353,
      "grad_norm": 0.04443359375,
      "learning_rate": 4.577309468757729e-05,
      "loss": 0.798,
      "num_input_tokens_seen": 76217632,
      "step": 131275
    },
    {
      "epoch": 19.553172475424486,
      "grad_norm": 0.0576171875,
      "learning_rate": 4.562102606035678e-05,
      "loss": 0.7962,
      "num_input_tokens_seen": 76220320,
      "step": 131280
    },
    {
      "epoch": 19.553917187965446,
      "grad_norm": 0.0269775390625,
      "learning_rate": 4.54692100721693e-05,
      "loss": 0.8098,
      "num_input_tokens_seen": 76223456,
      "step": 131285
    },
    {
      "epoch": 19.554661900506403,
      "grad_norm": 0.0380859375,
      "learning_rate": 4.5317646725581135e-05,
      "loss": 0.7889,
      "num_input_tokens_seen": 76225984,
      "step": 131290
    },
    {
      "epoch": 19.555406613047364,
      "grad_norm": 0.03955078125,
      "learning_rate": 4.5166336023153564e-05,
      "loss": 0.8008,
      "num_input_tokens_seen": 76228896,
      "step": 131295
    },
    {
      "epoch": 19.556151325588324,
      "grad_norm": 0.0771484375,
      "learning_rate": 4.5015277967439534e-05,
      "loss": 0.7854,
      "num_input_tokens_seen": 76231488,
      "step": 131300
    },
    {
      "epoch": 19.55689603812928,
      "grad_norm": 0.035888671875,
      "learning_rate": 4.486447256099368e-05,
      "loss": 0.7928,
      "num_input_tokens_seen": 76234560,
      "step": 131305
    },
    {
      "epoch": 19.55764075067024,
      "grad_norm": 0.07958984375,
      "learning_rate": 4.4713919806363966e-05,
      "loss": 0.7985,
      "num_input_tokens_seen": 76237760,
      "step": 131310
    },
    {
      "epoch": 19.558385463211202,
      "grad_norm": 0.059326171875,
      "learning_rate": 4.456361970609002e-05,
      "loss": 0.7917,
      "num_input_tokens_seen": 76240672,
      "step": 131315
    },
    {
      "epoch": 19.55913017575216,
      "grad_norm": 0.03955078125,
      "learning_rate": 4.4413572262714805e-05,
      "loss": 0.7884,
      "num_input_tokens_seen": 76243616,
      "step": 131320
    },
    {
      "epoch": 19.55987488829312,
      "grad_norm": 0.19921875,
      "learning_rate": 4.426377747877297e-05,
      "loss": 0.7969,
      "num_input_tokens_seen": 76246720,
      "step": 131325
    },
    {
      "epoch": 19.560619600834077,
      "grad_norm": 0.04248046875,
      "learning_rate": 4.411423535679415e-05,
      "loss": 0.7839,
      "num_input_tokens_seen": 76249664,
      "step": 131330
    },
    {
      "epoch": 19.561364313375037,
      "grad_norm": 0.04052734375,
      "learning_rate": 4.3964945899304665e-05,
      "loss": 0.7891,
      "num_input_tokens_seen": 76252768,
      "step": 131335
    },
    {
      "epoch": 19.562109025915998,
      "grad_norm": 0.043212890625,
      "learning_rate": 4.3815909108825824e-05,
      "loss": 0.7818,
      "num_input_tokens_seen": 76255552,
      "step": 131340
    },
    {
      "epoch": 19.562853738456955,
      "grad_norm": 0.051025390625,
      "learning_rate": 4.366712498787728e-05,
      "loss": 0.7851,
      "num_input_tokens_seen": 76258560,
      "step": 131345
    },
    {
      "epoch": 19.563598450997915,
      "grad_norm": 0.039794921875,
      "learning_rate": 4.3518593538972026e-05,
      "loss": 0.7849,
      "num_input_tokens_seen": 76261408,
      "step": 131350
    },
    {
      "epoch": 19.564343163538872,
      "grad_norm": 0.048095703125,
      "learning_rate": 4.3370314764618056e-05,
      "loss": 0.8107,
      "num_input_tokens_seen": 76264576,
      "step": 131355
    },
    {
      "epoch": 19.565087876079833,
      "grad_norm": 0.047119140625,
      "learning_rate": 4.3222288667321695e-05,
      "loss": 0.7983,
      "num_input_tokens_seen": 76267616,
      "step": 131360
    },
    {
      "epoch": 19.565832588620793,
      "grad_norm": 0.05419921875,
      "learning_rate": 4.307451524958428e-05,
      "loss": 0.8085,
      "num_input_tokens_seen": 76270368,
      "step": 131365
    },
    {
      "epoch": 19.56657730116175,
      "grad_norm": 0.041259765625,
      "learning_rate": 4.2926994513900474e-05,
      "loss": 0.8108,
      "num_input_tokens_seen": 76273216,
      "step": 131370
    },
    {
      "epoch": 19.56732201370271,
      "grad_norm": 0.0498046875,
      "learning_rate": 4.277972646276329e-05,
      "loss": 0.7927,
      "num_input_tokens_seen": 76276064,
      "step": 131375
    },
    {
      "epoch": 19.56806672624367,
      "grad_norm": 0.039306640625,
      "learning_rate": 4.26327110986624e-05,
      "loss": 0.7903,
      "num_input_tokens_seen": 76278688,
      "step": 131380
    },
    {
      "epoch": 19.56881143878463,
      "grad_norm": 0.0478515625,
      "learning_rate": 4.248594842407749e-05,
      "loss": 0.7922,
      "num_input_tokens_seen": 76281344,
      "step": 131385
    },
    {
      "epoch": 19.56955615132559,
      "grad_norm": 0.041015625,
      "learning_rate": 4.233943844149157e-05,
      "loss": 0.8094,
      "num_input_tokens_seen": 76284544,
      "step": 131390
    },
    {
      "epoch": 19.570300863866546,
      "grad_norm": 0.03369140625,
      "learning_rate": 4.2193181153377666e-05,
      "loss": 0.8092,
      "num_input_tokens_seen": 76287360,
      "step": 131395
    },
    {
      "epoch": 19.571045576407506,
      "grad_norm": 0.038330078125,
      "learning_rate": 4.2047176562208795e-05,
      "loss": 0.7939,
      "num_input_tokens_seen": 76290144,
      "step": 131400
    },
    {
      "epoch": 19.571790288948467,
      "grad_norm": 0.06396484375,
      "learning_rate": 4.190142467044966e-05,
      "loss": 0.7869,
      "num_input_tokens_seen": 76293056,
      "step": 131405
    },
    {
      "epoch": 19.572535001489424,
      "grad_norm": 0.0289306640625,
      "learning_rate": 4.175592548056495e-05,
      "loss": 0.8099,
      "num_input_tokens_seen": 76295936,
      "step": 131410
    },
    {
      "epoch": 19.573279714030384,
      "grad_norm": 0.031982421875,
      "learning_rate": 4.16106789950077e-05,
      "loss": 0.7868,
      "num_input_tokens_seen": 76298592,
      "step": 131415
    },
    {
      "epoch": 19.574024426571345,
      "grad_norm": 0.03955078125,
      "learning_rate": 4.146568521623761e-05,
      "loss": 0.7867,
      "num_input_tokens_seen": 76301408,
      "step": 131420
    },
    {
      "epoch": 19.574769139112302,
      "grad_norm": 0.130859375,
      "learning_rate": 4.1320944146701065e-05,
      "loss": 0.8055,
      "num_input_tokens_seen": 76304224,
      "step": 131425
    },
    {
      "epoch": 19.575513851653263,
      "grad_norm": 0.039794921875,
      "learning_rate": 4.117645578884277e-05,
      "loss": 0.7956,
      "num_input_tokens_seen": 76307072,
      "step": 131430
    },
    {
      "epoch": 19.57625856419422,
      "grad_norm": 0.030517578125,
      "learning_rate": 4.103222014510577e-05,
      "loss": 0.7854,
      "num_input_tokens_seen": 76310240,
      "step": 131435
    },
    {
      "epoch": 19.57700327673518,
      "grad_norm": 0.031982421875,
      "learning_rate": 4.088823721792478e-05,
      "loss": 0.8,
      "num_input_tokens_seen": 76312992,
      "step": 131440
    },
    {
      "epoch": 19.57774798927614,
      "grad_norm": 0.0294189453125,
      "learning_rate": 4.074450700973453e-05,
      "loss": 0.8077,
      "num_input_tokens_seen": 76316032,
      "step": 131445
    },
    {
      "epoch": 19.578492701817098,
      "grad_norm": 0.03466796875,
      "learning_rate": 4.060102952296141e-05,
      "loss": 0.7957,
      "num_input_tokens_seen": 76318976,
      "step": 131450
    },
    {
      "epoch": 19.579237414358058,
      "grad_norm": 0.04931640625,
      "learning_rate": 4.0457804760030136e-05,
      "loss": 0.8013,
      "num_input_tokens_seen": 76321728,
      "step": 131455
    },
    {
      "epoch": 19.57998212689902,
      "grad_norm": 0.03662109375,
      "learning_rate": 4.031483272335878e-05,
      "loss": 0.7964,
      "num_input_tokens_seen": 76324512,
      "step": 131460
    },
    {
      "epoch": 19.580726839439976,
      "grad_norm": 0.051025390625,
      "learning_rate": 4.017211341536542e-05,
      "loss": 0.7896,
      "num_input_tokens_seen": 76327648,
      "step": 131465
    },
    {
      "epoch": 19.581471551980936,
      "grad_norm": 0.037353515625,
      "learning_rate": 4.0029646838459774e-05,
      "loss": 0.7839,
      "num_input_tokens_seen": 76330336,
      "step": 131470
    },
    {
      "epoch": 19.582216264521893,
      "grad_norm": 0.06494140625,
      "learning_rate": 3.988743299504826e-05,
      "loss": 0.7937,
      "num_input_tokens_seen": 76333088,
      "step": 131475
    },
    {
      "epoch": 19.582960977062854,
      "grad_norm": 0.042236328125,
      "learning_rate": 3.9745471887533965e-05,
      "loss": 0.786,
      "num_input_tokens_seen": 76336288,
      "step": 131480
    },
    {
      "epoch": 19.583705689603814,
      "grad_norm": 0.06005859375,
      "learning_rate": 3.960376351831329e-05,
      "loss": 0.7866,
      "num_input_tokens_seen": 76339008,
      "step": 131485
    },
    {
      "epoch": 19.58445040214477,
      "grad_norm": 0.049072265625,
      "learning_rate": 3.946230788978433e-05,
      "loss": 0.7968,
      "num_input_tokens_seen": 76342080,
      "step": 131490
    },
    {
      "epoch": 19.58519511468573,
      "grad_norm": 0.34375,
      "learning_rate": 3.93211050043335e-05,
      "loss": 0.7966,
      "num_input_tokens_seen": 76344800,
      "step": 131495
    },
    {
      "epoch": 19.58593982722669,
      "grad_norm": 0.03466796875,
      "learning_rate": 3.9180154864347225e-05,
      "loss": 0.7986,
      "num_input_tokens_seen": 76347680,
      "step": 131500
    },
    {
      "epoch": 19.58668453976765,
      "grad_norm": 0.03759765625,
      "learning_rate": 3.9039457472206936e-05,
      "loss": 0.8108,
      "num_input_tokens_seen": 76350624,
      "step": 131505
    },
    {
      "epoch": 19.58742925230861,
      "grad_norm": 0.03564453125,
      "learning_rate": 3.8899012830287405e-05,
      "loss": 0.7849,
      "num_input_tokens_seen": 76353664,
      "step": 131510
    },
    {
      "epoch": 19.588173964849567,
      "grad_norm": 0.0380859375,
      "learning_rate": 3.875882094096505e-05,
      "loss": 0.7906,
      "num_input_tokens_seen": 76356608,
      "step": 131515
    },
    {
      "epoch": 19.588918677390527,
      "grad_norm": 0.041748046875,
      "learning_rate": 3.8618881806606326e-05,
      "loss": 0.7754,
      "num_input_tokens_seen": 76359328,
      "step": 131520
    },
    {
      "epoch": 19.589663389931488,
      "grad_norm": 0.0458984375,
      "learning_rate": 3.847919542957601e-05,
      "loss": 0.8184,
      "num_input_tokens_seen": 76362144,
      "step": 131525
    },
    {
      "epoch": 19.590408102472445,
      "grad_norm": 0.0439453125,
      "learning_rate": 3.833976181223386e-05,
      "loss": 0.7985,
      "num_input_tokens_seen": 76365408,
      "step": 131530
    },
    {
      "epoch": 19.591152815013405,
      "grad_norm": 0.0458984375,
      "learning_rate": 3.820058095693468e-05,
      "loss": 0.788,
      "num_input_tokens_seen": 76368128,
      "step": 131535
    },
    {
      "epoch": 19.591897527554362,
      "grad_norm": 0.03564453125,
      "learning_rate": 3.8061652866029915e-05,
      "loss": 0.8037,
      "num_input_tokens_seen": 76371200,
      "step": 131540
    },
    {
      "epoch": 19.592642240095323,
      "grad_norm": 0.06396484375,
      "learning_rate": 3.792297754186602e-05,
      "loss": 0.7891,
      "num_input_tokens_seen": 76373856,
      "step": 131545
    },
    {
      "epoch": 19.593386952636283,
      "grad_norm": 0.0654296875,
      "learning_rate": 3.7784554986787785e-05,
      "loss": 0.8232,
      "num_input_tokens_seen": 76376736,
      "step": 131550
    },
    {
      "epoch": 19.59413166517724,
      "grad_norm": 0.05419921875,
      "learning_rate": 3.764638520313168e-05,
      "loss": 0.7996,
      "num_input_tokens_seen": 76379744,
      "step": 131555
    },
    {
      "epoch": 19.5948763777182,
      "grad_norm": 0.040283203125,
      "learning_rate": 3.750846819323417e-05,
      "loss": 0.791,
      "num_input_tokens_seen": 76382336,
      "step": 131560
    },
    {
      "epoch": 19.59562109025916,
      "grad_norm": 0.0252685546875,
      "learning_rate": 3.737080395942172e-05,
      "loss": 0.7846,
      "num_input_tokens_seen": 76385216,
      "step": 131565
    },
    {
      "epoch": 19.59636580280012,
      "grad_norm": 0.050537109375,
      "learning_rate": 3.723339250402413e-05,
      "loss": 0.7926,
      "num_input_tokens_seen": 76387968,
      "step": 131570
    },
    {
      "epoch": 19.59711051534108,
      "grad_norm": 0.04345703125,
      "learning_rate": 3.709623382936122e-05,
      "loss": 0.7984,
      "num_input_tokens_seen": 76390816,
      "step": 131575
    },
    {
      "epoch": 19.597855227882036,
      "grad_norm": 0.035400390625,
      "learning_rate": 3.695932793774781e-05,
      "loss": 0.7809,
      "num_input_tokens_seen": 76393696,
      "step": 131580
    },
    {
      "epoch": 19.598599940422996,
      "grad_norm": 0.07861328125,
      "learning_rate": 3.682267483150037e-05,
      "loss": 0.7935,
      "num_input_tokens_seen": 76396544,
      "step": 131585
    },
    {
      "epoch": 19.599344652963957,
      "grad_norm": 0.03076171875,
      "learning_rate": 3.6686274512925386e-05,
      "loss": 0.8054,
      "num_input_tokens_seen": 76399456,
      "step": 131590
    },
    {
      "epoch": 19.600089365504914,
      "grad_norm": 0.0242919921875,
      "learning_rate": 3.655012698432769e-05,
      "loss": 0.8127,
      "num_input_tokens_seen": 76402176,
      "step": 131595
    },
    {
      "epoch": 19.600834078045875,
      "grad_norm": 0.050537109375,
      "learning_rate": 3.641423224800877e-05,
      "loss": 0.7859,
      "num_input_tokens_seen": 76404992,
      "step": 131600
    },
    {
      "epoch": 19.601578790586835,
      "grad_norm": 0.03759765625,
      "learning_rate": 3.627859030626179e-05,
      "loss": 0.804,
      "num_input_tokens_seen": 76407936,
      "step": 131605
    },
    {
      "epoch": 19.602323503127792,
      "grad_norm": 0.06787109375,
      "learning_rate": 3.614320116137992e-05,
      "loss": 0.7947,
      "num_input_tokens_seen": 76410528,
      "step": 131610
    },
    {
      "epoch": 19.603068215668753,
      "grad_norm": 0.039306640625,
      "learning_rate": 3.6008064815649665e-05,
      "loss": 0.8407,
      "num_input_tokens_seen": 76413056,
      "step": 131615
    },
    {
      "epoch": 19.60381292820971,
      "grad_norm": 0.0277099609375,
      "learning_rate": 3.587318127135419e-05,
      "loss": 0.8119,
      "num_input_tokens_seen": 76415776,
      "step": 131620
    },
    {
      "epoch": 19.60455764075067,
      "grad_norm": 0.035400390625,
      "learning_rate": 3.573855053077335e-05,
      "loss": 0.8131,
      "num_input_tokens_seen": 76418784,
      "step": 131625
    },
    {
      "epoch": 19.60530235329163,
      "grad_norm": 0.059814453125,
      "learning_rate": 3.5604172596180315e-05,
      "loss": 0.7986,
      "num_input_tokens_seen": 76421888,
      "step": 131630
    },
    {
      "epoch": 19.606047065832588,
      "grad_norm": 0.051025390625,
      "learning_rate": 3.547004746984495e-05,
      "loss": 0.7951,
      "num_input_tokens_seen": 76424768,
      "step": 131635
    },
    {
      "epoch": 19.606791778373548,
      "grad_norm": 0.04541015625,
      "learning_rate": 3.533617515403542e-05,
      "loss": 0.7853,
      "num_input_tokens_seen": 76427744,
      "step": 131640
    },
    {
      "epoch": 19.60753649091451,
      "grad_norm": 0.054443359375,
      "learning_rate": 3.520255565100994e-05,
      "loss": 0.7903,
      "num_input_tokens_seen": 76430496,
      "step": 131645
    },
    {
      "epoch": 19.608281203455466,
      "grad_norm": 0.07080078125,
      "learning_rate": 3.506918896302835e-05,
      "loss": 0.8029,
      "num_input_tokens_seen": 76433440,
      "step": 131650
    },
    {
      "epoch": 19.609025915996426,
      "grad_norm": 0.080078125,
      "learning_rate": 3.493607509234386e-05,
      "loss": 0.7863,
      "num_input_tokens_seen": 76436576,
      "step": 131655
    },
    {
      "epoch": 19.609770628537383,
      "grad_norm": 0.0615234375,
      "learning_rate": 3.4803214041204674e-05,
      "loss": 0.7927,
      "num_input_tokens_seen": 76439488,
      "step": 131660
    },
    {
      "epoch": 19.610515341078344,
      "grad_norm": 0.054443359375,
      "learning_rate": 3.4670605811855655e-05,
      "loss": 0.7802,
      "num_input_tokens_seen": 76442464,
      "step": 131665
    },
    {
      "epoch": 19.611260053619304,
      "grad_norm": 0.03759765625,
      "learning_rate": 3.453825040653502e-05,
      "loss": 0.8119,
      "num_input_tokens_seen": 76445664,
      "step": 131670
    },
    {
      "epoch": 19.61200476616026,
      "grad_norm": 0.037353515625,
      "learning_rate": 3.440614782748097e-05,
      "loss": 0.8031,
      "num_input_tokens_seen": 76448352,
      "step": 131675
    },
    {
      "epoch": 19.61274947870122,
      "grad_norm": 0.043701171875,
      "learning_rate": 3.427429807692506e-05,
      "loss": 0.7914,
      "num_input_tokens_seen": 76451200,
      "step": 131680
    },
    {
      "epoch": 19.613494191242182,
      "grad_norm": 0.04296875,
      "learning_rate": 3.414270115709383e-05,
      "loss": 0.7966,
      "num_input_tokens_seen": 76454240,
      "step": 131685
    },
    {
      "epoch": 19.61423890378314,
      "grad_norm": 0.040283203125,
      "learning_rate": 3.401135707021219e-05,
      "loss": 0.8088,
      "num_input_tokens_seen": 76456800,
      "step": 131690
    },
    {
      "epoch": 19.6149836163241,
      "grad_norm": 0.032470703125,
      "learning_rate": 3.388026581849668e-05,
      "loss": 0.7885,
      "num_input_tokens_seen": 76459808,
      "step": 131695
    },
    {
      "epoch": 19.615728328865057,
      "grad_norm": 0.057373046875,
      "learning_rate": 3.3749427404163864e-05,
      "loss": 0.7992,
      "num_input_tokens_seen": 76463168,
      "step": 131700
    },
    {
      "epoch": 19.616473041406017,
      "grad_norm": 0.056884765625,
      "learning_rate": 3.361884182942365e-05,
      "loss": 0.8231,
      "num_input_tokens_seen": 76466144,
      "step": 131705
    },
    {
      "epoch": 19.617217753946978,
      "grad_norm": 0.0771484375,
      "learning_rate": 3.348850909648093e-05,
      "loss": 0.7858,
      "num_input_tokens_seen": 76469280,
      "step": 131710
    },
    {
      "epoch": 19.617962466487935,
      "grad_norm": 0.0537109375,
      "learning_rate": 3.335842920754062e-05,
      "loss": 0.7995,
      "num_input_tokens_seen": 76471968,
      "step": 131715
    },
    {
      "epoch": 19.618707179028895,
      "grad_norm": 0.06591796875,
      "learning_rate": 3.3228602164795946e-05,
      "loss": 0.7811,
      "num_input_tokens_seen": 76474944,
      "step": 131720
    },
    {
      "epoch": 19.619451891569852,
      "grad_norm": 0.05224609375,
      "learning_rate": 3.30990279704435e-05,
      "loss": 0.8109,
      "num_input_tokens_seen": 76477888,
      "step": 131725
    },
    {
      "epoch": 19.620196604110813,
      "grad_norm": 0.036865234375,
      "learning_rate": 3.296970662667153e-05,
      "loss": 0.8139,
      "num_input_tokens_seen": 76480672,
      "step": 131730
    },
    {
      "epoch": 19.620941316651773,
      "grad_norm": 0.05224609375,
      "learning_rate": 3.284063813566495e-05,
      "loss": 0.7852,
      "num_input_tokens_seen": 76483552,
      "step": 131735
    },
    {
      "epoch": 19.62168602919273,
      "grad_norm": 0.05126953125,
      "learning_rate": 3.271182249960203e-05,
      "loss": 0.8014,
      "num_input_tokens_seen": 76486432,
      "step": 131740
    },
    {
      "epoch": 19.62243074173369,
      "grad_norm": 0.053955078125,
      "learning_rate": 3.258325972066267e-05,
      "loss": 0.8098,
      "num_input_tokens_seen": 76489504,
      "step": 131745
    },
    {
      "epoch": 19.62317545427465,
      "grad_norm": 0.044921875,
      "learning_rate": 3.245494980101515e-05,
      "loss": 0.7826,
      "num_input_tokens_seen": 76492128,
      "step": 131750
    },
    {
      "epoch": 19.62392016681561,
      "grad_norm": 0.055419921875,
      "learning_rate": 3.232689274283107e-05,
      "loss": 0.8096,
      "num_input_tokens_seen": 76494816,
      "step": 131755
    },
    {
      "epoch": 19.62466487935657,
      "grad_norm": 0.052490234375,
      "learning_rate": 3.219908854827036e-05,
      "loss": 0.8123,
      "num_input_tokens_seen": 76497888,
      "step": 131760
    },
    {
      "epoch": 19.625409591897526,
      "grad_norm": 0.049560546875,
      "learning_rate": 3.207153721949296e-05,
      "loss": 0.7923,
      "num_input_tokens_seen": 76500960,
      "step": 131765
    },
    {
      "epoch": 19.626154304438487,
      "grad_norm": 0.09326171875,
      "learning_rate": 3.194423875865548e-05,
      "loss": 0.7903,
      "num_input_tokens_seen": 76503808,
      "step": 131770
    },
    {
      "epoch": 19.626899016979447,
      "grad_norm": 0.057373046875,
      "learning_rate": 3.181719316790621e-05,
      "loss": 0.792,
      "num_input_tokens_seen": 76507008,
      "step": 131775
    },
    {
      "epoch": 19.627643729520404,
      "grad_norm": 0.04296875,
      "learning_rate": 3.1690400449393416e-05,
      "loss": 0.8063,
      "num_input_tokens_seen": 76510112,
      "step": 131780
    },
    {
      "epoch": 19.628388442061365,
      "grad_norm": 0.07568359375,
      "learning_rate": 3.1563860605257064e-05,
      "loss": 0.8045,
      "num_input_tokens_seen": 76513312,
      "step": 131785
    },
    {
      "epoch": 19.629133154602325,
      "grad_norm": 0.05322265625,
      "learning_rate": 3.1437573637637104e-05,
      "loss": 0.8,
      "num_input_tokens_seen": 76516096,
      "step": 131790
    },
    {
      "epoch": 19.629877867143282,
      "grad_norm": 0.0712890625,
      "learning_rate": 3.1311539548665165e-05,
      "loss": 0.8193,
      "num_input_tokens_seen": 76519008,
      "step": 131795
    },
    {
      "epoch": 19.630622579684243,
      "grad_norm": 0.06494140625,
      "learning_rate": 3.1185758340472876e-05,
      "loss": 0.796,
      "num_input_tokens_seen": 76521536,
      "step": 131800
    },
    {
      "epoch": 19.6313672922252,
      "grad_norm": 0.059814453125,
      "learning_rate": 3.106023001518188e-05,
      "loss": 0.8027,
      "num_input_tokens_seen": 76524352,
      "step": 131805
    },
    {
      "epoch": 19.63211200476616,
      "grad_norm": 0.043212890625,
      "learning_rate": 3.093495457491546e-05,
      "loss": 0.7771,
      "num_input_tokens_seen": 76527232,
      "step": 131810
    },
    {
      "epoch": 19.63285671730712,
      "grad_norm": 0.04296875,
      "learning_rate": 3.080993202178861e-05,
      "loss": 0.7982,
      "num_input_tokens_seen": 76530208,
      "step": 131815
    },
    {
      "epoch": 19.633601429848078,
      "grad_norm": 0.03759765625,
      "learning_rate": 3.0685162357912964e-05,
      "loss": 0.7977,
      "num_input_tokens_seen": 76533056,
      "step": 131820
    },
    {
      "epoch": 19.634346142389038,
      "grad_norm": 0.060791015625,
      "learning_rate": 3.056064558539851e-05,
      "loss": 0.7866,
      "num_input_tokens_seen": 76536000,
      "step": 131825
    },
    {
      "epoch": 19.63509085493,
      "grad_norm": 0.05126953125,
      "learning_rate": 3.0436381706348547e-05,
      "loss": 0.786,
      "num_input_tokens_seen": 76538720,
      "step": 131830
    },
    {
      "epoch": 19.635835567470956,
      "grad_norm": 0.035400390625,
      "learning_rate": 3.031237072285975e-05,
      "loss": 0.8057,
      "num_input_tokens_seen": 76541920,
      "step": 131835
    },
    {
      "epoch": 19.636580280011916,
      "grad_norm": 0.058837890625,
      "learning_rate": 3.018861263703043e-05,
      "loss": 0.8175,
      "num_input_tokens_seen": 76544992,
      "step": 131840
    },
    {
      "epoch": 19.637324992552873,
      "grad_norm": 0.0888671875,
      "learning_rate": 3.0065107450948923e-05,
      "loss": 0.7949,
      "num_input_tokens_seen": 76547808,
      "step": 131845
    },
    {
      "epoch": 19.638069705093834,
      "grad_norm": 0.03662109375,
      "learning_rate": 2.9941855166703557e-05,
      "loss": 0.8091,
      "num_input_tokens_seen": 76550880,
      "step": 131850
    },
    {
      "epoch": 19.638814417634794,
      "grad_norm": 0.091796875,
      "learning_rate": 2.981885578637433e-05,
      "loss": 0.7755,
      "num_input_tokens_seen": 76553664,
      "step": 131855
    },
    {
      "epoch": 19.63955913017575,
      "grad_norm": 0.03759765625,
      "learning_rate": 2.9696109312041252e-05,
      "loss": 0.779,
      "num_input_tokens_seen": 76556512,
      "step": 131860
    },
    {
      "epoch": 19.640303842716712,
      "grad_norm": 0.050048828125,
      "learning_rate": 2.957361574577766e-05,
      "loss": 0.7892,
      "num_input_tokens_seen": 76559264,
      "step": 131865
    },
    {
      "epoch": 19.64104855525767,
      "grad_norm": 0.1796875,
      "learning_rate": 2.9451375089651897e-05,
      "loss": 0.8174,
      "num_input_tokens_seen": 76562272,
      "step": 131870
    },
    {
      "epoch": 19.64179326779863,
      "grad_norm": 0.059814453125,
      "learning_rate": 2.9329387345730648e-05,
      "loss": 0.7927,
      "num_input_tokens_seen": 76565056,
      "step": 131875
    },
    {
      "epoch": 19.64253798033959,
      "grad_norm": 0.03369140625,
      "learning_rate": 2.920765251607227e-05,
      "loss": 0.8041,
      "num_input_tokens_seen": 76567776,
      "step": 131880
    },
    {
      "epoch": 19.643282692880547,
      "grad_norm": 0.04833984375,
      "learning_rate": 2.908617060273677e-05,
      "loss": 0.7813,
      "num_input_tokens_seen": 76570464,
      "step": 131885
    },
    {
      "epoch": 19.644027405421507,
      "grad_norm": 0.06494140625,
      "learning_rate": 2.8964941607774187e-05,
      "loss": 0.781,
      "num_input_tokens_seen": 76573088,
      "step": 131890
    },
    {
      "epoch": 19.644772117962468,
      "grad_norm": 0.043701171875,
      "learning_rate": 2.8843965533231206e-05,
      "loss": 0.7948,
      "num_input_tokens_seen": 76576128,
      "step": 131895
    },
    {
      "epoch": 19.645516830503425,
      "grad_norm": 0.0380859375,
      "learning_rate": 2.8723242381156198e-05,
      "loss": 0.8216,
      "num_input_tokens_seen": 76578976,
      "step": 131900
    },
    {
      "epoch": 19.646261543044385,
      "grad_norm": 0.037841796875,
      "learning_rate": 2.8602772153584197e-05,
      "loss": 0.7744,
      "num_input_tokens_seen": 76581696,
      "step": 131905
    },
    {
      "epoch": 19.647006255585342,
      "grad_norm": 0.050537109375,
      "learning_rate": 2.848255485255191e-05,
      "loss": 0.7921,
      "num_input_tokens_seen": 76584704,
      "step": 131910
    },
    {
      "epoch": 19.647750968126303,
      "grad_norm": 0.0673828125,
      "learning_rate": 2.836259048009271e-05,
      "loss": 0.7968,
      "num_input_tokens_seen": 76587872,
      "step": 131915
    },
    {
      "epoch": 19.648495680667263,
      "grad_norm": 0.10986328125,
      "learning_rate": 2.8242879038228307e-05,
      "loss": 0.7907,
      "num_input_tokens_seen": 76590784,
      "step": 131920
    },
    {
      "epoch": 19.64924039320822,
      "grad_norm": 0.036865234375,
      "learning_rate": 2.812342052898542e-05,
      "loss": 0.785,
      "num_input_tokens_seen": 76593632,
      "step": 131925
    },
    {
      "epoch": 19.64998510574918,
      "grad_norm": 0.052734375,
      "learning_rate": 2.8004214954380767e-05,
      "loss": 0.7892,
      "num_input_tokens_seen": 76596416,
      "step": 131930
    },
    {
      "epoch": 19.65072981829014,
      "grad_norm": 0.03857421875,
      "learning_rate": 2.788526231642774e-05,
      "loss": 0.8034,
      "num_input_tokens_seen": 76599648,
      "step": 131935
    },
    {
      "epoch": 19.6514745308311,
      "grad_norm": 0.0576171875,
      "learning_rate": 2.7766562617134726e-05,
      "loss": 0.8077,
      "num_input_tokens_seen": 76602784,
      "step": 131940
    },
    {
      "epoch": 19.65221924337206,
      "grad_norm": 0.053466796875,
      "learning_rate": 2.7648115858510123e-05,
      "loss": 0.7923,
      "num_input_tokens_seen": 76605536,
      "step": 131945
    },
    {
      "epoch": 19.652963955913016,
      "grad_norm": 0.055908203125,
      "learning_rate": 2.7529922042554e-05,
      "loss": 0.8042,
      "num_input_tokens_seen": 76608256,
      "step": 131950
    },
    {
      "epoch": 19.653708668453977,
      "grad_norm": 0.03857421875,
      "learning_rate": 2.7411981171259758e-05,
      "loss": 0.7981,
      "num_input_tokens_seen": 76611008,
      "step": 131955
    },
    {
      "epoch": 19.654453380994937,
      "grad_norm": 0.03955078125,
      "learning_rate": 2.72942932466258e-05,
      "loss": 0.8045,
      "num_input_tokens_seen": 76613952,
      "step": 131960
    },
    {
      "epoch": 19.655198093535894,
      "grad_norm": 0.05029296875,
      "learning_rate": 2.717685827063554e-05,
      "loss": 0.7953,
      "num_input_tokens_seen": 76616960,
      "step": 131965
    },
    {
      "epoch": 19.655942806076855,
      "grad_norm": 0.051513671875,
      "learning_rate": 2.7059676245274055e-05,
      "loss": 0.7904,
      "num_input_tokens_seen": 76619936,
      "step": 131970
    },
    {
      "epoch": 19.656687518617815,
      "grad_norm": 0.060791015625,
      "learning_rate": 2.6942747172523095e-05,
      "loss": 0.7934,
      "num_input_tokens_seen": 76622656,
      "step": 131975
    },
    {
      "epoch": 19.657432231158772,
      "grad_norm": 0.0546875,
      "learning_rate": 2.6826071054354417e-05,
      "loss": 0.7876,
      "num_input_tokens_seen": 76625536,
      "step": 131980
    },
    {
      "epoch": 19.658176943699733,
      "grad_norm": 0.028564453125,
      "learning_rate": 2.6709647892743105e-05,
      "loss": 0.8014,
      "num_input_tokens_seen": 76628512,
      "step": 131985
    },
    {
      "epoch": 19.65892165624069,
      "grad_norm": 0.057373046875,
      "learning_rate": 2.6593477689654255e-05,
      "loss": 0.8084,
      "num_input_tokens_seen": 76631456,
      "step": 131990
    },
    {
      "epoch": 19.65966636878165,
      "grad_norm": 0.0419921875,
      "learning_rate": 2.647756044704963e-05,
      "loss": 0.7869,
      "num_input_tokens_seen": 76634304,
      "step": 131995
    },
    {
      "epoch": 19.66041108132261,
      "grad_norm": 0.02734375,
      "learning_rate": 2.6361896166887664e-05,
      "loss": 0.8025,
      "num_input_tokens_seen": 76637280,
      "step": 132000
    },
    {
      "epoch": 19.661155793863568,
      "grad_norm": 0.053955078125,
      "learning_rate": 2.6246484851123462e-05,
      "loss": 0.7931,
      "num_input_tokens_seen": 76640160,
      "step": 132005
    },
    {
      "epoch": 19.66190050640453,
      "grad_norm": 0.03857421875,
      "learning_rate": 2.6131326501705464e-05,
      "loss": 0.7942,
      "num_input_tokens_seen": 76643168,
      "step": 132010
    },
    {
      "epoch": 19.662645218945485,
      "grad_norm": 0.040283203125,
      "learning_rate": 2.601642112058211e-05,
      "loss": 0.8009,
      "num_input_tokens_seen": 76645888,
      "step": 132015
    },
    {
      "epoch": 19.663389931486446,
      "grad_norm": 0.04736328125,
      "learning_rate": 2.5901768709690186e-05,
      "loss": 0.8126,
      "num_input_tokens_seen": 76648896,
      "step": 132020
    },
    {
      "epoch": 19.664134644027406,
      "grad_norm": 0.039306640625,
      "learning_rate": 2.5787369270969806e-05,
      "loss": 0.7998,
      "num_input_tokens_seen": 76652224,
      "step": 132025
    },
    {
      "epoch": 19.664879356568363,
      "grad_norm": 0.0654296875,
      "learning_rate": 2.5673222806352757e-05,
      "loss": 0.8021,
      "num_input_tokens_seen": 76655008,
      "step": 132030
    },
    {
      "epoch": 19.665624069109324,
      "grad_norm": 0.057861328125,
      "learning_rate": 2.5559329317767497e-05,
      "loss": 0.8162,
      "num_input_tokens_seen": 76657792,
      "step": 132035
    },
    {
      "epoch": 19.666368781650284,
      "grad_norm": 0.04638671875,
      "learning_rate": 2.544568880713749e-05,
      "loss": 0.8248,
      "num_input_tokens_seen": 76660512,
      "step": 132040
    },
    {
      "epoch": 19.66711349419124,
      "grad_norm": 0.30859375,
      "learning_rate": 2.533230127638286e-05,
      "loss": 0.8092,
      "num_input_tokens_seen": 76663488,
      "step": 132045
    },
    {
      "epoch": 19.667858206732202,
      "grad_norm": 0.07861328125,
      "learning_rate": 2.5219166727420417e-05,
      "loss": 0.8225,
      "num_input_tokens_seen": 76666144,
      "step": 132050
    },
    {
      "epoch": 19.66860291927316,
      "grad_norm": 0.0478515625,
      "learning_rate": 2.510628516216029e-05,
      "loss": 0.8013,
      "num_input_tokens_seen": 76668864,
      "step": 132055
    },
    {
      "epoch": 19.66934763181412,
      "grad_norm": 0.04736328125,
      "learning_rate": 2.499365658250929e-05,
      "loss": 0.8133,
      "num_input_tokens_seen": 76671936,
      "step": 132060
    },
    {
      "epoch": 19.67009234435508,
      "grad_norm": 0.04052734375,
      "learning_rate": 2.4881280990370902e-05,
      "loss": 0.8022,
      "num_input_tokens_seen": 76674880,
      "step": 132065
    },
    {
      "epoch": 19.670837056896037,
      "grad_norm": 0.044921875,
      "learning_rate": 2.476915838764193e-05,
      "loss": 0.7997,
      "num_input_tokens_seen": 76677600,
      "step": 132070
    },
    {
      "epoch": 19.671581769436997,
      "grad_norm": 0.0751953125,
      "learning_rate": 2.46572887762192e-05,
      "loss": 0.8129,
      "num_input_tokens_seen": 76680768,
      "step": 132075
    },
    {
      "epoch": 19.672326481977958,
      "grad_norm": 0.051025390625,
      "learning_rate": 2.4545672157991192e-05,
      "loss": 0.8127,
      "num_input_tokens_seen": 76683424,
      "step": 132080
    },
    {
      "epoch": 19.673071194518915,
      "grad_norm": 0.08251953125,
      "learning_rate": 2.4434308534843074e-05,
      "loss": 0.798,
      "num_input_tokens_seen": 76686432,
      "step": 132085
    },
    {
      "epoch": 19.673815907059875,
      "grad_norm": 0.037109375,
      "learning_rate": 2.4323197908658334e-05,
      "loss": 0.8101,
      "num_input_tokens_seen": 76689120,
      "step": 132090
    },
    {
      "epoch": 19.674560619600832,
      "grad_norm": 0.064453125,
      "learning_rate": 2.421234028131047e-05,
      "loss": 0.794,
      "num_input_tokens_seen": 76692096,
      "step": 132095
    },
    {
      "epoch": 19.675305332141793,
      "grad_norm": 0.07861328125,
      "learning_rate": 2.4101735654676324e-05,
      "loss": 0.7861,
      "num_input_tokens_seen": 76695040,
      "step": 132100
    },
    {
      "epoch": 19.676050044682754,
      "grad_norm": 0.06298828125,
      "learning_rate": 2.3991384030621066e-05,
      "loss": 0.7856,
      "num_input_tokens_seen": 76697952,
      "step": 132105
    },
    {
      "epoch": 19.67679475722371,
      "grad_norm": 0.076171875,
      "learning_rate": 2.388128541101153e-05,
      "loss": 0.7895,
      "num_input_tokens_seen": 76700512,
      "step": 132110
    },
    {
      "epoch": 19.67753946976467,
      "grad_norm": 0.0419921875,
      "learning_rate": 2.377143979770624e-05,
      "loss": 0.7897,
      "num_input_tokens_seen": 76703680,
      "step": 132115
    },
    {
      "epoch": 19.67828418230563,
      "grad_norm": 0.05712890625,
      "learning_rate": 2.366184719256037e-05,
      "loss": 0.7806,
      "num_input_tokens_seen": 76706400,
      "step": 132120
    },
    {
      "epoch": 19.67902889484659,
      "grad_norm": 0.055419921875,
      "learning_rate": 2.355250759742744e-05,
      "loss": 0.8163,
      "num_input_tokens_seen": 76709344,
      "step": 132125
    },
    {
      "epoch": 19.67977360738755,
      "grad_norm": 0.09326171875,
      "learning_rate": 2.3443421014154307e-05,
      "loss": 0.7931,
      "num_input_tokens_seen": 76712352,
      "step": 132130
    },
    {
      "epoch": 19.680518319928506,
      "grad_norm": 0.03125,
      "learning_rate": 2.3334587444581166e-05,
      "loss": 0.7902,
      "num_input_tokens_seen": 76715264,
      "step": 132135
    },
    {
      "epoch": 19.681263032469467,
      "grad_norm": 0.0556640625,
      "learning_rate": 2.3226006890549875e-05,
      "loss": 0.8062,
      "num_input_tokens_seen": 76718272,
      "step": 132140
    },
    {
      "epoch": 19.682007745010427,
      "grad_norm": 0.0301513671875,
      "learning_rate": 2.3117679353892307e-05,
      "loss": 0.7842,
      "num_input_tokens_seen": 76721408,
      "step": 132145
    },
    {
      "epoch": 19.682752457551384,
      "grad_norm": 0.049560546875,
      "learning_rate": 2.300960483644032e-05,
      "loss": 0.7973,
      "num_input_tokens_seen": 76724064,
      "step": 132150
    },
    {
      "epoch": 19.683497170092345,
      "grad_norm": 0.034912109375,
      "learning_rate": 2.2901783340019132e-05,
      "loss": 0.7914,
      "num_input_tokens_seen": 76726880,
      "step": 132155
    },
    {
      "epoch": 19.684241882633305,
      "grad_norm": 0.07470703125,
      "learning_rate": 2.2794214866450613e-05,
      "loss": 0.7803,
      "num_input_tokens_seen": 76729728,
      "step": 132160
    },
    {
      "epoch": 19.684986595174262,
      "grad_norm": 0.05029296875,
      "learning_rate": 2.2686899417551642e-05,
      "loss": 0.8011,
      "num_input_tokens_seen": 76732320,
      "step": 132165
    },
    {
      "epoch": 19.685731307715223,
      "grad_norm": 0.06494140625,
      "learning_rate": 2.257983699513577e-05,
      "loss": 0.7909,
      "num_input_tokens_seen": 76735424,
      "step": 132170
    },
    {
      "epoch": 19.68647602025618,
      "grad_norm": 0.33203125,
      "learning_rate": 2.2473027601009887e-05,
      "loss": 0.8208,
      "num_input_tokens_seen": 76738240,
      "step": 132175
    },
    {
      "epoch": 19.68722073279714,
      "grad_norm": 0.05908203125,
      "learning_rate": 2.2366471236980876e-05,
      "loss": 0.7886,
      "num_input_tokens_seen": 76741216,
      "step": 132180
    },
    {
      "epoch": 19.6879654453381,
      "grad_norm": 0.050048828125,
      "learning_rate": 2.2260167904847304e-05,
      "loss": 0.8051,
      "num_input_tokens_seen": 76743872,
      "step": 132185
    },
    {
      "epoch": 19.688710157879058,
      "grad_norm": 0.055419921875,
      "learning_rate": 2.21541176064044e-05,
      "loss": 0.7857,
      "num_input_tokens_seen": 76746752,
      "step": 132190
    },
    {
      "epoch": 19.68945487042002,
      "grad_norm": 0.052734375,
      "learning_rate": 2.2048320343445726e-05,
      "loss": 0.7932,
      "num_input_tokens_seen": 76749696,
      "step": 132195
    },
    {
      "epoch": 19.69019958296098,
      "grad_norm": 0.03662109375,
      "learning_rate": 2.194277611775819e-05,
      "loss": 0.7979,
      "num_input_tokens_seen": 76753024,
      "step": 132200
    },
    {
      "epoch": 19.690944295501936,
      "grad_norm": 0.038330078125,
      "learning_rate": 2.1837484931123697e-05,
      "loss": 0.8052,
      "num_input_tokens_seen": 76756160,
      "step": 132205
    },
    {
      "epoch": 19.691689008042896,
      "grad_norm": 0.06396484375,
      "learning_rate": 2.173244678532249e-05,
      "loss": 0.7876,
      "num_input_tokens_seen": 76759168,
      "step": 132210
    },
    {
      "epoch": 19.692433720583853,
      "grad_norm": 0.03564453125,
      "learning_rate": 2.162766168212815e-05,
      "loss": 0.7938,
      "num_input_tokens_seen": 76761760,
      "step": 132215
    },
    {
      "epoch": 19.693178433124814,
      "grad_norm": 0.040283203125,
      "learning_rate": 2.1523129623310932e-05,
      "loss": 0.7989,
      "num_input_tokens_seen": 76764672,
      "step": 132220
    },
    {
      "epoch": 19.693923145665774,
      "grad_norm": 0.0234375,
      "learning_rate": 2.141885061063775e-05,
      "loss": 0.7875,
      "num_input_tokens_seen": 76767328,
      "step": 132225
    },
    {
      "epoch": 19.69466785820673,
      "grad_norm": 0.03955078125,
      "learning_rate": 2.1314824645868867e-05,
      "loss": 0.8013,
      "num_input_tokens_seen": 76770016,
      "step": 132230
    },
    {
      "epoch": 19.695412570747692,
      "grad_norm": 0.0634765625,
      "learning_rate": 2.1211051730762876e-05,
      "loss": 0.8032,
      "num_input_tokens_seen": 76773312,
      "step": 132235
    },
    {
      "epoch": 19.69615728328865,
      "grad_norm": 0.0751953125,
      "learning_rate": 2.1107531867071703e-05,
      "loss": 0.8172,
      "num_input_tokens_seen": 76776512,
      "step": 132240
    },
    {
      "epoch": 19.69690199582961,
      "grad_norm": 0.0556640625,
      "learning_rate": 2.100426505654562e-05,
      "loss": 0.8071,
      "num_input_tokens_seen": 76779264,
      "step": 132245
    },
    {
      "epoch": 19.69764670837057,
      "grad_norm": 0.0556640625,
      "learning_rate": 2.090125130092657e-05,
      "loss": 0.7977,
      "num_input_tokens_seen": 76782240,
      "step": 132250
    },
    {
      "epoch": 19.698391420911527,
      "grad_norm": 0.03955078125,
      "learning_rate": 2.0798490601958154e-05,
      "loss": 0.8103,
      "num_input_tokens_seen": 76785056,
      "step": 132255
    },
    {
      "epoch": 19.699136133452487,
      "grad_norm": 0.046142578125,
      "learning_rate": 2.0695982961375646e-05,
      "loss": 0.7941,
      "num_input_tokens_seen": 76787744,
      "step": 132260
    },
    {
      "epoch": 19.699880845993448,
      "grad_norm": 0.036376953125,
      "learning_rate": 2.0593728380911003e-05,
      "loss": 0.8014,
      "num_input_tokens_seen": 76790976,
      "step": 132265
    },
    {
      "epoch": 19.700625558534405,
      "grad_norm": 0.029296875,
      "learning_rate": 2.0491726862289505e-05,
      "loss": 0.7889,
      "num_input_tokens_seen": 76793760,
      "step": 132270
    },
    {
      "epoch": 19.701370271075366,
      "grad_norm": 0.041015625,
      "learning_rate": 2.0389978407234775e-05,
      "loss": 0.8015,
      "num_input_tokens_seen": 76796640,
      "step": 132275
    },
    {
      "epoch": 19.702114983616323,
      "grad_norm": 0.04541015625,
      "learning_rate": 2.028848301746877e-05,
      "loss": 0.8146,
      "num_input_tokens_seen": 76799872,
      "step": 132280
    },
    {
      "epoch": 19.702859696157283,
      "grad_norm": 0.0458984375,
      "learning_rate": 2.0187240694703455e-05,
      "loss": 0.8047,
      "num_input_tokens_seen": 76802656,
      "step": 132285
    },
    {
      "epoch": 19.703604408698244,
      "grad_norm": 0.06884765625,
      "learning_rate": 2.0086251440649128e-05,
      "loss": 0.8068,
      "num_input_tokens_seen": 76805568,
      "step": 132290
    },
    {
      "epoch": 19.7043491212392,
      "grad_norm": 0.052490234375,
      "learning_rate": 1.9985515257012752e-05,
      "loss": 0.8089,
      "num_input_tokens_seen": 76808480,
      "step": 132295
    },
    {
      "epoch": 19.70509383378016,
      "grad_norm": 0.0517578125,
      "learning_rate": 1.988503214549797e-05,
      "loss": 0.8217,
      "num_input_tokens_seen": 76811328,
      "step": 132300
    },
    {
      "epoch": 19.70583854632112,
      "grad_norm": 0.042724609375,
      "learning_rate": 1.9784802107798427e-05,
      "loss": 0.8218,
      "num_input_tokens_seen": 76814400,
      "step": 132305
    },
    {
      "epoch": 19.70658325886208,
      "grad_norm": 0.04833984375,
      "learning_rate": 1.9684825145611096e-05,
      "loss": 0.8165,
      "num_input_tokens_seen": 76817088,
      "step": 132310
    },
    {
      "epoch": 19.70732797140304,
      "grad_norm": 0.0517578125,
      "learning_rate": 1.9585101260621297e-05,
      "loss": 0.7985,
      "num_input_tokens_seen": 76819840,
      "step": 132315
    },
    {
      "epoch": 19.708072683943996,
      "grad_norm": 0.036376953125,
      "learning_rate": 1.9485630454517677e-05,
      "loss": 0.8068,
      "num_input_tokens_seen": 76822880,
      "step": 132320
    },
    {
      "epoch": 19.708817396484957,
      "grad_norm": 0.0498046875,
      "learning_rate": 1.938641272897723e-05,
      "loss": 0.7961,
      "num_input_tokens_seen": 76826176,
      "step": 132325
    },
    {
      "epoch": 19.709562109025917,
      "grad_norm": 0.04248046875,
      "learning_rate": 1.9287448085678615e-05,
      "loss": 0.7964,
      "num_input_tokens_seen": 76829184,
      "step": 132330
    },
    {
      "epoch": 19.710306821566874,
      "grad_norm": 0.05029296875,
      "learning_rate": 1.9188736526293825e-05,
      "loss": 0.795,
      "num_input_tokens_seen": 76832064,
      "step": 132335
    },
    {
      "epoch": 19.711051534107835,
      "grad_norm": 0.044189453125,
      "learning_rate": 1.9090278052488195e-05,
      "loss": 0.7929,
      "num_input_tokens_seen": 76835040,
      "step": 132340
    },
    {
      "epoch": 19.711796246648795,
      "grad_norm": 0.09619140625,
      "learning_rate": 1.899207266592706e-05,
      "loss": 0.7976,
      "num_input_tokens_seen": 76837856,
      "step": 132345
    },
    {
      "epoch": 19.712540959189752,
      "grad_norm": 0.059814453125,
      "learning_rate": 1.889412036826743e-05,
      "loss": 0.7883,
      "num_input_tokens_seen": 76840672,
      "step": 132350
    },
    {
      "epoch": 19.713285671730713,
      "grad_norm": 0.043701171875,
      "learning_rate": 1.879642116116631e-05,
      "loss": 0.7955,
      "num_input_tokens_seen": 76843520,
      "step": 132355
    },
    {
      "epoch": 19.71403038427167,
      "grad_norm": 0.07470703125,
      "learning_rate": 1.8698975046274047e-05,
      "loss": 0.7925,
      "num_input_tokens_seen": 76846528,
      "step": 132360
    },
    {
      "epoch": 19.71477509681263,
      "grad_norm": 0.034423828125,
      "learning_rate": 1.860178202523599e-05,
      "loss": 0.7984,
      "num_input_tokens_seen": 76849344,
      "step": 132365
    },
    {
      "epoch": 19.71551980935359,
      "grad_norm": 0.095703125,
      "learning_rate": 1.8504842099694163e-05,
      "loss": 0.807,
      "num_input_tokens_seen": 76852704,
      "step": 132370
    },
    {
      "epoch": 19.716264521894548,
      "grad_norm": 0.03857421875,
      "learning_rate": 1.8408155271287253e-05,
      "loss": 0.797,
      "num_input_tokens_seen": 76855424,
      "step": 132375
    },
    {
      "epoch": 19.71700923443551,
      "grad_norm": 0.046630859375,
      "learning_rate": 1.8311721541647284e-05,
      "loss": 0.7989,
      "num_input_tokens_seen": 76858496,
      "step": 132380
    },
    {
      "epoch": 19.717753946976465,
      "grad_norm": 0.04248046875,
      "learning_rate": 1.8215540912404627e-05,
      "loss": 0.7834,
      "num_input_tokens_seen": 76860960,
      "step": 132385
    },
    {
      "epoch": 19.718498659517426,
      "grad_norm": 0.035400390625,
      "learning_rate": 1.8119613385182975e-05,
      "loss": 0.7986,
      "num_input_tokens_seen": 76863904,
      "step": 132390
    },
    {
      "epoch": 19.719243372058386,
      "grad_norm": 0.059814453125,
      "learning_rate": 1.8023938961604368e-05,
      "loss": 0.7848,
      "num_input_tokens_seen": 76866688,
      "step": 132395
    },
    {
      "epoch": 19.719988084599343,
      "grad_norm": 0.057861328125,
      "learning_rate": 1.7928517643282515e-05,
      "loss": 0.7828,
      "num_input_tokens_seen": 76869728,
      "step": 132400
    },
    {
      "epoch": 19.720732797140304,
      "grad_norm": 0.041015625,
      "learning_rate": 1.783334943183279e-05,
      "loss": 0.7998,
      "num_input_tokens_seen": 76872864,
      "step": 132405
    },
    {
      "epoch": 19.721477509681264,
      "grad_norm": 0.0732421875,
      "learning_rate": 1.7738434328860575e-05,
      "loss": 0.7935,
      "num_input_tokens_seen": 76875840,
      "step": 132410
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 0.039306640625,
      "learning_rate": 1.7643772335971253e-05,
      "loss": 0.8105,
      "num_input_tokens_seen": 76878848,
      "step": 132415
    },
    {
      "epoch": 19.722966934763182,
      "grad_norm": 0.045654296875,
      "learning_rate": 1.754936345476188e-05,
      "loss": 0.8126,
      "num_input_tokens_seen": 76881792,
      "step": 132420
    },
    {
      "epoch": 19.72371164730414,
      "grad_norm": 0.05419921875,
      "learning_rate": 1.745520768682951e-05,
      "loss": 0.795,
      "num_input_tokens_seen": 76884832,
      "step": 132425
    },
    {
      "epoch": 19.7244563598451,
      "grad_norm": 0.07470703125,
      "learning_rate": 1.736130503376454e-05,
      "loss": 0.78,
      "num_input_tokens_seen": 76887616,
      "step": 132430
    },
    {
      "epoch": 19.72520107238606,
      "grad_norm": 0.0439453125,
      "learning_rate": 1.7267655497150702e-05,
      "loss": 0.7985,
      "num_input_tokens_seen": 76890816,
      "step": 132435
    },
    {
      "epoch": 19.725945784927017,
      "grad_norm": 0.042724609375,
      "learning_rate": 1.7174259078573396e-05,
      "loss": 0.8014,
      "num_input_tokens_seen": 76894112,
      "step": 132440
    },
    {
      "epoch": 19.726690497467978,
      "grad_norm": 0.038330078125,
      "learning_rate": 1.7081115779608023e-05,
      "loss": 0.8194,
      "num_input_tokens_seen": 76897056,
      "step": 132445
    },
    {
      "epoch": 19.727435210008938,
      "grad_norm": 0.04150390625,
      "learning_rate": 1.6988225601829997e-05,
      "loss": 0.7935,
      "num_input_tokens_seen": 76899872,
      "step": 132450
    },
    {
      "epoch": 19.728179922549895,
      "grad_norm": 0.0458984375,
      "learning_rate": 1.689558854680806e-05,
      "loss": 0.824,
      "num_input_tokens_seen": 76902944,
      "step": 132455
    },
    {
      "epoch": 19.728924635090856,
      "grad_norm": 0.051513671875,
      "learning_rate": 1.6803204616105958e-05,
      "loss": 0.8126,
      "num_input_tokens_seen": 76905504,
      "step": 132460
    },
    {
      "epoch": 19.729669347631813,
      "grad_norm": 0.0654296875,
      "learning_rate": 1.671107381128578e-05,
      "loss": 0.8116,
      "num_input_tokens_seen": 76908640,
      "step": 132465
    },
    {
      "epoch": 19.730414060172773,
      "grad_norm": 0.044189453125,
      "learning_rate": 1.661919613390461e-05,
      "loss": 0.7993,
      "num_input_tokens_seen": 76911808,
      "step": 132470
    },
    {
      "epoch": 19.731158772713734,
      "grad_norm": 0.040283203125,
      "learning_rate": 1.6527571585512878e-05,
      "loss": 0.7987,
      "num_input_tokens_seen": 76914688,
      "step": 132475
    },
    {
      "epoch": 19.73190348525469,
      "grad_norm": 0.04638671875,
      "learning_rate": 1.6436200167659344e-05,
      "loss": 0.797,
      "num_input_tokens_seen": 76917536,
      "step": 132480
    },
    {
      "epoch": 19.73264819779565,
      "grad_norm": 0.0625,
      "learning_rate": 1.634508188188777e-05,
      "loss": 0.7745,
      "num_input_tokens_seen": 76920320,
      "step": 132485
    },
    {
      "epoch": 19.73339291033661,
      "grad_norm": 0.04296875,
      "learning_rate": 1.625421672973859e-05,
      "loss": 0.8221,
      "num_input_tokens_seen": 76923392,
      "step": 132490
    },
    {
      "epoch": 19.73413762287757,
      "grad_norm": 0.05224609375,
      "learning_rate": 1.6163604712743917e-05,
      "loss": 0.8115,
      "num_input_tokens_seen": 76926112,
      "step": 132495
    },
    {
      "epoch": 19.73488233541853,
      "grad_norm": 0.09814453125,
      "learning_rate": 1.6073245832435856e-05,
      "loss": 0.7813,
      "num_input_tokens_seen": 76929344,
      "step": 132500
    },
    {
      "epoch": 19.735627047959486,
      "grad_norm": 0.0400390625,
      "learning_rate": 1.5983140090343185e-05,
      "loss": 0.8128,
      "num_input_tokens_seen": 76932352,
      "step": 132505
    },
    {
      "epoch": 19.736371760500447,
      "grad_norm": 0.04833984375,
      "learning_rate": 1.5893287487984687e-05,
      "loss": 0.7942,
      "num_input_tokens_seen": 76935040,
      "step": 132510
    },
    {
      "epoch": 19.737116473041407,
      "grad_norm": 0.0556640625,
      "learning_rate": 1.5803688026880814e-05,
      "loss": 0.7745,
      "num_input_tokens_seen": 76937920,
      "step": 132515
    },
    {
      "epoch": 19.737861185582364,
      "grad_norm": 0.035400390625,
      "learning_rate": 1.571434170854369e-05,
      "loss": 0.7983,
      "num_input_tokens_seen": 76940544,
      "step": 132520
    },
    {
      "epoch": 19.738605898123325,
      "grad_norm": 0.0732421875,
      "learning_rate": 1.5625248534483772e-05,
      "loss": 0.8014,
      "num_input_tokens_seen": 76943488,
      "step": 132525
    },
    {
      "epoch": 19.73935061066428,
      "grad_norm": 0.04345703125,
      "learning_rate": 1.553640850620652e-05,
      "loss": 0.7877,
      "num_input_tokens_seen": 76946048,
      "step": 132530
    },
    {
      "epoch": 19.740095323205242,
      "grad_norm": 0.06640625,
      "learning_rate": 1.5447821625210745e-05,
      "loss": 0.7953,
      "num_input_tokens_seen": 76948960,
      "step": 132535
    },
    {
      "epoch": 19.740840035746203,
      "grad_norm": 0.06201171875,
      "learning_rate": 1.5359487892995237e-05,
      "loss": 0.8251,
      "num_input_tokens_seen": 76952064,
      "step": 132540
    },
    {
      "epoch": 19.74158474828716,
      "grad_norm": 0.041259765625,
      "learning_rate": 1.527140731105214e-05,
      "loss": 0.8091,
      "num_input_tokens_seen": 76954720,
      "step": 132545
    },
    {
      "epoch": 19.74232946082812,
      "grad_norm": 0.04248046875,
      "learning_rate": 1.5183579880868602e-05,
      "loss": 0.8012,
      "num_input_tokens_seen": 76957664,
      "step": 132550
    },
    {
      "epoch": 19.74307417336908,
      "grad_norm": 0.051513671875,
      "learning_rate": 1.5096005603926764e-05,
      "loss": 0.7874,
      "num_input_tokens_seen": 76960704,
      "step": 132555
    },
    {
      "epoch": 19.743818885910038,
      "grad_norm": 0.040283203125,
      "learning_rate": 1.5008684481710443e-05,
      "loss": 0.7903,
      "num_input_tokens_seen": 76963584,
      "step": 132560
    },
    {
      "epoch": 19.744563598451,
      "grad_norm": 0.05859375,
      "learning_rate": 1.4921616515690128e-05,
      "loss": 0.8017,
      "num_input_tokens_seen": 76966336,
      "step": 132565
    },
    {
      "epoch": 19.74530831099196,
      "grad_norm": 0.032470703125,
      "learning_rate": 1.483480170733964e-05,
      "loss": 0.8037,
      "num_input_tokens_seen": 76969216,
      "step": 132570
    },
    {
      "epoch": 19.746053023532916,
      "grad_norm": 0.036865234375,
      "learning_rate": 1.4748240058126138e-05,
      "loss": 0.7893,
      "num_input_tokens_seen": 76972640,
      "step": 132575
    },
    {
      "epoch": 19.746797736073876,
      "grad_norm": 0.050537109375,
      "learning_rate": 1.4661931569510123e-05,
      "loss": 0.8006,
      "num_input_tokens_seen": 76975488,
      "step": 132580
    },
    {
      "epoch": 19.747542448614833,
      "grad_norm": 0.0830078125,
      "learning_rate": 1.4575876242950425e-05,
      "loss": 0.7832,
      "num_input_tokens_seen": 76978272,
      "step": 132585
    },
    {
      "epoch": 19.748287161155794,
      "grad_norm": 0.06005859375,
      "learning_rate": 1.4490074079899218e-05,
      "loss": 0.795,
      "num_input_tokens_seen": 76981152,
      "step": 132590
    },
    {
      "epoch": 19.749031873696755,
      "grad_norm": 0.04931640625,
      "learning_rate": 1.4404525081810338e-05,
      "loss": 0.7943,
      "num_input_tokens_seen": 76983936,
      "step": 132595
    },
    {
      "epoch": 19.74977658623771,
      "grad_norm": 0.099609375,
      "learning_rate": 1.4319229250124299e-05,
      "loss": 0.7965,
      "num_input_tokens_seen": 76986944,
      "step": 132600
    },
    {
      "epoch": 19.750521298778672,
      "grad_norm": 0.02734375,
      "learning_rate": 1.4234186586284947e-05,
      "loss": 0.7809,
      "num_input_tokens_seen": 76990464,
      "step": 132605
    },
    {
      "epoch": 19.75126601131963,
      "grad_norm": 0.0498046875,
      "learning_rate": 1.4149397091727799e-05,
      "loss": 0.7934,
      "num_input_tokens_seen": 76993312,
      "step": 132610
    },
    {
      "epoch": 19.75201072386059,
      "grad_norm": 0.052978515625,
      "learning_rate": 1.4064860767885046e-05,
      "loss": 0.8022,
      "num_input_tokens_seen": 76996032,
      "step": 132615
    },
    {
      "epoch": 19.75275543640155,
      "grad_norm": 0.06103515625,
      "learning_rate": 1.3980577616187206e-05,
      "loss": 0.795,
      "num_input_tokens_seen": 76998880,
      "step": 132620
    },
    {
      "epoch": 19.753500148942507,
      "grad_norm": 0.04443359375,
      "learning_rate": 1.3896547638054812e-05,
      "loss": 0.7852,
      "num_input_tokens_seen": 77002016,
      "step": 132625
    },
    {
      "epoch": 19.754244861483468,
      "grad_norm": 0.02734375,
      "learning_rate": 1.381277083491006e-05,
      "loss": 0.7964,
      "num_input_tokens_seen": 77004672,
      "step": 132630
    },
    {
      "epoch": 19.754989574024428,
      "grad_norm": 0.06396484375,
      "learning_rate": 1.3729247208166816e-05,
      "loss": 0.8281,
      "num_input_tokens_seen": 77007584,
      "step": 132635
    },
    {
      "epoch": 19.755734286565385,
      "grad_norm": 0.046142578125,
      "learning_rate": 1.3645976759235623e-05,
      "loss": 0.7974,
      "num_input_tokens_seen": 77010528,
      "step": 132640
    },
    {
      "epoch": 19.756478999106346,
      "grad_norm": 0.044189453125,
      "learning_rate": 1.356295948952535e-05,
      "loss": 0.8173,
      "num_input_tokens_seen": 77013440,
      "step": 132645
    },
    {
      "epoch": 19.757223711647303,
      "grad_norm": 0.03759765625,
      "learning_rate": 1.3480195400436545e-05,
      "loss": 0.7838,
      "num_input_tokens_seen": 77016320,
      "step": 132650
    },
    {
      "epoch": 19.757968424188263,
      "grad_norm": 0.030517578125,
      "learning_rate": 1.3397684493369754e-05,
      "loss": 0.7925,
      "num_input_tokens_seen": 77019328,
      "step": 132655
    },
    {
      "epoch": 19.758713136729224,
      "grad_norm": 0.06689453125,
      "learning_rate": 1.3315426769715532e-05,
      "loss": 0.7894,
      "num_input_tokens_seen": 77022368,
      "step": 132660
    },
    {
      "epoch": 19.75945784927018,
      "grad_norm": 0.03564453125,
      "learning_rate": 1.3233422230867764e-05,
      "loss": 0.7892,
      "num_input_tokens_seen": 77025600,
      "step": 132665
    },
    {
      "epoch": 19.76020256181114,
      "grad_norm": 0.056640625,
      "learning_rate": 1.315167087820701e-05,
      "loss": 0.7915,
      "num_input_tokens_seen": 77028384,
      "step": 132670
    },
    {
      "epoch": 19.7609472743521,
      "grad_norm": 0.0361328125,
      "learning_rate": 1.3070172713118833e-05,
      "loss": 0.7995,
      "num_input_tokens_seen": 77030944,
      "step": 132675
    },
    {
      "epoch": 19.76169198689306,
      "grad_norm": 0.06884765625,
      "learning_rate": 1.2988927736977129e-05,
      "loss": 0.8098,
      "num_input_tokens_seen": 77033696,
      "step": 132680
    },
    {
      "epoch": 19.76243669943402,
      "grad_norm": 0.0281982421875,
      "learning_rate": 1.2907935951154136e-05,
      "loss": 0.7901,
      "num_input_tokens_seen": 77036576,
      "step": 132685
    },
    {
      "epoch": 19.763181411974976,
      "grad_norm": 0.0595703125,
      "learning_rate": 1.2827197357020425e-05,
      "loss": 0.7947,
      "num_input_tokens_seen": 77039584,
      "step": 132690
    },
    {
      "epoch": 19.763926124515937,
      "grad_norm": 0.039794921875,
      "learning_rate": 1.2746711955939905e-05,
      "loss": 0.7844,
      "num_input_tokens_seen": 77042560,
      "step": 132695
    },
    {
      "epoch": 19.764670837056897,
      "grad_norm": 0.06298828125,
      "learning_rate": 1.2666479749269821e-05,
      "loss": 0.7794,
      "num_input_tokens_seen": 77045440,
      "step": 132700
    },
    {
      "epoch": 19.765415549597854,
      "grad_norm": 0.28125,
      "learning_rate": 1.2586500738367423e-05,
      "loss": 0.832,
      "num_input_tokens_seen": 77048192,
      "step": 132705
    },
    {
      "epoch": 19.766160262138815,
      "grad_norm": 0.138671875,
      "learning_rate": 1.2506774924583297e-05,
      "loss": 0.7791,
      "num_input_tokens_seen": 77051072,
      "step": 132710
    },
    {
      "epoch": 19.766904974679775,
      "grad_norm": 0.33203125,
      "learning_rate": 1.2427302309266363e-05,
      "loss": 0.8311,
      "num_input_tokens_seen": 77053888,
      "step": 132715
    },
    {
      "epoch": 19.767649687220732,
      "grad_norm": 0.17578125,
      "learning_rate": 1.2348082893755551e-05,
      "loss": 0.7698,
      "num_input_tokens_seen": 77056800,
      "step": 132720
    },
    {
      "epoch": 19.768394399761693,
      "grad_norm": 0.041015625,
      "learning_rate": 1.2269116679391455e-05,
      "loss": 0.7986,
      "num_input_tokens_seen": 77059744,
      "step": 132725
    },
    {
      "epoch": 19.76913911230265,
      "grad_norm": 0.05126953125,
      "learning_rate": 1.2190403667509675e-05,
      "loss": 0.7904,
      "num_input_tokens_seen": 77062720,
      "step": 132730
    },
    {
      "epoch": 19.76988382484361,
      "grad_norm": 0.09033203125,
      "learning_rate": 1.2111943859435815e-05,
      "loss": 0.7792,
      "num_input_tokens_seen": 77065632,
      "step": 132735
    },
    {
      "epoch": 19.77062853738457,
      "grad_norm": 0.02880859375,
      "learning_rate": 1.2033737256498811e-05,
      "loss": 0.8013,
      "num_input_tokens_seen": 77068416,
      "step": 132740
    },
    {
      "epoch": 19.771373249925528,
      "grad_norm": 0.037841796875,
      "learning_rate": 1.195578386001761e-05,
      "loss": 0.7965,
      "num_input_tokens_seen": 77071424,
      "step": 132745
    },
    {
      "epoch": 19.77211796246649,
      "grad_norm": 0.03857421875,
      "learning_rate": 1.1878083671311157e-05,
      "loss": 0.7998,
      "num_input_tokens_seen": 77074336,
      "step": 132750
    },
    {
      "epoch": 19.772862675007445,
      "grad_norm": 0.052734375,
      "learning_rate": 1.1800636691691735e-05,
      "loss": 0.8012,
      "num_input_tokens_seen": 77077376,
      "step": 132755
    },
    {
      "epoch": 19.773607387548406,
      "grad_norm": 0.0302734375,
      "learning_rate": 1.172344292246663e-05,
      "loss": 0.8114,
      "num_input_tokens_seen": 77080032,
      "step": 132760
    },
    {
      "epoch": 19.774352100089367,
      "grad_norm": 0.055908203125,
      "learning_rate": 1.1646502364939803e-05,
      "loss": 0.7954,
      "num_input_tokens_seen": 77082720,
      "step": 132765
    },
    {
      "epoch": 19.775096812630323,
      "grad_norm": 0.080078125,
      "learning_rate": 1.1569815020411878e-05,
      "loss": 0.7888,
      "num_input_tokens_seen": 77085888,
      "step": 132770
    },
    {
      "epoch": 19.775841525171284,
      "grad_norm": 0.0478515625,
      "learning_rate": 1.149338089018015e-05,
      "loss": 0.8131,
      "num_input_tokens_seen": 77088960,
      "step": 132775
    },
    {
      "epoch": 19.776586237712245,
      "grad_norm": 0.0458984375,
      "learning_rate": 1.1417199975531922e-05,
      "loss": 0.8213,
      "num_input_tokens_seen": 77092064,
      "step": 132780
    },
    {
      "epoch": 19.7773309502532,
      "grad_norm": 0.03564453125,
      "learning_rate": 1.134127227775783e-05,
      "loss": 0.7941,
      "num_input_tokens_seen": 77094880,
      "step": 132785
    },
    {
      "epoch": 19.778075662794162,
      "grad_norm": 0.040771484375,
      "learning_rate": 1.1265597798138515e-05,
      "loss": 0.8025,
      "num_input_tokens_seen": 77097568,
      "step": 132790
    },
    {
      "epoch": 19.77882037533512,
      "grad_norm": 0.061767578125,
      "learning_rate": 1.119017653795462e-05,
      "loss": 0.792,
      "num_input_tokens_seen": 77100416,
      "step": 132795
    },
    {
      "epoch": 19.77956508787608,
      "grad_norm": 0.056884765625,
      "learning_rate": 1.1115008498476797e-05,
      "loss": 0.7817,
      "num_input_tokens_seen": 77103360,
      "step": 132800
    },
    {
      "epoch": 19.78030980041704,
      "grad_norm": 0.061279296875,
      "learning_rate": 1.1040093680977358e-05,
      "loss": 0.7977,
      "num_input_tokens_seen": 77106272,
      "step": 132805
    },
    {
      "epoch": 19.781054512957997,
      "grad_norm": 0.046875,
      "learning_rate": 1.096543208672196e-05,
      "loss": 0.7774,
      "num_input_tokens_seen": 77109024,
      "step": 132810
    },
    {
      "epoch": 19.781799225498958,
      "grad_norm": 0.03955078125,
      "learning_rate": 1.089102371697126e-05,
      "loss": 0.7912,
      "num_input_tokens_seen": 77111808,
      "step": 132815
    },
    {
      "epoch": 19.782543938039918,
      "grad_norm": 0.05029296875,
      "learning_rate": 1.081686857298092e-05,
      "loss": 0.8034,
      "num_input_tokens_seen": 77114816,
      "step": 132820
    },
    {
      "epoch": 19.783288650580875,
      "grad_norm": 0.04296875,
      "learning_rate": 1.0742966656006603e-05,
      "loss": 0.788,
      "num_input_tokens_seen": 77117728,
      "step": 132825
    },
    {
      "epoch": 19.784033363121836,
      "grad_norm": 0.050537109375,
      "learning_rate": 1.0669317967295643e-05,
      "loss": 0.7847,
      "num_input_tokens_seen": 77120416,
      "step": 132830
    },
    {
      "epoch": 19.784778075662793,
      "grad_norm": 0.05322265625,
      "learning_rate": 1.0595922508092047e-05,
      "loss": 0.7831,
      "num_input_tokens_seen": 77123648,
      "step": 132835
    },
    {
      "epoch": 19.785522788203753,
      "grad_norm": 0.048095703125,
      "learning_rate": 1.052278027963649e-05,
      "loss": 0.8007,
      "num_input_tokens_seen": 77126464,
      "step": 132840
    },
    {
      "epoch": 19.786267500744714,
      "grad_norm": 0.03564453125,
      "learning_rate": 1.0449891283162982e-05,
      "loss": 0.7916,
      "num_input_tokens_seen": 77129312,
      "step": 132845
    },
    {
      "epoch": 19.78701221328567,
      "grad_norm": 0.033203125,
      "learning_rate": 1.0377255519903871e-05,
      "loss": 0.7927,
      "num_input_tokens_seen": 77132032,
      "step": 132850
    },
    {
      "epoch": 19.78775692582663,
      "grad_norm": 0.08984375,
      "learning_rate": 1.0304872991086511e-05,
      "loss": 0.7899,
      "num_input_tokens_seen": 77134880,
      "step": 132855
    },
    {
      "epoch": 19.788501638367592,
      "grad_norm": 0.041748046875,
      "learning_rate": 1.0232743697933255e-05,
      "loss": 0.7937,
      "num_input_tokens_seen": 77137472,
      "step": 132860
    },
    {
      "epoch": 19.78924635090855,
      "grad_norm": 0.111328125,
      "learning_rate": 1.0160867641663129e-05,
      "loss": 0.8061,
      "num_input_tokens_seen": 77140320,
      "step": 132865
    },
    {
      "epoch": 19.78999106344951,
      "grad_norm": 0.04296875,
      "learning_rate": 1.0089244823490161e-05,
      "loss": 0.7952,
      "num_input_tokens_seen": 77143104,
      "step": 132870
    },
    {
      "epoch": 19.790735775990466,
      "grad_norm": 0.05517578125,
      "learning_rate": 1.0017875244623386e-05,
      "loss": 0.7957,
      "num_input_tokens_seen": 77146240,
      "step": 132875
    },
    {
      "epoch": 19.791480488531427,
      "grad_norm": 0.046875,
      "learning_rate": 9.94675890627017e-06,
      "loss": 0.8091,
      "num_input_tokens_seen": 77149344,
      "step": 132880
    },
    {
      "epoch": 19.792225201072387,
      "grad_norm": 0.0615234375,
      "learning_rate": 9.87589580963122e-06,
      "loss": 0.793,
      "num_input_tokens_seen": 77152352,
      "step": 132885
    },
    {
      "epoch": 19.792969913613344,
      "grad_norm": 0.05712890625,
      "learning_rate": 9.805285955903908e-06,
      "loss": 0.7913,
      "num_input_tokens_seen": 77155168,
      "step": 132890
    },
    {
      "epoch": 19.793714626154305,
      "grad_norm": 0.05615234375,
      "learning_rate": 9.734929346280618e-06,
      "loss": 0.798,
      "num_input_tokens_seen": 77157920,
      "step": 132895
    },
    {
      "epoch": 19.794459338695262,
      "grad_norm": 0.051513671875,
      "learning_rate": 9.664825981950398e-06,
      "loss": 0.7885,
      "num_input_tokens_seen": 77160896,
      "step": 132900
    },
    {
      "epoch": 19.795204051236222,
      "grad_norm": 0.037353515625,
      "learning_rate": 9.594975864097299e-06,
      "loss": 0.7921,
      "num_input_tokens_seen": 77163904,
      "step": 132905
    },
    {
      "epoch": 19.795948763777183,
      "grad_norm": 0.039306640625,
      "learning_rate": 9.525378993902045e-06,
      "loss": 0.7833,
      "num_input_tokens_seen": 77166688,
      "step": 132910
    },
    {
      "epoch": 19.79669347631814,
      "grad_norm": 0.051025390625,
      "learning_rate": 9.45603537254036e-06,
      "loss": 0.7916,
      "num_input_tokens_seen": 77169760,
      "step": 132915
    },
    {
      "epoch": 19.7974381888591,
      "grad_norm": 0.05419921875,
      "learning_rate": 9.386945001181312e-06,
      "loss": 0.797,
      "num_input_tokens_seen": 77172704,
      "step": 132920
    },
    {
      "epoch": 19.79818290140006,
      "grad_norm": 0.040283203125,
      "learning_rate": 9.31810788099563e-06,
      "loss": 0.801,
      "num_input_tokens_seen": 77175680,
      "step": 132925
    },
    {
      "epoch": 19.798927613941018,
      "grad_norm": 0.0439453125,
      "learning_rate": 9.24952401314405e-06,
      "loss": 0.8083,
      "num_input_tokens_seen": 77178784,
      "step": 132930
    },
    {
      "epoch": 19.79967232648198,
      "grad_norm": 0.037841796875,
      "learning_rate": 9.181193398787313e-06,
      "loss": 0.809,
      "num_input_tokens_seen": 77181536,
      "step": 132935
    },
    {
      "epoch": 19.800417039022935,
      "grad_norm": 0.03662109375,
      "learning_rate": 9.113116039076164e-06,
      "loss": 0.7916,
      "num_input_tokens_seen": 77184512,
      "step": 132940
    },
    {
      "epoch": 19.801161751563896,
      "grad_norm": 0.041748046875,
      "learning_rate": 9.045291935164679e-06,
      "loss": 0.8111,
      "num_input_tokens_seen": 77187200,
      "step": 132945
    },
    {
      "epoch": 19.801906464104857,
      "grad_norm": 0.07080078125,
      "learning_rate": 8.977721088195277e-06,
      "loss": 0.7867,
      "num_input_tokens_seen": 77189952,
      "step": 132950
    },
    {
      "epoch": 19.802651176645814,
      "grad_norm": 0.06494140625,
      "learning_rate": 8.910403499312046e-06,
      "loss": 0.7985,
      "num_input_tokens_seen": 77192896,
      "step": 132955
    },
    {
      "epoch": 19.803395889186774,
      "grad_norm": 0.036376953125,
      "learning_rate": 8.843339169650744e-06,
      "loss": 0.7884,
      "num_input_tokens_seen": 77195872,
      "step": 132960
    },
    {
      "epoch": 19.804140601727735,
      "grad_norm": 0.034423828125,
      "learning_rate": 8.77652810034546e-06,
      "loss": 0.8095,
      "num_input_tokens_seen": 77198752,
      "step": 132965
    },
    {
      "epoch": 19.80488531426869,
      "grad_norm": 0.042724609375,
      "learning_rate": 8.709970292523627e-06,
      "loss": 0.7838,
      "num_input_tokens_seen": 77201600,
      "step": 132970
    },
    {
      "epoch": 19.805630026809652,
      "grad_norm": 0.044921875,
      "learning_rate": 8.643665747311014e-06,
      "loss": 0.7999,
      "num_input_tokens_seen": 77204512,
      "step": 132975
    },
    {
      "epoch": 19.80637473935061,
      "grad_norm": 0.045654296875,
      "learning_rate": 8.577614465826722e-06,
      "loss": 0.8013,
      "num_input_tokens_seen": 77207136,
      "step": 132980
    },
    {
      "epoch": 19.80711945189157,
      "grad_norm": 0.035888671875,
      "learning_rate": 8.511816449186526e-06,
      "loss": 0.7742,
      "num_input_tokens_seen": 77210112,
      "step": 132985
    },
    {
      "epoch": 19.80786416443253,
      "grad_norm": 0.076171875,
      "learning_rate": 8.446271698504537e-06,
      "loss": 0.7891,
      "num_input_tokens_seen": 77213088,
      "step": 132990
    },
    {
      "epoch": 19.808608876973487,
      "grad_norm": 0.038818359375,
      "learning_rate": 8.380980214883204e-06,
      "loss": 0.8008,
      "num_input_tokens_seen": 77215968,
      "step": 132995
    },
    {
      "epoch": 19.809353589514448,
      "grad_norm": 0.0286865234375,
      "learning_rate": 8.315941999429977e-06,
      "loss": 0.8189,
      "num_input_tokens_seen": 77218912,
      "step": 133000
    },
    {
      "epoch": 19.81009830205541,
      "grad_norm": 0.031494140625,
      "learning_rate": 8.251157053242308e-06,
      "loss": 0.8062,
      "num_input_tokens_seen": 77221760,
      "step": 133005
    },
    {
      "epoch": 19.810843014596365,
      "grad_norm": 0.026611328125,
      "learning_rate": 8.186625377414324e-06,
      "loss": 0.7995,
      "num_input_tokens_seen": 77224672,
      "step": 133010
    },
    {
      "epoch": 19.811587727137326,
      "grad_norm": 0.0274658203125,
      "learning_rate": 8.12234697303682e-06,
      "loss": 0.789,
      "num_input_tokens_seen": 77227712,
      "step": 133015
    },
    {
      "epoch": 19.812332439678283,
      "grad_norm": 0.044921875,
      "learning_rate": 8.058321841193927e-06,
      "loss": 0.8123,
      "num_input_tokens_seen": 77230528,
      "step": 133020
    },
    {
      "epoch": 19.813077152219243,
      "grad_norm": 0.031005859375,
      "learning_rate": 7.994549982968113e-06,
      "loss": 0.8023,
      "num_input_tokens_seen": 77233344,
      "step": 133025
    },
    {
      "epoch": 19.813821864760204,
      "grad_norm": 0.054443359375,
      "learning_rate": 7.931031399438514e-06,
      "loss": 0.8015,
      "num_input_tokens_seen": 77236384,
      "step": 133030
    },
    {
      "epoch": 19.81456657730116,
      "grad_norm": 0.041748046875,
      "learning_rate": 7.867766091677608e-06,
      "loss": 0.8002,
      "num_input_tokens_seen": 77239520,
      "step": 133035
    },
    {
      "epoch": 19.81531128984212,
      "grad_norm": 0.047607421875,
      "learning_rate": 7.804754060751206e-06,
      "loss": 0.7945,
      "num_input_tokens_seen": 77242432,
      "step": 133040
    },
    {
      "epoch": 19.816056002383082,
      "grad_norm": 0.0390625,
      "learning_rate": 7.741995307728455e-06,
      "loss": 0.7995,
      "num_input_tokens_seen": 77245280,
      "step": 133045
    },
    {
      "epoch": 19.81680071492404,
      "grad_norm": 0.055419921875,
      "learning_rate": 7.679489833665176e-06,
      "loss": 0.8007,
      "num_input_tokens_seen": 77248320,
      "step": 133050
    },
    {
      "epoch": 19.817545427465,
      "grad_norm": 0.045166015625,
      "learning_rate": 7.6172376396205215e-06,
      "loss": 0.7891,
      "num_input_tokens_seen": 77251424,
      "step": 133055
    },
    {
      "epoch": 19.818290140005956,
      "grad_norm": 0.03369140625,
      "learning_rate": 7.555238726645319e-06,
      "loss": 0.8031,
      "num_input_tokens_seen": 77254432,
      "step": 133060
    },
    {
      "epoch": 19.819034852546917,
      "grad_norm": 0.0703125,
      "learning_rate": 7.493493095785397e-06,
      "loss": 0.7738,
      "num_input_tokens_seen": 77257600,
      "step": 133065
    },
    {
      "epoch": 19.819779565087877,
      "grad_norm": 0.0291748046875,
      "learning_rate": 7.432000748086586e-06,
      "loss": 0.7967,
      "num_input_tokens_seen": 77260512,
      "step": 133070
    },
    {
      "epoch": 19.820524277628834,
      "grad_norm": 0.040771484375,
      "learning_rate": 7.370761684584726e-06,
      "loss": 0.7935,
      "num_input_tokens_seen": 77263840,
      "step": 133075
    },
    {
      "epoch": 19.821268990169795,
      "grad_norm": 0.12060546875,
      "learning_rate": 7.309775906317317e-06,
      "loss": 0.7974,
      "num_input_tokens_seen": 77266752,
      "step": 133080
    },
    {
      "epoch": 19.822013702710755,
      "grad_norm": 0.040771484375,
      "learning_rate": 7.249043414313538e-06,
      "loss": 0.8106,
      "num_input_tokens_seen": 77269568,
      "step": 133085
    },
    {
      "epoch": 19.822758415251712,
      "grad_norm": 0.037841796875,
      "learning_rate": 7.188564209599235e-06,
      "loss": 0.8006,
      "num_input_tokens_seen": 77272416,
      "step": 133090
    },
    {
      "epoch": 19.823503127792673,
      "grad_norm": 0.043701171875,
      "learning_rate": 7.128338293195257e-06,
      "loss": 0.784,
      "num_input_tokens_seen": 77274912,
      "step": 133095
    },
    {
      "epoch": 19.82424784033363,
      "grad_norm": 0.03515625,
      "learning_rate": 7.06836566612079e-06,
      "loss": 0.8047,
      "num_input_tokens_seen": 77277632,
      "step": 133100
    },
    {
      "epoch": 19.82499255287459,
      "grad_norm": 0.08984375,
      "learning_rate": 7.008646329386692e-06,
      "loss": 0.7671,
      "num_input_tokens_seen": 77280864,
      "step": 133105
    },
    {
      "epoch": 19.82573726541555,
      "grad_norm": 0.04150390625,
      "learning_rate": 6.949180284005485e-06,
      "loss": 0.7903,
      "num_input_tokens_seen": 77283680,
      "step": 133110
    },
    {
      "epoch": 19.826481977956508,
      "grad_norm": 0.0693359375,
      "learning_rate": 6.889967530978036e-06,
      "loss": 0.8127,
      "num_input_tokens_seen": 77286752,
      "step": 133115
    },
    {
      "epoch": 19.82722669049747,
      "grad_norm": 0.042724609375,
      "learning_rate": 6.831008071306876e-06,
      "loss": 0.7749,
      "num_input_tokens_seen": 77289632,
      "step": 133120
    },
    {
      "epoch": 19.827971403038426,
      "grad_norm": 0.052734375,
      "learning_rate": 6.772301905987876e-06,
      "loss": 0.8339,
      "num_input_tokens_seen": 77292512,
      "step": 133125
    },
    {
      "epoch": 19.828716115579386,
      "grad_norm": 0.038330078125,
      "learning_rate": 6.71384903601191e-06,
      "loss": 0.7903,
      "num_input_tokens_seen": 77295232,
      "step": 133130
    },
    {
      "epoch": 19.829460828120347,
      "grad_norm": 0.043212890625,
      "learning_rate": 6.655649462366519e-06,
      "loss": 0.7957,
      "num_input_tokens_seen": 77298112,
      "step": 133135
    },
    {
      "epoch": 19.830205540661304,
      "grad_norm": 0.06591796875,
      "learning_rate": 6.597703186035919e-06,
      "loss": 0.7986,
      "num_input_tokens_seen": 77301248,
      "step": 133140
    },
    {
      "epoch": 19.830950253202264,
      "grad_norm": 0.046875,
      "learning_rate": 6.5400102079976595e-06,
      "loss": 0.8137,
      "num_input_tokens_seen": 77304096,
      "step": 133145
    },
    {
      "epoch": 19.831694965743225,
      "grad_norm": 0.058349609375,
      "learning_rate": 6.482570529227627e-06,
      "loss": 0.7789,
      "num_input_tokens_seen": 77306880,
      "step": 133150
    },
    {
      "epoch": 19.83243967828418,
      "grad_norm": 0.053466796875,
      "learning_rate": 6.425384150696711e-06,
      "loss": 0.7903,
      "num_input_tokens_seen": 77309664,
      "step": 133155
    },
    {
      "epoch": 19.833184390825142,
      "grad_norm": 0.03955078125,
      "learning_rate": 6.368451073369141e-06,
      "loss": 0.7803,
      "num_input_tokens_seen": 77312800,
      "step": 133160
    },
    {
      "epoch": 19.8339291033661,
      "grad_norm": 0.031982421875,
      "learning_rate": 6.311771298209145e-06,
      "loss": 0.7858,
      "num_input_tokens_seen": 77315648,
      "step": 133165
    },
    {
      "epoch": 19.83467381590706,
      "grad_norm": 0.04638671875,
      "learning_rate": 6.255344826170961e-06,
      "loss": 0.7869,
      "num_input_tokens_seen": 77318688,
      "step": 133170
    },
    {
      "epoch": 19.83541852844802,
      "grad_norm": 0.03759765625,
      "learning_rate": 6.1991716582104896e-06,
      "loss": 0.8196,
      "num_input_tokens_seen": 77321472,
      "step": 133175
    },
    {
      "epoch": 19.836163240988977,
      "grad_norm": 0.047119140625,
      "learning_rate": 6.143251795276971e-06,
      "loss": 0.8144,
      "num_input_tokens_seen": 77324384,
      "step": 133180
    },
    {
      "epoch": 19.836907953529938,
      "grad_norm": 0.04638671875,
      "learning_rate": 6.087585238312986e-06,
      "loss": 0.7957,
      "num_input_tokens_seen": 77327264,
      "step": 133185
    },
    {
      "epoch": 19.8376526660709,
      "grad_norm": 0.0380859375,
      "learning_rate": 6.032171988261114e-06,
      "loss": 0.7954,
      "num_input_tokens_seen": 77330272,
      "step": 133190
    },
    {
      "epoch": 19.838397378611855,
      "grad_norm": 0.06005859375,
      "learning_rate": 5.977012046055607e-06,
      "loss": 0.803,
      "num_input_tokens_seen": 77333216,
      "step": 133195
    },
    {
      "epoch": 19.839142091152816,
      "grad_norm": 0.04150390625,
      "learning_rate": 5.9221054126290526e-06,
      "loss": 0.8105,
      "num_input_tokens_seen": 77335840,
      "step": 133200
    },
    {
      "epoch": 19.839886803693773,
      "grad_norm": 0.053466796875,
      "learning_rate": 5.867452088910707e-06,
      "loss": 0.7732,
      "num_input_tokens_seen": 77338528,
      "step": 133205
    },
    {
      "epoch": 19.840631516234733,
      "grad_norm": 0.03955078125,
      "learning_rate": 5.813052075821501e-06,
      "loss": 0.8061,
      "num_input_tokens_seen": 77341472,
      "step": 133210
    },
    {
      "epoch": 19.841376228775694,
      "grad_norm": 0.031005859375,
      "learning_rate": 5.7589053742806985e-06,
      "loss": 0.8085,
      "num_input_tokens_seen": 77344320,
      "step": 133215
    },
    {
      "epoch": 19.84212094131665,
      "grad_norm": 0.055419921875,
      "learning_rate": 5.705011985204233e-06,
      "loss": 0.7971,
      "num_input_tokens_seen": 77347200,
      "step": 133220
    },
    {
      "epoch": 19.84286565385761,
      "grad_norm": 0.03515625,
      "learning_rate": 5.651371909501379e-06,
      "loss": 0.7889,
      "num_input_tokens_seen": 77349856,
      "step": 133225
    },
    {
      "epoch": 19.843610366398572,
      "grad_norm": 0.05810546875,
      "learning_rate": 5.597985148079743e-06,
      "loss": 0.801,
      "num_input_tokens_seen": 77352480,
      "step": 133230
    },
    {
      "epoch": 19.84435507893953,
      "grad_norm": 0.032470703125,
      "learning_rate": 5.544851701841935e-06,
      "loss": 0.7975,
      "num_input_tokens_seen": 77355168,
      "step": 133235
    },
    {
      "epoch": 19.84509979148049,
      "grad_norm": 0.08349609375,
      "learning_rate": 5.491971571682241e-06,
      "loss": 0.7769,
      "num_input_tokens_seen": 77358240,
      "step": 133240
    },
    {
      "epoch": 19.845844504021446,
      "grad_norm": 0.051513671875,
      "learning_rate": 5.439344758496611e-06,
      "loss": 0.802,
      "num_input_tokens_seen": 77361120,
      "step": 133245
    },
    {
      "epoch": 19.846589216562407,
      "grad_norm": 0.056396484375,
      "learning_rate": 5.386971263172668e-06,
      "loss": 0.7851,
      "num_input_tokens_seen": 77364000,
      "step": 133250
    },
    {
      "epoch": 19.847333929103367,
      "grad_norm": 0.050048828125,
      "learning_rate": 5.334851086596371e-06,
      "loss": 0.8137,
      "num_input_tokens_seen": 77366752,
      "step": 133255
    },
    {
      "epoch": 19.848078641644324,
      "grad_norm": 0.0301513671875,
      "learning_rate": 5.282984229648679e-06,
      "loss": 0.8088,
      "num_input_tokens_seen": 77369728,
      "step": 133260
    },
    {
      "epoch": 19.848823354185285,
      "grad_norm": 0.0537109375,
      "learning_rate": 5.231370693203896e-06,
      "loss": 0.8094,
      "num_input_tokens_seen": 77372896,
      "step": 133265
    },
    {
      "epoch": 19.849568066726242,
      "grad_norm": 0.043701171875,
      "learning_rate": 5.180010478136321e-06,
      "loss": 0.7803,
      "num_input_tokens_seen": 77375840,
      "step": 133270
    },
    {
      "epoch": 19.850312779267203,
      "grad_norm": 0.03759765625,
      "learning_rate": 5.128903585311927e-06,
      "loss": 0.7895,
      "num_input_tokens_seen": 77378688,
      "step": 133275
    },
    {
      "epoch": 19.851057491808163,
      "grad_norm": 0.03662109375,
      "learning_rate": 5.078050015595026e-06,
      "loss": 0.8288,
      "num_input_tokens_seen": 77381760,
      "step": 133280
    },
    {
      "epoch": 19.85180220434912,
      "grad_norm": 0.054931640625,
      "learning_rate": 5.027449769843262e-06,
      "loss": 0.7944,
      "num_input_tokens_seen": 77384352,
      "step": 133285
    },
    {
      "epoch": 19.85254691689008,
      "grad_norm": 0.043701171875,
      "learning_rate": 4.977102848912617e-06,
      "loss": 0.8058,
      "num_input_tokens_seen": 77387264,
      "step": 133290
    },
    {
      "epoch": 19.85329162943104,
      "grad_norm": 0.05615234375,
      "learning_rate": 4.92700925365408e-06,
      "loss": 0.7952,
      "num_input_tokens_seen": 77389888,
      "step": 133295
    },
    {
      "epoch": 19.854036341971998,
      "grad_norm": 0.034423828125,
      "learning_rate": 4.8771689849136375e-06,
      "loss": 0.8003,
      "num_input_tokens_seen": 77392928,
      "step": 133300
    },
    {
      "epoch": 19.85478105451296,
      "grad_norm": 0.0791015625,
      "learning_rate": 4.827582043532286e-06,
      "loss": 0.7923,
      "num_input_tokens_seen": 77395936,
      "step": 133305
    },
    {
      "epoch": 19.855525767053916,
      "grad_norm": 0.0294189453125,
      "learning_rate": 4.778248430349352e-06,
      "loss": 0.8104,
      "num_input_tokens_seen": 77398848,
      "step": 133310
    },
    {
      "epoch": 19.856270479594876,
      "grad_norm": 0.0556640625,
      "learning_rate": 4.729168146197504e-06,
      "loss": 0.7907,
      "num_input_tokens_seen": 77401664,
      "step": 133315
    },
    {
      "epoch": 19.857015192135837,
      "grad_norm": 0.0634765625,
      "learning_rate": 4.680341191904413e-06,
      "loss": 0.8159,
      "num_input_tokens_seen": 77404608,
      "step": 133320
    },
    {
      "epoch": 19.857759904676794,
      "grad_norm": 0.0228271484375,
      "learning_rate": 4.631767568297751e-06,
      "loss": 0.7935,
      "num_input_tokens_seen": 77407552,
      "step": 133325
    },
    {
      "epoch": 19.858504617217754,
      "grad_norm": 0.0537109375,
      "learning_rate": 4.5834472761968615e-06,
      "loss": 0.7761,
      "num_input_tokens_seen": 77410624,
      "step": 133330
    },
    {
      "epoch": 19.859249329758715,
      "grad_norm": 0.076171875,
      "learning_rate": 4.535380316417758e-06,
      "loss": 0.8052,
      "num_input_tokens_seen": 77413344,
      "step": 133335
    },
    {
      "epoch": 19.85999404229967,
      "grad_norm": 0.03125,
      "learning_rate": 4.4875666897714605e-06,
      "loss": 0.7839,
      "num_input_tokens_seen": 77415936,
      "step": 133340
    },
    {
      "epoch": 19.860738754840632,
      "grad_norm": 0.046142578125,
      "learning_rate": 4.440006397068985e-06,
      "loss": 0.7902,
      "num_input_tokens_seen": 77418944,
      "step": 133345
    },
    {
      "epoch": 19.86148346738159,
      "grad_norm": 0.05517578125,
      "learning_rate": 4.392699439109693e-06,
      "loss": 0.7902,
      "num_input_tokens_seen": 77422016,
      "step": 133350
    },
    {
      "epoch": 19.86222817992255,
      "grad_norm": 0.04296875,
      "learning_rate": 4.345645816696275e-06,
      "loss": 0.7973,
      "num_input_tokens_seen": 77424672,
      "step": 133355
    },
    {
      "epoch": 19.86297289246351,
      "grad_norm": 0.047607421875,
      "learning_rate": 4.298845530623096e-06,
      "loss": 0.7786,
      "num_input_tokens_seen": 77427616,
      "step": 133360
    },
    {
      "epoch": 19.863717605004467,
      "grad_norm": 0.0771484375,
      "learning_rate": 4.25229858167786e-06,
      "loss": 0.7882,
      "num_input_tokens_seen": 77430368,
      "step": 133365
    },
    {
      "epoch": 19.864462317545428,
      "grad_norm": 0.045654296875,
      "learning_rate": 4.206004970649934e-06,
      "loss": 0.7972,
      "num_input_tokens_seen": 77433312,
      "step": 133370
    },
    {
      "epoch": 19.86520703008639,
      "grad_norm": 0.033203125,
      "learning_rate": 4.159964698320362e-06,
      "loss": 0.7941,
      "num_input_tokens_seen": 77436128,
      "step": 133375
    },
    {
      "epoch": 19.865951742627345,
      "grad_norm": 0.045166015625,
      "learning_rate": 4.1141777654685185e-06,
      "loss": 0.8006,
      "num_input_tokens_seen": 77438944,
      "step": 133380
    },
    {
      "epoch": 19.866696455168306,
      "grad_norm": 0.032470703125,
      "learning_rate": 4.068644172865454e-06,
      "loss": 0.7883,
      "num_input_tokens_seen": 77441824,
      "step": 133385
    },
    {
      "epoch": 19.867441167709263,
      "grad_norm": 0.0322265625,
      "learning_rate": 4.023363921280554e-06,
      "loss": 0.7825,
      "num_input_tokens_seen": 77444576,
      "step": 133390
    },
    {
      "epoch": 19.868185880250223,
      "grad_norm": 0.049072265625,
      "learning_rate": 3.978337011479871e-06,
      "loss": 0.7748,
      "num_input_tokens_seen": 77447872,
      "step": 133395
    },
    {
      "epoch": 19.868930592791184,
      "grad_norm": 0.03369140625,
      "learning_rate": 3.933563444224464e-06,
      "loss": 0.8046,
      "num_input_tokens_seen": 77450976,
      "step": 133400
    },
    {
      "epoch": 19.86967530533214,
      "grad_norm": 0.03173828125,
      "learning_rate": 3.889043220270394e-06,
      "loss": 0.8025,
      "num_input_tokens_seen": 77453856,
      "step": 133405
    },
    {
      "epoch": 19.8704200178731,
      "grad_norm": 0.0252685546875,
      "learning_rate": 3.8447763403703924e-06,
      "loss": 0.7879,
      "num_input_tokens_seen": 77456736,
      "step": 133410
    },
    {
      "epoch": 19.87116473041406,
      "grad_norm": 0.040283203125,
      "learning_rate": 3.80076280527053e-06,
      "loss": 0.8064,
      "num_input_tokens_seen": 77459488,
      "step": 133415
    },
    {
      "epoch": 19.87190944295502,
      "grad_norm": 0.0654296875,
      "learning_rate": 3.75700261571521e-06,
      "loss": 0.7985,
      "num_input_tokens_seen": 77462464,
      "step": 133420
    },
    {
      "epoch": 19.87265415549598,
      "grad_norm": 0.05615234375,
      "learning_rate": 3.713495772443842e-06,
      "loss": 0.783,
      "num_input_tokens_seen": 77465280,
      "step": 133425
    },
    {
      "epoch": 19.873398868036936,
      "grad_norm": 0.10986328125,
      "learning_rate": 3.670242276190838e-06,
      "loss": 0.8052,
      "num_input_tokens_seen": 77467936,
      "step": 133430
    },
    {
      "epoch": 19.874143580577897,
      "grad_norm": 0.042236328125,
      "learning_rate": 3.6272421276889455e-06,
      "loss": 0.7802,
      "num_input_tokens_seen": 77470816,
      "step": 133435
    },
    {
      "epoch": 19.874888293118858,
      "grad_norm": 0.039306640625,
      "learning_rate": 3.5844953276609193e-06,
      "loss": 0.8064,
      "num_input_tokens_seen": 77473472,
      "step": 133440
    },
    {
      "epoch": 19.875633005659815,
      "grad_norm": 0.30078125,
      "learning_rate": 3.5420018768328453e-06,
      "loss": 0.8578,
      "num_input_tokens_seen": 77476864,
      "step": 133445
    },
    {
      "epoch": 19.876377718200775,
      "grad_norm": 0.05712890625,
      "learning_rate": 3.4997617759208174e-06,
      "loss": 0.8129,
      "num_input_tokens_seen": 77479744,
      "step": 133450
    },
    {
      "epoch": 19.877122430741732,
      "grad_norm": 0.0537109375,
      "learning_rate": 3.4577750256392648e-06,
      "loss": 0.8076,
      "num_input_tokens_seen": 77482400,
      "step": 133455
    },
    {
      "epoch": 19.877867143282693,
      "grad_norm": 0.052490234375,
      "learning_rate": 3.4160416266959536e-06,
      "loss": 0.7845,
      "num_input_tokens_seen": 77485600,
      "step": 133460
    },
    {
      "epoch": 19.878611855823653,
      "grad_norm": 0.04931640625,
      "learning_rate": 3.3745615797969863e-06,
      "loss": 0.792,
      "num_input_tokens_seen": 77488544,
      "step": 133465
    },
    {
      "epoch": 19.87935656836461,
      "grad_norm": 0.044677734375,
      "learning_rate": 3.333334885641803e-06,
      "loss": 0.8049,
      "num_input_tokens_seen": 77491360,
      "step": 133470
    },
    {
      "epoch": 19.88010128090557,
      "grad_norm": 0.2197265625,
      "learning_rate": 3.2923615449298447e-06,
      "loss": 0.8291,
      "num_input_tokens_seen": 77494112,
      "step": 133475
    },
    {
      "epoch": 19.88084599344653,
      "grad_norm": 0.0703125,
      "learning_rate": 3.2516415583505595e-06,
      "loss": 0.8075,
      "num_input_tokens_seen": 77497120,
      "step": 133480
    },
    {
      "epoch": 19.881590705987488,
      "grad_norm": 0.056640625,
      "learning_rate": 3.211174926591731e-06,
      "loss": 0.804,
      "num_input_tokens_seen": 77500000,
      "step": 133485
    },
    {
      "epoch": 19.88233541852845,
      "grad_norm": 0.064453125,
      "learning_rate": 3.1709616503394766e-06,
      "loss": 0.7872,
      "num_input_tokens_seen": 77502688,
      "step": 133490
    },
    {
      "epoch": 19.883080131069406,
      "grad_norm": 0.037841796875,
      "learning_rate": 3.1310017302715875e-06,
      "loss": 0.7859,
      "num_input_tokens_seen": 77505312,
      "step": 133495
    },
    {
      "epoch": 19.883824843610366,
      "grad_norm": 0.045654296875,
      "learning_rate": 3.0912951670625245e-06,
      "loss": 0.7942,
      "num_input_tokens_seen": 77508352,
      "step": 133500
    },
    {
      "epoch": 19.884569556151327,
      "grad_norm": 0.0732421875,
      "learning_rate": 3.051841961385082e-06,
      "loss": 0.7944,
      "num_input_tokens_seen": 77511584,
      "step": 133505
    },
    {
      "epoch": 19.885314268692284,
      "grad_norm": 0.030029296875,
      "learning_rate": 3.0126421139037295e-06,
      "loss": 0.8159,
      "num_input_tokens_seen": 77514592,
      "step": 133510
    },
    {
      "epoch": 19.886058981233244,
      "grad_norm": 0.048828125,
      "learning_rate": 2.9736956252812697e-06,
      "loss": 0.803,
      "num_input_tokens_seen": 77517408,
      "step": 133515
    },
    {
      "epoch": 19.886803693774205,
      "grad_norm": 0.07080078125,
      "learning_rate": 2.9350024961755095e-06,
      "loss": 0.8085,
      "num_input_tokens_seen": 77520256,
      "step": 133520
    },
    {
      "epoch": 19.88754840631516,
      "grad_norm": 0.033447265625,
      "learning_rate": 2.8965627272425906e-06,
      "loss": 0.7915,
      "num_input_tokens_seen": 77523360,
      "step": 133525
    },
    {
      "epoch": 19.888293118856122,
      "grad_norm": 0.0693359375,
      "learning_rate": 2.8583763191269984e-06,
      "loss": 0.7953,
      "num_input_tokens_seen": 77526016,
      "step": 133530
    },
    {
      "epoch": 19.88903783139708,
      "grad_norm": 0.047607421875,
      "learning_rate": 2.8204432724798776e-06,
      "loss": 0.7945,
      "num_input_tokens_seen": 77529248,
      "step": 133535
    },
    {
      "epoch": 19.88978254393804,
      "grad_norm": 0.056396484375,
      "learning_rate": 2.7827635879357215e-06,
      "loss": 0.7898,
      "num_input_tokens_seen": 77532096,
      "step": 133540
    },
    {
      "epoch": 19.890527256479,
      "grad_norm": 0.056640625,
      "learning_rate": 2.7453372661373486e-06,
      "loss": 0.7835,
      "num_input_tokens_seen": 77534720,
      "step": 133545
    },
    {
      "epoch": 19.891271969019957,
      "grad_norm": 0.0810546875,
      "learning_rate": 2.7081643077125906e-06,
      "loss": 0.7729,
      "num_input_tokens_seen": 77537536,
      "step": 133550
    },
    {
      "epoch": 19.892016681560918,
      "grad_norm": 0.06103515625,
      "learning_rate": 2.671244713289278e-06,
      "loss": 0.7942,
      "num_input_tokens_seen": 77540608,
      "step": 133555
    },
    {
      "epoch": 19.89276139410188,
      "grad_norm": 0.060302734375,
      "learning_rate": 2.634578483495242e-06,
      "loss": 0.7992,
      "num_input_tokens_seen": 77543360,
      "step": 133560
    },
    {
      "epoch": 19.893506106642835,
      "grad_norm": 0.035888671875,
      "learning_rate": 2.598165618946657e-06,
      "loss": 0.7803,
      "num_input_tokens_seen": 77546464,
      "step": 133565
    },
    {
      "epoch": 19.894250819183796,
      "grad_norm": 0.045654296875,
      "learning_rate": 2.562006120258031e-06,
      "loss": 0.8122,
      "num_input_tokens_seen": 77549376,
      "step": 133570
    },
    {
      "epoch": 19.894995531724753,
      "grad_norm": 0.0255126953125,
      "learning_rate": 2.5260999880422074e-06,
      "loss": 0.7947,
      "num_input_tokens_seen": 77552128,
      "step": 133575
    },
    {
      "epoch": 19.895740244265713,
      "grad_norm": 0.044189453125,
      "learning_rate": 2.4904472229053676e-06,
      "loss": 0.8036,
      "num_input_tokens_seen": 77555104,
      "step": 133580
    },
    {
      "epoch": 19.896484956806674,
      "grad_norm": 0.04541015625,
      "learning_rate": 2.4550478254503627e-06,
      "loss": 0.7935,
      "num_input_tokens_seen": 77557920,
      "step": 133585
    },
    {
      "epoch": 19.89722966934763,
      "grad_norm": 0.0419921875,
      "learning_rate": 2.4199017962750478e-06,
      "loss": 0.7992,
      "num_input_tokens_seen": 77560992,
      "step": 133590
    },
    {
      "epoch": 19.89797438188859,
      "grad_norm": 0.036865234375,
      "learning_rate": 2.385009135970617e-06,
      "loss": 0.8029,
      "num_input_tokens_seen": 77564032,
      "step": 133595
    },
    {
      "epoch": 19.898719094429552,
      "grad_norm": 0.0400390625,
      "learning_rate": 2.350369845129929e-06,
      "loss": 0.7891,
      "num_input_tokens_seen": 77567296,
      "step": 133600
    },
    {
      "epoch": 19.89946380697051,
      "grad_norm": 0.035400390625,
      "learning_rate": 2.3159839243358516e-06,
      "loss": 0.7932,
      "num_input_tokens_seen": 77570112,
      "step": 133605
    },
    {
      "epoch": 19.90020851951147,
      "grad_norm": 0.042236328125,
      "learning_rate": 2.2818513741712507e-06,
      "loss": 0.8027,
      "num_input_tokens_seen": 77572928,
      "step": 133610
    },
    {
      "epoch": 19.900953232052427,
      "grad_norm": 0.09228515625,
      "learning_rate": 2.247972195210668e-06,
      "loss": 0.8028,
      "num_input_tokens_seen": 77575520,
      "step": 133615
    },
    {
      "epoch": 19.901697944593387,
      "grad_norm": 0.039306640625,
      "learning_rate": 2.2143463880286427e-06,
      "loss": 0.7899,
      "num_input_tokens_seen": 77578272,
      "step": 133620
    },
    {
      "epoch": 19.902442657134348,
      "grad_norm": 0.036376953125,
      "learning_rate": 2.1809739531913894e-06,
      "loss": 0.8015,
      "num_input_tokens_seen": 77581344,
      "step": 133625
    },
    {
      "epoch": 19.903187369675305,
      "grad_norm": 0.054931640625,
      "learning_rate": 2.1478548912634566e-06,
      "loss": 0.8172,
      "num_input_tokens_seen": 77584352,
      "step": 133630
    },
    {
      "epoch": 19.903932082216265,
      "grad_norm": 0.0498046875,
      "learning_rate": 2.114989202806061e-06,
      "loss": 0.8096,
      "num_input_tokens_seen": 77587264,
      "step": 133635
    },
    {
      "epoch": 19.904676794757222,
      "grad_norm": 0.041259765625,
      "learning_rate": 2.0823768883704297e-06,
      "loss": 0.79,
      "num_input_tokens_seen": 77590112,
      "step": 133640
    },
    {
      "epoch": 19.905421507298183,
      "grad_norm": 0.0673828125,
      "learning_rate": 2.050017948511118e-06,
      "loss": 0.7912,
      "num_input_tokens_seen": 77592800,
      "step": 133645
    },
    {
      "epoch": 19.906166219839143,
      "grad_norm": 0.08984375,
      "learning_rate": 2.017912383772691e-06,
      "loss": 0.7753,
      "num_input_tokens_seen": 77595840,
      "step": 133650
    },
    {
      "epoch": 19.9069109323801,
      "grad_norm": 0.05029296875,
      "learning_rate": 1.9860601946997125e-06,
      "loss": 0.794,
      "num_input_tokens_seen": 77598752,
      "step": 133655
    },
    {
      "epoch": 19.90765564492106,
      "grad_norm": 0.052001953125,
      "learning_rate": 1.9544613818284207e-06,
      "loss": 0.7807,
      "num_input_tokens_seen": 77601632,
      "step": 133660
    },
    {
      "epoch": 19.90840035746202,
      "grad_norm": 0.055908203125,
      "learning_rate": 1.9231159456917222e-06,
      "loss": 0.8026,
      "num_input_tokens_seen": 77604320,
      "step": 133665
    },
    {
      "epoch": 19.909145070002978,
      "grad_norm": 0.045166015625,
      "learning_rate": 1.8920238868225248e-06,
      "loss": 0.8069,
      "num_input_tokens_seen": 77607040,
      "step": 133670
    },
    {
      "epoch": 19.90988978254394,
      "grad_norm": 0.05419921875,
      "learning_rate": 1.8611852057437427e-06,
      "loss": 0.8105,
      "num_input_tokens_seen": 77609856,
      "step": 133675
    },
    {
      "epoch": 19.910634495084896,
      "grad_norm": 0.044921875,
      "learning_rate": 1.8305999029766262e-06,
      "loss": 0.8026,
      "num_input_tokens_seen": 77612608,
      "step": 133680
    },
    {
      "epoch": 19.911379207625856,
      "grad_norm": 0.05712890625,
      "learning_rate": 1.8002679790374286e-06,
      "loss": 0.8042,
      "num_input_tokens_seen": 77615584,
      "step": 133685
    },
    {
      "epoch": 19.912123920166817,
      "grad_norm": 0.05615234375,
      "learning_rate": 1.7701894344390733e-06,
      "loss": 0.8014,
      "num_input_tokens_seen": 77618272,
      "step": 133690
    },
    {
      "epoch": 19.912868632707774,
      "grad_norm": 0.059814453125,
      "learning_rate": 1.7403642696911524e-06,
      "loss": 0.8065,
      "num_input_tokens_seen": 77621504,
      "step": 133695
    },
    {
      "epoch": 19.913613345248734,
      "grad_norm": 0.03564453125,
      "learning_rate": 1.7107924852949317e-06,
      "loss": 0.7958,
      "num_input_tokens_seen": 77624064,
      "step": 133700
    },
    {
      "epoch": 19.914358057789695,
      "grad_norm": 0.0380859375,
      "learning_rate": 1.681474081751677e-06,
      "loss": 0.7867,
      "num_input_tokens_seen": 77627168,
      "step": 133705
    },
    {
      "epoch": 19.915102770330652,
      "grad_norm": 0.059326171875,
      "learning_rate": 1.6524090595576579e-06,
      "loss": 0.8044,
      "num_input_tokens_seen": 77630048,
      "step": 133710
    },
    {
      "epoch": 19.915847482871612,
      "grad_norm": 0.06298828125,
      "learning_rate": 1.6235974192008174e-06,
      "loss": 0.8056,
      "num_input_tokens_seen": 77632768,
      "step": 133715
    },
    {
      "epoch": 19.91659219541257,
      "grad_norm": 0.048095703125,
      "learning_rate": 1.5950391611707636e-06,
      "loss": 0.8162,
      "num_input_tokens_seen": 77635968,
      "step": 133720
    },
    {
      "epoch": 19.91733690795353,
      "grad_norm": 0.0498046875,
      "learning_rate": 1.5667342859487787e-06,
      "loss": 0.799,
      "num_input_tokens_seen": 77638752,
      "step": 133725
    },
    {
      "epoch": 19.91808162049449,
      "grad_norm": 0.06298828125,
      "learning_rate": 1.5386827940128133e-06,
      "loss": 0.804,
      "num_input_tokens_seen": 77641728,
      "step": 133730
    },
    {
      "epoch": 19.918826333035447,
      "grad_norm": 0.0299072265625,
      "learning_rate": 1.5108846858358227e-06,
      "loss": 0.794,
      "num_input_tokens_seen": 77644672,
      "step": 133735
    },
    {
      "epoch": 19.919571045576408,
      "grad_norm": 0.041748046875,
      "learning_rate": 1.4833399618907617e-06,
      "loss": 0.7974,
      "num_input_tokens_seen": 77647808,
      "step": 133740
    },
    {
      "epoch": 19.92031575811737,
      "grad_norm": 0.0908203125,
      "learning_rate": 1.4560486226389279e-06,
      "loss": 0.7914,
      "num_input_tokens_seen": 77650848,
      "step": 133745
    },
    {
      "epoch": 19.921060470658325,
      "grad_norm": 0.06640625,
      "learning_rate": 1.4290106685449498e-06,
      "loss": 0.7939,
      "num_input_tokens_seen": 77653664,
      "step": 133750
    },
    {
      "epoch": 19.921805183199286,
      "grad_norm": 0.0257568359375,
      "learning_rate": 1.4022261000617986e-06,
      "loss": 0.8008,
      "num_input_tokens_seen": 77656352,
      "step": 133755
    },
    {
      "epoch": 19.922549895740243,
      "grad_norm": 0.0654296875,
      "learning_rate": 1.3756949176457755e-06,
      "loss": 0.8131,
      "num_input_tokens_seen": 77659264,
      "step": 133760
    },
    {
      "epoch": 19.923294608281203,
      "grad_norm": 0.029296875,
      "learning_rate": 1.3494171217431904e-06,
      "loss": 0.7999,
      "num_input_tokens_seen": 77662144,
      "step": 133765
    },
    {
      "epoch": 19.924039320822164,
      "grad_norm": 0.0849609375,
      "learning_rate": 1.3233927127986876e-06,
      "loss": 0.797,
      "num_input_tokens_seen": 77664960,
      "step": 133770
    },
    {
      "epoch": 19.92478403336312,
      "grad_norm": 0.053955078125,
      "learning_rate": 1.2976216912519155e-06,
      "loss": 0.7787,
      "num_input_tokens_seen": 77667936,
      "step": 133775
    },
    {
      "epoch": 19.92552874590408,
      "grad_norm": 0.0301513671875,
      "learning_rate": 1.272104057535861e-06,
      "loss": 0.7894,
      "num_input_tokens_seen": 77670816,
      "step": 133780
    },
    {
      "epoch": 19.92627345844504,
      "grad_norm": 0.064453125,
      "learning_rate": 1.2468398120851764e-06,
      "loss": 0.7892,
      "num_input_tokens_seen": 77673696,
      "step": 133785
    },
    {
      "epoch": 19.927018170986,
      "grad_norm": 0.03564453125,
      "learning_rate": 1.221828955324522e-06,
      "loss": 0.7809,
      "num_input_tokens_seen": 77676608,
      "step": 133790
    },
    {
      "epoch": 19.92776288352696,
      "grad_norm": 0.06494140625,
      "learning_rate": 1.1970714876768928e-06,
      "loss": 0.8027,
      "num_input_tokens_seen": 77679392,
      "step": 133795
    },
    {
      "epoch": 19.928507596067917,
      "grad_norm": 0.042236328125,
      "learning_rate": 1.172567409561953e-06,
      "loss": 0.7993,
      "num_input_tokens_seen": 77682176,
      "step": 133800
    },
    {
      "epoch": 19.929252308608877,
      "grad_norm": 0.04541015625,
      "learning_rate": 1.1483167213910405e-06,
      "loss": 0.7929,
      "num_input_tokens_seen": 77684736,
      "step": 133805
    },
    {
      "epoch": 19.929997021149838,
      "grad_norm": 0.03125,
      "learning_rate": 1.1243194235754926e-06,
      "loss": 0.7978,
      "num_input_tokens_seen": 77687584,
      "step": 133810
    },
    {
      "epoch": 19.930741733690795,
      "grad_norm": 0.05126953125,
      "learning_rate": 1.100575516519986e-06,
      "loss": 0.8087,
      "num_input_tokens_seen": 77690336,
      "step": 133815
    },
    {
      "epoch": 19.931486446231755,
      "grad_norm": 0.036865234375,
      "learning_rate": 1.077085000625866e-06,
      "loss": 0.7733,
      "num_input_tokens_seen": 77693280,
      "step": 133820
    },
    {
      "epoch": 19.932231158772712,
      "grad_norm": 0.048095703125,
      "learning_rate": 1.0538478762911475e-06,
      "loss": 0.7958,
      "num_input_tokens_seen": 77696096,
      "step": 133825
    },
    {
      "epoch": 19.932975871313673,
      "grad_norm": 0.03857421875,
      "learning_rate": 1.0308641439071841e-06,
      "loss": 0.8035,
      "num_input_tokens_seen": 77698912,
      "step": 133830
    },
    {
      "epoch": 19.933720583854633,
      "grad_norm": 0.05029296875,
      "learning_rate": 1.008133803861999e-06,
      "loss": 0.7822,
      "num_input_tokens_seen": 77701600,
      "step": 133835
    },
    {
      "epoch": 19.93446529639559,
      "grad_norm": 0.0615234375,
      "learning_rate": 9.856568565402845e-07,
      "loss": 0.8144,
      "num_input_tokens_seen": 77704448,
      "step": 133840
    },
    {
      "epoch": 19.93521000893655,
      "grad_norm": 0.039306640625,
      "learning_rate": 9.634333023217366e-07,
      "loss": 0.7913,
      "num_input_tokens_seen": 77707424,
      "step": 133845
    },
    {
      "epoch": 19.93595472147751,
      "grad_norm": 0.047607421875,
      "learning_rate": 9.414631415810559e-07,
      "loss": 0.7898,
      "num_input_tokens_seen": 77710112,
      "step": 133850
    },
    {
      "epoch": 19.93669943401847,
      "grad_norm": 0.048828125,
      "learning_rate": 9.197463746912771e-07,
      "loss": 0.8033,
      "num_input_tokens_seen": 77712992,
      "step": 133855
    },
    {
      "epoch": 19.93744414655943,
      "grad_norm": 0.045654296875,
      "learning_rate": 8.982830020171084e-07,
      "loss": 0.7878,
      "num_input_tokens_seen": 77715936,
      "step": 133860
    },
    {
      "epoch": 19.938188859100386,
      "grad_norm": 0.042236328125,
      "learning_rate": 8.770730239215929e-07,
      "loss": 0.7996,
      "num_input_tokens_seen": 77718880,
      "step": 133865
    },
    {
      "epoch": 19.938933571641346,
      "grad_norm": 0.0289306640625,
      "learning_rate": 8.561164407644428e-07,
      "loss": 0.8063,
      "num_input_tokens_seen": 77721952,
      "step": 133870
    },
    {
      "epoch": 19.939678284182307,
      "grad_norm": 0.05419921875,
      "learning_rate": 8.35413252898709e-07,
      "loss": 0.7995,
      "num_input_tokens_seen": 77724736,
      "step": 133875
    },
    {
      "epoch": 19.940422996723264,
      "grad_norm": 0.0458984375,
      "learning_rate": 8.149634606741119e-07,
      "loss": 0.8197,
      "num_input_tokens_seen": 77727648,
      "step": 133880
    },
    {
      "epoch": 19.941167709264224,
      "grad_norm": 0.07470703125,
      "learning_rate": 7.947670644353754e-07,
      "loss": 0.7888,
      "num_input_tokens_seen": 77730688,
      "step": 133885
    },
    {
      "epoch": 19.941912421805185,
      "grad_norm": 0.037109375,
      "learning_rate": 7.748240645255589e-07,
      "loss": 0.7872,
      "num_input_tokens_seen": 77733472,
      "step": 133890
    },
    {
      "epoch": 19.942657134346142,
      "grad_norm": 0.054443359375,
      "learning_rate": 7.551344612793942e-07,
      "loss": 0.7917,
      "num_input_tokens_seen": 77736416,
      "step": 133895
    },
    {
      "epoch": 19.943401846887102,
      "grad_norm": 0.052734375,
      "learning_rate": 7.356982550316137e-07,
      "loss": 0.8016,
      "num_input_tokens_seen": 77739104,
      "step": 133900
    },
    {
      "epoch": 19.94414655942806,
      "grad_norm": 0.051025390625,
      "learning_rate": 7.165154461102885e-07,
      "loss": 0.7889,
      "num_input_tokens_seen": 77741952,
      "step": 133905
    },
    {
      "epoch": 19.94489127196902,
      "grad_norm": 0.0654296875,
      "learning_rate": 6.975860348368278e-07,
      "loss": 0.7958,
      "num_input_tokens_seen": 77744832,
      "step": 133910
    },
    {
      "epoch": 19.94563598450998,
      "grad_norm": 0.08154296875,
      "learning_rate": 6.789100215343069e-07,
      "loss": 0.8002,
      "num_input_tokens_seen": 77747552,
      "step": 133915
    },
    {
      "epoch": 19.946380697050937,
      "grad_norm": 0.07958984375,
      "learning_rate": 6.604874065174737e-07,
      "loss": 0.8337,
      "num_input_tokens_seen": 77750432,
      "step": 133920
    },
    {
      "epoch": 19.947125409591898,
      "grad_norm": 0.0361328125,
      "learning_rate": 6.423181900960805e-07,
      "loss": 0.8061,
      "num_input_tokens_seen": 77753344,
      "step": 133925
    },
    {
      "epoch": 19.947870122132855,
      "grad_norm": 0.080078125,
      "learning_rate": 6.244023725782144e-07,
      "loss": 0.7879,
      "num_input_tokens_seen": 77756224,
      "step": 133930
    },
    {
      "epoch": 19.948614834673815,
      "grad_norm": 0.06494140625,
      "learning_rate": 6.06739954266966e-07,
      "loss": 0.775,
      "num_input_tokens_seen": 77758944,
      "step": 133935
    },
    {
      "epoch": 19.949359547214776,
      "grad_norm": 0.04150390625,
      "learning_rate": 5.893309354604304e-07,
      "loss": 0.791,
      "num_input_tokens_seen": 77761824,
      "step": 133940
    },
    {
      "epoch": 19.950104259755733,
      "grad_norm": 0.04052734375,
      "learning_rate": 5.721753164517063e-07,
      "loss": 0.7898,
      "num_input_tokens_seen": 77764672,
      "step": 133945
    },
    {
      "epoch": 19.950848972296694,
      "grad_norm": 0.031982421875,
      "learning_rate": 5.55273097530562e-07,
      "loss": 0.7974,
      "num_input_tokens_seen": 77767776,
      "step": 133950
    },
    {
      "epoch": 19.951593684837654,
      "grad_norm": 0.0517578125,
      "learning_rate": 5.386242789851004e-07,
      "loss": 0.7885,
      "num_input_tokens_seen": 77770720,
      "step": 133955
    },
    {
      "epoch": 19.95233839737861,
      "grad_norm": 0.0419921875,
      "learning_rate": 5.222288610934322e-07,
      "loss": 0.818,
      "num_input_tokens_seen": 77773504,
      "step": 133960
    },
    {
      "epoch": 19.95308310991957,
      "grad_norm": 0.03857421875,
      "learning_rate": 5.060868441336685e-07,
      "loss": 0.774,
      "num_input_tokens_seen": 77776800,
      "step": 133965
    },
    {
      "epoch": 19.95382782246053,
      "grad_norm": 0.05810546875,
      "learning_rate": 4.901982283805895e-07,
      "loss": 0.8278,
      "num_input_tokens_seen": 77779776,
      "step": 133970
    },
    {
      "epoch": 19.95457253500149,
      "grad_norm": 0.051513671875,
      "learning_rate": 4.7456301409898317e-07,
      "loss": 0.821,
      "num_input_tokens_seen": 77782720,
      "step": 133975
    },
    {
      "epoch": 19.95531724754245,
      "grad_norm": 0.05908203125,
      "learning_rate": 4.591812015553032e-07,
      "loss": 0.8058,
      "num_input_tokens_seen": 77785600,
      "step": 133980
    },
    {
      "epoch": 19.956061960083407,
      "grad_norm": 0.042236328125,
      "learning_rate": 4.440527910093417e-07,
      "loss": 0.774,
      "num_input_tokens_seen": 77788832,
      "step": 133985
    },
    {
      "epoch": 19.956806672624367,
      "grad_norm": 0.057861328125,
      "learning_rate": 4.2917778271589487e-07,
      "loss": 0.8118,
      "num_input_tokens_seen": 77791776,
      "step": 133990
    },
    {
      "epoch": 19.957551385165328,
      "grad_norm": 0.06201171875,
      "learning_rate": 4.145561769264283e-07,
      "loss": 0.7913,
      "num_input_tokens_seen": 77794720,
      "step": 133995
    },
    {
      "epoch": 19.958296097706285,
      "grad_norm": 0.09619140625,
      "learning_rate": 4.001879738890768e-07,
      "loss": 0.8092,
      "num_input_tokens_seen": 77797728,
      "step": 134000
    },
    {
      "epoch": 19.959040810247245,
      "grad_norm": 0.09326171875,
      "learning_rate": 3.8607317384531377e-07,
      "loss": 0.7951,
      "num_input_tokens_seen": 77800512,
      "step": 134005
    },
    {
      "epoch": 19.959785522788202,
      "grad_norm": 0.03125,
      "learning_rate": 3.722117770332822e-07,
      "loss": 0.8184,
      "num_input_tokens_seen": 77803136,
      "step": 134010
    },
    {
      "epoch": 19.960530235329163,
      "grad_norm": 0.042236328125,
      "learning_rate": 3.586037836877942e-07,
      "loss": 0.8001,
      "num_input_tokens_seen": 77806112,
      "step": 134015
    },
    {
      "epoch": 19.961274947870123,
      "grad_norm": 0.05908203125,
      "learning_rate": 3.4524919403866593e-07,
      "loss": 0.8042,
      "num_input_tokens_seen": 77808960,
      "step": 134020
    },
    {
      "epoch": 19.96201966041108,
      "grad_norm": 0.049072265625,
      "learning_rate": 3.321480083123829e-07,
      "loss": 0.789,
      "num_input_tokens_seen": 77811744,
      "step": 134025
    },
    {
      "epoch": 19.96276437295204,
      "grad_norm": 0.03466796875,
      "learning_rate": 3.1930022672876923e-07,
      "loss": 0.802,
      "num_input_tokens_seen": 77814848,
      "step": 134030
    },
    {
      "epoch": 19.963509085493,
      "grad_norm": 0.041259765625,
      "learning_rate": 3.0670584950598376e-07,
      "loss": 0.808,
      "num_input_tokens_seen": 77817952,
      "step": 134035
    },
    {
      "epoch": 19.96425379803396,
      "grad_norm": 0.054931640625,
      "learning_rate": 2.9436487685718935e-07,
      "loss": 0.794,
      "num_input_tokens_seen": 77820832,
      "step": 134040
    },
    {
      "epoch": 19.96499851057492,
      "grad_norm": 0.037841796875,
      "learning_rate": 2.8227730898888745e-07,
      "loss": 0.8008,
      "num_input_tokens_seen": 77823936,
      "step": 134045
    },
    {
      "epoch": 19.965743223115876,
      "grad_norm": 0.0257568359375,
      "learning_rate": 2.704431461059142e-07,
      "loss": 0.7845,
      "num_input_tokens_seen": 77826912,
      "step": 134050
    },
    {
      "epoch": 19.966487935656836,
      "grad_norm": 0.031494140625,
      "learning_rate": 2.588623884097751e-07,
      "loss": 0.8049,
      "num_input_tokens_seen": 77829888,
      "step": 134055
    },
    {
      "epoch": 19.967232648197797,
      "grad_norm": 0.06689453125,
      "learning_rate": 2.4753503609531434e-07,
      "loss": 0.788,
      "num_input_tokens_seen": 77832704,
      "step": 134060
    },
    {
      "epoch": 19.967977360738754,
      "grad_norm": 0.04931640625,
      "learning_rate": 2.364610893540453e-07,
      "loss": 0.8117,
      "num_input_tokens_seen": 77835648,
      "step": 134065
    },
    {
      "epoch": 19.968722073279714,
      "grad_norm": 0.1103515625,
      "learning_rate": 2.2564054837082015e-07,
      "loss": 0.8061,
      "num_input_tokens_seen": 77838688,
      "step": 134070
    },
    {
      "epoch": 19.969466785820675,
      "grad_norm": 0.04638671875,
      "learning_rate": 2.1507341333215635e-07,
      "loss": 0.7868,
      "num_input_tokens_seen": 77841632,
      "step": 134075
    },
    {
      "epoch": 19.970211498361632,
      "grad_norm": 0.09326171875,
      "learning_rate": 2.0475968441291404e-07,
      "loss": 0.7998,
      "num_input_tokens_seen": 77844544,
      "step": 134080
    },
    {
      "epoch": 19.970956210902592,
      "grad_norm": 0.0810546875,
      "learning_rate": 1.94699361791284e-07,
      "loss": 0.7966,
      "num_input_tokens_seen": 77847648,
      "step": 134085
    },
    {
      "epoch": 19.97170092344355,
      "grad_norm": 0.042236328125,
      "learning_rate": 1.848924456337997e-07,
      "loss": 0.7938,
      "num_input_tokens_seen": 77850432,
      "step": 134090
    },
    {
      "epoch": 19.97244563598451,
      "grad_norm": 0.05810546875,
      "learning_rate": 1.7533893610699457e-07,
      "loss": 0.7942,
      "num_input_tokens_seen": 77853440,
      "step": 134095
    },
    {
      "epoch": 19.97319034852547,
      "grad_norm": 0.0390625,
      "learning_rate": 1.6603883337407143e-07,
      "loss": 0.7989,
      "num_input_tokens_seen": 77856192,
      "step": 134100
    },
    {
      "epoch": 19.973935061066427,
      "grad_norm": 0.05322265625,
      "learning_rate": 1.5699213758990638e-07,
      "loss": 0.7751,
      "num_input_tokens_seen": 77859392,
      "step": 134105
    },
    {
      "epoch": 19.974679773607388,
      "grad_norm": 0.053955078125,
      "learning_rate": 1.4819884890771017e-07,
      "loss": 0.8192,
      "num_input_tokens_seen": 77862304,
      "step": 134110
    },
    {
      "epoch": 19.97542448614835,
      "grad_norm": 0.04052734375,
      "learning_rate": 1.3965896747736295e-07,
      "loss": 0.8138,
      "num_input_tokens_seen": 77865344,
      "step": 134115
    },
    {
      "epoch": 19.976169198689306,
      "grad_norm": 0.05517578125,
      "learning_rate": 1.3137249344208344e-07,
      "loss": 0.7975,
      "num_input_tokens_seen": 77868320,
      "step": 134120
    },
    {
      "epoch": 19.976913911230266,
      "grad_norm": 0.0361328125,
      "learning_rate": 1.2333942694175982e-07,
      "loss": 0.7987,
      "num_input_tokens_seen": 77871456,
      "step": 134125
    },
    {
      "epoch": 19.977658623771223,
      "grad_norm": 0.051025390625,
      "learning_rate": 1.1555976811294942e-07,
      "loss": 0.802,
      "num_input_tokens_seen": 77874656,
      "step": 134130
    },
    {
      "epoch": 19.978403336312184,
      "grad_norm": 0.04541015625,
      "learning_rate": 1.0803351708721375e-07,
      "loss": 0.8094,
      "num_input_tokens_seen": 77877536,
      "step": 134135
    },
    {
      "epoch": 19.979148048853144,
      "grad_norm": 0.0712890625,
      "learning_rate": 1.0076067398945287e-07,
      "loss": 0.8023,
      "num_input_tokens_seen": 77880704,
      "step": 134140
    },
    {
      "epoch": 19.9798927613941,
      "grad_norm": 0.04150390625,
      "learning_rate": 9.37412389462322e-08,
      "loss": 0.7932,
      "num_input_tokens_seen": 77883840,
      "step": 134145
    },
    {
      "epoch": 19.98063747393506,
      "grad_norm": 0.0238037109375,
      "learning_rate": 8.697521207245984e-08,
      "loss": 0.7978,
      "num_input_tokens_seen": 77886720,
      "step": 134150
    },
    {
      "epoch": 19.98138218647602,
      "grad_norm": 0.056640625,
      "learning_rate": 8.046259348304385e-08,
      "loss": 0.7722,
      "num_input_tokens_seen": 77889664,
      "step": 134155
    },
    {
      "epoch": 19.98212689901698,
      "grad_norm": 0.0498046875,
      "learning_rate": 7.420338329122699e-08,
      "loss": 0.7916,
      "num_input_tokens_seen": 77892448,
      "step": 134160
    },
    {
      "epoch": 19.98287161155794,
      "grad_norm": 0.048583984375,
      "learning_rate": 6.819758159859468e-08,
      "loss": 0.7972,
      "num_input_tokens_seen": 77895072,
      "step": 134165
    },
    {
      "epoch": 19.983616324098897,
      "grad_norm": 0.048828125,
      "learning_rate": 6.244518851006297e-08,
      "loss": 0.8127,
      "num_input_tokens_seen": 77897792,
      "step": 134170
    },
    {
      "epoch": 19.984361036639857,
      "grad_norm": 0.04736328125,
      "learning_rate": 5.6946204120555954e-08,
      "loss": 0.8112,
      "num_input_tokens_seen": 77900672,
      "step": 134175
    },
    {
      "epoch": 19.985105749180818,
      "grad_norm": 0.0947265625,
      "learning_rate": 5.1700628523332346e-08,
      "loss": 0.8235,
      "num_input_tokens_seen": 77903488,
      "step": 134180
    },
    {
      "epoch": 19.985850461721775,
      "grad_norm": 0.039794921875,
      "learning_rate": 4.670846180832022e-08,
      "loss": 0.7935,
      "num_input_tokens_seen": 77906336,
      "step": 134185
    },
    {
      "epoch": 19.986595174262735,
      "grad_norm": 0.037353515625,
      "learning_rate": 4.1969704057120967e-08,
      "loss": 0.8051,
      "num_input_tokens_seen": 77908960,
      "step": 134190
    },
    {
      "epoch": 19.987339886803692,
      "grad_norm": 0.034912109375,
      "learning_rate": 3.7484355353001316e-08,
      "loss": 0.8039,
      "num_input_tokens_seen": 77911872,
      "step": 134195
    },
    {
      "epoch": 19.988084599344653,
      "grad_norm": 0.03662109375,
      "learning_rate": 3.3252415770901324e-08,
      "loss": 0.8109,
      "num_input_tokens_seen": 77914496,
      "step": 134200
    },
    {
      "epoch": 19.988829311885613,
      "grad_norm": 0.039794921875,
      "learning_rate": 2.9273885380765028e-08,
      "loss": 0.7852,
      "num_input_tokens_seen": 77917472,
      "step": 134205
    },
    {
      "epoch": 19.98957402442657,
      "grad_norm": 0.037841796875,
      "learning_rate": 2.5548764250871158e-08,
      "loss": 0.8004,
      "num_input_tokens_seen": 77920352,
      "step": 134210
    },
    {
      "epoch": 19.99031873696753,
      "grad_norm": 0.039306640625,
      "learning_rate": 2.2077052442837086e-08,
      "loss": 0.785,
      "num_input_tokens_seen": 77923168,
      "step": 134215
    },
    {
      "epoch": 19.99106344950849,
      "grad_norm": 0.04736328125,
      "learning_rate": 1.885875001828019e-08,
      "loss": 0.803,
      "num_input_tokens_seen": 77926368,
      "step": 134220
    },
    {
      "epoch": 19.99180816204945,
      "grad_norm": 0.060791015625,
      "learning_rate": 1.5893857028825842e-08,
      "loss": 0.8026,
      "num_input_tokens_seen": 77929440,
      "step": 134225
    },
    {
      "epoch": 19.99255287459041,
      "grad_norm": 0.08544921875,
      "learning_rate": 1.3182373526099411e-08,
      "loss": 0.8162,
      "num_input_tokens_seen": 77932192,
      "step": 134230
    },
    {
      "epoch": 19.993297587131366,
      "grad_norm": 0.0439453125,
      "learning_rate": 1.072429955506493e-08,
      "loss": 0.7904,
      "num_input_tokens_seen": 77935136,
      "step": 134235
    },
    {
      "epoch": 19.994042299672326,
      "grad_norm": 0.038330078125,
      "learning_rate": 8.519635157355765e-09,
      "loss": 0.81,
      "num_input_tokens_seen": 77938048,
      "step": 134240
    },
    {
      "epoch": 19.994787012213287,
      "grad_norm": 0.0458984375,
      "learning_rate": 6.5683803712746065e-09,
      "loss": 0.8076,
      "num_input_tokens_seen": 77940864,
      "step": 134245
    },
    {
      "epoch": 19.995531724754244,
      "grad_norm": 0.03076171875,
      "learning_rate": 4.8705352284628135e-09,
      "loss": 0.7787,
      "num_input_tokens_seen": 77943904,
      "step": 134250
    },
    {
      "epoch": 19.996276437295204,
      "grad_norm": 0.05078125,
      "learning_rate": 3.4260997572310714e-09,
      "loss": 0.8164,
      "num_input_tokens_seen": 77946912,
      "step": 134255
    },
    {
      "epoch": 19.997021149836165,
      "grad_norm": 0.06640625,
      "learning_rate": 2.2350739842247333e-09,
      "loss": 0.7934,
      "num_input_tokens_seen": 77949824,
      "step": 134260
    },
    {
      "epoch": 19.997765862377122,
      "grad_norm": 0.0546875,
      "learning_rate": 1.2974579277624798e-09,
      "loss": 0.7855,
      "num_input_tokens_seen": 77953024,
      "step": 134265
    },
    {
      "epoch": 19.998510574918082,
      "grad_norm": 0.05810546875,
      "learning_rate": 6.132516028323209e-10,
      "loss": 0.8183,
      "num_input_tokens_seen": 77956032,
      "step": 134270
    },
    {
      "epoch": 19.99925528745904,
      "grad_norm": 0.040283203125,
      "learning_rate": 1.8245502275693325e-10,
      "loss": 0.7991,
      "num_input_tokens_seen": 77959072,
      "step": 134275
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.0693359375,
      "learning_rate": 5.068195862989455e-12,
      "loss": 0.8063,
      "num_input_tokens_seen": 77961608,
      "step": 134280
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.7992360591888428,
      "eval_runtime": 70.5641,
      "eval_samples_per_second": 42.288,
      "eval_steps_per_second": 10.572,
      "num_input_tokens_seen": 77961608,
      "step": 134280
    },
    {
      "epoch": 20.0,
      "num_input_tokens_seen": 77961608,
      "step": 134280,
      "total_flos": 3.5105759720879555e+18,
      "train_loss": 0.8024387844978857,
      "train_runtime": 29897.2604,
      "train_samples_per_second": 17.964,
      "train_steps_per_second": 4.491
    }
  ],
  "logging_steps": 5,
  "max_steps": 134280,
  "num_input_tokens_seen": 77961608,
  "num_train_epochs": 20,
  "save_steps": 6714,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.5105759720879555e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}